9.12 Gastrointestinal disease: public health aspects

9.12 Gastrointestinal disease: public health aspects
Oxford Textbook of Public Health

Gastrointestinal disease: public health aspects

R.F.A. Logan, M.J.G. Farthing, and M.J.S. Langman

Intestinal infection

Methods of assessment

Disease patterns

Factors influencing disease frequency

Peptic ulcer

Methods of assessment

Factors influencing disease frequency or severity
Inflammatory bowel disease

Methods of assessment

Factors affecting disease patterns
Gallstones (cholelithiasis)

Method of assessment

Disease patterns

Factors influencing disease frequency

Prospects for prevention

Disease patterns

Aetiological factors

Prospects for prevention
Diverticular disease of the colon

Disease patterns

Factors influencing disease frequency

Prospects for intervention
Coeliac disease

Methods of assessment

Patterns of disease

Factors influencing disease frequency


Factors influencing disease frequency
Liver disease

Cirrhosis of the liver

Prospects for prevention
Prospects for the control of gastrointestinal disease in the new millennium
Chapter References

This chapter considers the prospects for preventing infective and chronic digestive disease; cancer is discussed elsewhere.
Chronic gastrointestinal illness presents particular problems for applying public health measures to reduce the risk of contracting disease or ameliorating its impact. Our understanding of causes is rudimentary. The scarcity of reliable comparative data makes it difficult to decide if particular illnesses present special problems in one place or seldom cause significant disability elsewhere. Furthermore, the variety and precision of diagnostic measures are changing rapidly. Therefore it would be unwise to assume that illness, which in the past was primarily diagnosed by radiographic means but now is being increasingly assessed by fibre-optic methods, will necessarily represent the same range and proportions of severity in both periods.
If, despite these problems, the value of public health measures in prevention is to be examined, the strengths and frailties of individual indicators of disease presence and activity should be assessed. Thus for each disease where possible we have (i) reviewed the methods of measuring disease frequency before considering (ii) what is known of the environmental factors influencing occurrence and (iii) what prospects exist for prevention or improved control.
Intestinal infection
Infection is the most common affliction of the gastrointestinal tract worldwide. The major burden of intestinal infections is borne by individuals living in the developing world, where up to 4 million preschool children still die each year of acute dehydrating diarrhoeal disease of predominantly infectious aetiology (Farthing et al. 1993). Although prevalence is higher in the developing world, intestinal infection is now increasingly recognized in the majority of industrialized countries in which foodborne, and to a lesser extent waterborne infections are among the most common. Acute diarrhoea due to certain viruses is aetiologically important in infants and young children, although the mortality rate is extremely low. The vast majority of acute intestinal infections resolve without specific therapy providing there is adequate replacement of fluid and electrolyte losses (Farthing 1994). However, certain infections such as dysenteric shigellosis, invasive bacterial diarrhoea with systemic features, and some parasitic infections such as amoebiasis and giardiasis, do require specific antibiotic therapy (Farthing et al. 1992). Even chronic infections such as tuberculosis and schistosomiasis are amenable to drug therapy. Thus, although intestinal infections are numerically important in terms of disease morbidity worldwide and still produce an unacceptably high mortality in infants and preschool children, most are self-limiting and with appropriate public health interventions, the majority are preventable. Thus, intestinal infection continues to challenge clinicians and public health services; control and prevention of these diseases are achievable on the basis of current knowledge, providing appropriate infrastructures are put in place.
Methods of assessment
It is assumed that the local prevalence of intestinal infection is underestimated worldwide. There are marked geographic differences in the structures in place to collect such information, and even in those countries that attempt such an exercise, the quality of data collected is also variable. On a worldwide basis, the World Health Organization collects information on diarrhoeal diseases and publishes this on a yearly basis as a function of the Diarrhoeal Diseases Control Programme. In the United States, the Centers for Disease Control, Atlanta, produces regular reports on infectious diarrhoeal disease based on notifications and in-depth investigation of outbreaks. These processes are, of course, highly selective and only partially reflect the overall burden of infectious diarrhoeal disease. In the United Kingdom, the Communicable Disease Surveillance Centre, Colindale, functions in a similar way to the Centers for Disease Control and publishes the number of reported cases of specific infections on a yearly basis. Cholera, dysentery, food poisoning (with or without microbiological confirmation), typhoid fever, and tuberculosis are all statutorily notifiable infections in the United Kingdom. The Communicable Disease Surveillance Centre data for certain intestinal infections from 1980 to 1998 is shown in Fig. 1. The rise in Salmonella species and Campylobacter species infections, for example, is thought to be due to a true increase in the incidence of these diseases, rather than more vigorous reporting or efficient data collection. However, only a proportion of intestinal infections are notified to the Communicable Disease Surveillance Centre, and a substantial proportion are unlikely to even come to the notice of a general practitioner. A United Kingdom national study supported by the Department of Health has examined the background prevalence of a broad range of enteropathogens within the general practice setting in patients attending with and without diarrhoeal disease (Wheeler et al. 1999). The incidence of infectious intestinal disease was 19.4 per 100 person-years; however, only one in six people presented to a general practitioner. The ratio of cases in the community to cases reaching national surveillance was lower for bacterial pathogens than for viruses. Thus, the study indicates that there are about 9.4 million cases of infective diarrhoea each year, although current surveillance methods markedly underestimate the size of the problem.

Fig. 1 Laboratory reports of selected gastrointestinal infections in England and Wales, 1980 to 1998.

Disease patterns
Clinically, gastrointestinal infection can be classified on the basis of symptoms: (i) acute watery diarrhoea, (ii) diarrhoea with blood (dysentery), and (iii) chronic diarrhoea with or without evidence of steatorrhoea. The clinical presentation can often give an indication of the type of enteropathogen involved (Table 1) although this approach is by no means totally reliable. Acute watery diarrhoea is usually self-limiting and, provided that the illness begins to resolve within 5 to 7 days, no specific investigation or treatment is routinely required if fluid and electrolyte losses are replaced appropriately (Farthing 1994). However, it is wise to identify the enteropathogens responsible for dysentery and chronic diarrhoea as specific antimicrobial chemotherapy may be required. The prevalence of certain bacterial intestinal infections is increasing in the United Kingdom according to Communicable Disease Surveillance Centre data (Fig. 1). Salmonella species and Campylobacter jejuni dominate the bacterial causes of diarrhoea, a fact which has been attributed largely to the transmission of these organisms in food such as chickens and eggs (Telzak et al. 1990). Enterohaemorrhagic Escherichia coli, particularly serotype O157:H7 (also known as verocytotoxin-producing E. coli) has been responsible for a large number of community outbreaks of diarrhoea and dysentery associated with a significant mortality; major complications responsible for death include the haemolytic uraemic syndrome and thrombotic thrombocytopenic purpura.

Table 1 Clinical patterns and aetiology of infective diarrhoea

Prevalence data on enteric viral infections is sparse because of the difficulties in making a firm diagnosis. However, clinical evidence would suggest that these infections (rotavirus and enteric adenovirus types 40 and 41) are extremely common in infants and young children, and worldwide are now the most common causes of acute diarrhoea in this age group. The small round structured viruses such as the Norwalk family of viruses are also common, but produce a milder illness often with vomiting as a major component. These infections affect adults and there is an extremely high secondary attack rate producing outbreaks in families and in other situations where humans live or work in close proximity.
Intestinal infections are generally transmitted by the faecal–oral route. The organism itself or a protected, dormant form of the organism (cyst or oocyst) is excreted in faeces which can then be transmitted to other humans by (i) direct person-to-person contact or (ii) food and/or water. Thus, for many human intestinal infections, humans remain the major reservoirs. Certain infections such as amoebiasis and shigellosis are exclusively human pathogens, although under certain experimental circumstances, models of infection can be created in animals. However, some human enteropathogens, notably Salmonella species and Campylobacter jejuni, have large animal reservoirs, and thus these infections fulfil the criteria for a zoonosis. Similarly, certain parasitic infections such as cryptosporidiosis and possibly giardiasis are also carried by a variety of wild and domestic animals and can produce clinical disease. There is increasing concern that these animal reservoirs may be responsible for contamination of surface domestic water supplies and food products (Baird-Parker 1990). Molecular analysis has shown that there are at least two distinct strains of Cryptosporidium parvum: one which is only isolated from humans and does not colonize animals (type H), and another which is isolated from cattle but is also able to cause diarrhoea in humans (type C). Contamination of water supplies with cysts of Cryptosporidium parvum or Giardia intestinalis presents a major problem for domestic water suppliers since the cysts can survive for weeks or months in cool water and are unreliably inactivated or destroyed by the concentrations of chlorine that will render water acceptable with respect to coliform count (Porter et al. 1988; Sorvillo et al. 1992). There is some evidence to suggest that some enteric viruses such as small round structured viruses and possibly rotavirus are transmitted by aerosol created during vomiting (Reid et al. 1988). In this case, viral particles would be inhaled into the pharynx and then swallowed.
Enteropathogens vary in their prevalence in different seasons in the year. The majority of bacterial enteropathogens do not show marked variations in seasonality although in the developing world cholera outbreaks generally occur during times of heavy rainfall and flooding. Intestinal parasitic infections also tend to be more common in the cooler, wetter months of the year, situations which tend to favour the survival of cysts in the environment. Some enteric virus infections demonstrate seasonality with a dramatic increase in the cases of rotavirus infection during the winter months and a marked fall during the warm, dry summer months (Cukor and Blacklow 1984). However, enteric adenoviruses produce a similar number of cases throughout the year.
Geographic location
Intestinal infections are more common in developing countries than in the industrialized world where there has been more investment in sanitation and domestic water supplies. Other variables which increase the prevalence of intestinal infection include water shortages, which diminish the opportunity for maintenance of personal hygiene, crowded living conditions, the use of ‘night soil’ (human faeces) for fertilizing crops, and the proximity of humans to their domestic and farm animals. Certain infections such as cholera and those due to the intestinal helminths, hookworm, schistosomiasis, Strongyloides stercoralis, and others are limited almost exclusively to the tropics and subtropics and are usually only seen in the industrialized world as imported infections. The developing world thus carries the major burden of enteric infection, both with respect to case load and the spectrum of enteropathogens. Nevertheless, there is real concern that there has been true increase in the prevalence of intestinal infections in the industrialized world because of the increasing importance of foodborne disease and a variety of other risk factors outlined below.
Factors influencing disease frequency
Travel, immunodeficiency states, institutional residence or day-care attendance, reduced gastric-acid secretion, and extremes of age are the main factors known to increase susceptibility to intestinal infection. Travellers from the industrialized to the developing world are well recognized to be at risk of many of the infections shown in Table 1, although by far the most common is that due to enterotoxigenic E. coli (Farthing et al. 1992, 1993). Individuals with inherited or acquired immunodeficiency, particularly that relating to infection with HIV, are more commonly affected by the conventional bacterial enteropathogens and have a special susceptibility to the intracellular protozoa, Cryptosporidium parvum and Microsporidium species. These often produce chronic debilitating illnesses which are poorly responsive to medical therapy. Those at the extremes of age appear to be more susceptible to infection, presumably because of impaired immunological defences and the reduced ability to cope physiologically with the losses of fluid and electrolytes; morbidity from infection is increased in these individuals. People living in institutions or children attending day-centres appear to be more susceptible to enteric infections, largely because of the ease with which infection can be transmitted to large numbers of individuals living in close proximity to one another. This is particularly evident in shigellosis and giardiasis in which the infective dose may be as few as 10 organisms. The increased movement of individuals from the developing world to the industrialized countries has brought with it diseases that have previously not assumed great importance to those living in the developed world. Intestinal tuberculosis, for example, has increased dramatically in the United Kingdom, due almost exclusively to the increasing number of immigrants from the Indian subcontinent (Cook 1985). Similarly, some years ago a substantial number of cases of intestinal schistosomiasis was reported in New York City, again due exclusively to immigrants from the Caribbean. Finally, there is evidence that reduction in the concentration of gastric acid, either pharmacologically or following gastric surgery, results in an increased susceptibility to enteric bacterial infections. This has been demonstrated with Salmonella species where it appears that the elderly population are particularly at risk (Neal et al. 1994).
Other environmental and behavioural factors are also emerging as potential risk factors for the acquisition of intestinal infection. Seawater, freshwater lakes, and swimming pools have all been associated with the acquisition of intestinal infection, notably Cryptosporidium parvum, Giardia intestinalis and also enterohaemorrhagic E. coli (Walker 1992). Recent epidemiological analyses indicate that the relative risk of acquiring an infection is closely related to whether the swimmer swallows any of the water. Changes in eating habits in the West may also account for part of the increase in foodborne diarrhoeal disease. This includes the ingestion of raw fish, and of foods containing uncooked or partially cooked eggs such as mayonnaise, mousse, and certain Italian desserts.
Primary prevention of gastrointestinal infection can be achieved by interruption of the faecal–oral transmission route. This involves the widespread availability of high quality potable water, adequate sanitation to ensure safe disposal of faeces, and clear guidelines on personal hygiene to minimize person-to-person transmission. The importance of these measures is emphasized by the frequent rapid development of epidemics of gastrointestinal infection in areas where the appropriate infrastructure is disrupted by natural disasters or by war. In addition, guidelines and, where necessary, legislation are required to ensure the highest standards in animal husbandry, food production, and subsequent food handling with regular surveillance procedures to ensure that these standards are maintained.
There is compelling evidence that broad-spectrum antibiotics, taken at approximately half the therapeutic dose, can prevent certain bacterial infections of the intestine, notably the organisms that are responsible for travellers’ diarrhoea and cholera (Farthing et al. 1992). Their use in travellers’ diarrhoea remains highly controversial but, while not generally recommended, may be of value in certain situations, particularly in high-risk individuals. However, tetracycline is recommended for prophylaxis in family members during cholera outbreaks. Non-antibiotic approaches such as bismuth preparations have also been used to prevent travellers’ diarrhoea, the most widely used being bismuth subsalicylate. This compound is less effective than broad-spectrum antibiotics but is free of the potential adverse effects of the latter, and does not increase the emergence of antibiotic resistant bacteria.
The concept that the gut can be colonized with harmless bacteria that will protect against the effects of bacterial and possibly viral enteropathogens has been considered for several decades. Lactobacilli, particularly those used to produce yoghurt, have been used empirically for this purpose, although the evidence of efficacy is extremely poor (Oksanen et al. 1990) or non-existent (Katelaris et al. 1995). However, a recent study using Bifidobacterium bifidium and Streptococcus thermophilus showed that these organisms do appear to protect infants from rotavirus infection (Saavedra et al. 1994). Further evidence is awaited to confirm the validity of this approach in the prevention of other intestinal infections.
Although parenteral vaccines have been available for cholera and typhoid for many years, their efficacy is low. There have been major difficulties in developing effective oral vaccines but this approach is generally regarded as being the most effective for the production of protective immune responses in the gut, the site where they are most needed (Levine 1990). A whole cell-B subunit oral cholera vaccine has been subjected to extensive field trials and has been shown to produce moderately good but relatively short-lasting immunity to cholera (Peltola et al. 1991). More recently, a genetically engineered live oral cholera vaccine has been developed which appears to be as effective in a single dose, producing good protection without diarrhoea (Suharyono et al. 1992). There is some evidence to suggest that these cholera vaccines also protect against travellers’ diarrhoea caused by other organisms and thus, their application may be wider than was initially envisaged (Peltola et al. 1991). Vaccines are also under development for Salmonella species, Shigella species, and enterotoxigenic E. coli. There have also been extensive efforts to develop a rotavirus vaccine that will be safe, effective, and affordable in the developing world. A ‘tetravalent’ rhesus assortment vaccine containing an RNA segment which encodes the major neutralization antigen of sertype 1, 2, or 4 human rotavirus has been subjected to field trials in the developing world and in the United States. Although there was initially great optimism, there is now evidence that this vaccine causes intussusception in well-nourished children in the United States, possibly as the result of an exaggerated immune response in Peyer’s patches.
Peptic ulcer
Disability due to gastric or duodenal ulceration presents a major health problem in Western populations, where 10 per cent or more may be affected at some time in their lives. It is also a problem in many less-developed areas. Ulcer frequency varies greatly from time to time, and from place to place, reflecting the interplay of environmental and genetic influences, with environmental factors being of critical importance.
Methods of assessment
Mortality rates
Death usually occurs from ulcer complications or as a sequel to an operation. Few die as a result of a peptic ulcer, but mortality rates rise steeply with advancing age. In the United Kingdom the chance of dying from peptic ulcer is several hundred times greater in people over 65 years of age than in those under 35 (Table 2).

Table 2 Ulcer of the stomach and duodenum 1991: death rates per million population for England and Wales

Those who die from ulcer are likely to do so because of the severity of complications, or because of innate inability to withstand the stresses of illness because of age or coincident disease.
Autopsy surveys
Careful scrutiny of the stomach and duodenum at autopsy can give reliable information about the current frequency of ulcer, and about past frequency as judged by stigmata of previous surgery. However, it may be difficult to decide if there is duodenal or gastric scarring due to previous ulceration, and individual judgement may vary. Matters are complicated in two more ways. First, the group of people in whom autopsy examination is performed do not necessarily form a reasonable sample of those dying. Second, the autopsy rate in many countries is falling steadily, and it would be dangerous to assume that the findings at a certain time when, say, two-thirds of all those dying in hospital were subjected to autopsy, can be simply related to findings at a time when a third or less are being so examined.
Hospital admission rates
Provided that hospital diagnostic indices are reliably compiled, admission rates can act as proxies for measures of disease occurrence. Data on peptic ulceration must be interpreted with particular caution. The introduction of highly effective pharmacological treatments has for all practical purposes eliminated the need for hospital admission for medical treatment, and elective surgery is now seldom needed.
Ulcer complication rates present attractions as measures. Perforation and severe haemorrhage are clinically obvious, and are mandatory reasons for hospital admission. However, in making comparisons it should be noted that the proportions of ulcers which bleed or perforate may vary from time to time, and from place to place.
Admission rates with ulcer complications are distinctly higher in the elderly than in the young in the United Kingdom and the United States, a pattern which may not necessarily be followed in non-European countries. Data obtained in the United Kingdom (Walt et al. 1986) and elsewhere suggest that ulcer-perforation rates have risen steeply in the elderly, particularly in women, and may have fallen greatly in the young (Table 3). A small proportion of the rise in the elderly may be attributable to the use of non-steroidal anti-inflammatory drugs, but most is not. One attractive hypothesis is that the elderly form a cohort exposed in youth to Helicobacter pylori who contracted infection and hence a lifelong susceptibility to ulcer. If this is true then it presupposes that those born more recently have for some reason failed to become infected. Interacting influences are also important because many more, amongst non-takers of non-steroidal anti-inflammatory drugs, become infected with H. pylori than develop peptic ulcers. Factors to be considered must plainly include smoking, although the elderly are not, in general, a group of heavy smokers. Outside predisposing factors remain poorly understood.

Table 3 Admission rates for gastric and duodenal ulcer perforation in England and Wales and the United States (per 100 000 population per year)

Diagnostic rates
Very few investigators have examined the overall frequency with which diagnoses are made over time, almost certainly because of doubts over the effects of changes in the type and intensity of application of diagnostic methods.
Sickness certification
Certified sickness rates are unsatisfactory proxies for true disease occurrence when symptoms are poor guides to the presence of specific disease, and when functional abdominal symptoms are very common.
Taking all these factors into account, the best comparator indices from time to time and place to place must be complication and death rates. Direct diagnostic figures are plainly the best measures in individual studies of, for instance, risk associated with infection or drug use, and can make valuable contributions within well-described stable study areas (Primatesta et al. 1994).
Factors influencing disease frequency or severity
The two influences which have emerged as being of particular importance in affecting ulcer occurrence or the development of complications have been infection with H. pylori, and use of non-steroidal anti-inflammatory drugs.
Helicobacter pylori
Descriptions of gastric antral colonization by spiral-shaped organisms originally described as Campylobacter, and later recategorized as H. pylori, have been amply confirmed. The organism is an almost constant concomitant of duodenal ulceration in those who do not use non-steroidal anti-inflammatory drugs, it is detectable in a majority of those with gastric ulceration and a high proportion of those with antral gastric cancer, and it can also be found in many normal individuals who have, or tend to have, antral gastritis but not ulceration (Taylor and Blaser 1991; Labenz and Borsch 1994; Nomura et al. 1994). The evidence for a causal role in peptic, particularly duodenal, ulceration depends upon two main features. Firstly there is the almost constant association between ulcer occurrence and the presence of infection, and secondly eradication of infection by antibiotic therapy results in prolonged ulcer remission.
Unsolved questions include lack of understanding why infection in many people causes an acute or chronic gastritis, but no ulceration. This lack of ulceration does not seem wholly accounted for by the absence of other risk factors, notably smoking.
Although Helicobacter infection appears to be a substantial risk factor for gastric carcinoma, and distal carcinoma in particular, proximal disease and gastro-oesophageal reflux do not seem to be related, and attention has been drawn to opposing trends in the occurrence of peptic ulcer (falling) and of reflux disease (rising) in the United States (El Serag and Sonnenberg 1998).
Calculating the disease burden likely to be imposed by infection is hindered because the base upon which estimates should be founded is unclear. Serological studies using antibodies to Helicobacter as markers of infection, whether present or past, show clearly that in Western populations there is an increase with age in the proportions in whom raised titres can be detected, so that at age 60 half or more of all adults may have serological evidence of infection. Examination of age-stratified data indicates that positive serology is detectable in childhood in many individuals but that there is a cumulative increase with age. By contrast, in poorer and tropical populations seropositivity tends to be more frequent, and to be detectable at younger ages, the impression therefore being that infection is more easily acquired in infancy, in conditions where hygiene is poor; an analogy with infective diarrhoea is obvious (Klein et al. 1991; Mendall et al. 1992).
Analysis is complicated in at least two important ways. Firstly, since the frequency of ulcer is much lower than the frequency of infection, some explanation is needed of why most people who are infected do not develop ulcers. Secondly, comparison of results of serological studies conducted at different times indicates that the chances of becoming infected may have changed greatly, with a reduction in frequency in Western populations.
If such a reduction were generally occurring, and if there were a reasonably constant increment of risk of peptic ulceration imposed by infection, then one would predict that the peptic ulcer population would be tending to become progressively older as a cohort of the general population with a high rate of infection. There is evidence that this is what is happening. Thus in the United Kingdom the proportion of older individuals presenting with ulcer complications has risen (Walt et al. 1986).
The trend is clearly detectable in women; the difference from men is unexplained. It could perhaps be accounted for by a reduced rate in men as they smoke less with advancing age, but it does not seem to be accounted for by differential rates of non-steroidal anti-inflammatory drug use. Evidence that new infection, or reinfection, is uncommon in older people comes from two main sources. Firstly prolonged remission can be induced by antimicrobial therapy in ulcer patients and that remission is accompanied by evidence that active Helicobacter infection remains undetectable. Secondly, repeated serological examination at prolonged intervals has shown that few new infections were detectable serologically. Thus, in Busselton, Western Australia, comparison of serological evidence of infection in blood samples taken in 1968 and in 1990 showed that only 7 per cent of those aged 20 to 44 years in 1968 became positive in 1990 (Cullen et al. 1993).
If these findings are correct then they have important implications. Firstly, attempts to eradicate infection in those with established ulcer disease are clearly justified by the likely prolonged remission. Whether treating individuals who have evidence of infection but do not have peptic ulcer disease is worthwhile is harder to judge. In those who have abdominal symptoms but no detectable ulcer the results of antimicrobial therapy have been mixed, and often disappointing (Moayyedi et al. 2000; Laine et al. 2001). Thus in the United States no relationship could be found between the presence of infection, the activity of gastritis, and the occurrence of specific symptoms (Schubert et al. 1992).
A further factor to consider is whether eradication of infection, by reducing the risk of antral atrophic gastritis, would reduce the frequency of gastric cancer for which both atrophic gastritis and Helicobactor pylori infection appear to be risk factors, probably interdependently. Such treatment as a public health measure would imply a massive effort if up to half the adult population needed treatment, as seems likely. Furthermore, there are suggestions that eradication of Helicobactor pylori infection can enhance gastric acid secretion in patients with corpus gastritis, and so could make oesophageal reflux disease more likely. This, taken with evidence that reflux disease is an important predisposing factor to oesophageal adenocarcinoma (Lagergren et al. 1999), which is also increasing in frequency in Western countries (Pera et al. 1993; Devesa et al. 1998), makes the value of eradication in those without ulcer even more speculative.
Primary prevention of infection clearly must also be considered. Available evidence indicates that organisms can occasionally be found in dental plaque, but it is more likely that faecal–oral transmission takes place (Thomas et al. 1992). Risks of infection have been shown to be raised where sanitation has been poor during infancy (Mendall et al. 1992). The need for primary prevention measures may be less obvious in Western countries where the risk of infection appears to be falling rapidly (Parsonnet et al. 1992) than in those where clean water supplies and good sanitation may be lacking (Klein et al. 1991). Finally, consideration has to be given to intrinsic differences between varieties of Helicobacter (Covacci et al. 1993; Atherton et al. 1997).
Non-steroidal anti-inflammatory drugs
There is now ample evidence that takers of non steroidal anti-inflammatory drugs (NSAIDs) are at increased risk of peptic ulcer, hospitalization for the disease in general, or admission with ulcer complications. The development of ulcers during NSAID use has been clearly shown during the course of studies in which the value of drug prophylaxis against NSAID-associated ulcer has been examined. In addition, cohort and retrospective case–control studies have shown risks of varying size associated with NSAID treatment. Table 4 sets out odds ratios associated with the use of different agents. The size of the apparent risk varies, sometimes markedly, from one to another, reflecting a varying mix of dose differences from one place to another, and of differing design features as well as simple chance effects as confidence intervals round point estimates are often wide. However, the consistency of data, associated with experimental verification and dose–response effects, indicate compellingly that the treatment is directly responsible.

Table 4 Risks of gastrointestinal hospital admission of ulcer complications associated with non-aspirin non-steroidal drug use

The public health implications are complex. NSAIDs are used as analgesic and anti-inflammatory agents and in cardiovascular prophylaxis. Walt et al. calculated that about one-third of all peptic ulcer bleeding in the United Kingdom could be attributed to aspirin or non-aspirin NSAID use. Of this third about two-thirds were attributable to non-aspirin NSAID use, and a third to aspirin. NSAIDs are commonly used in managing patients with joint disease, apart from anti-ulcer-drug prophylaxis. Prevention depends on substituting safer selective selective NSAIDs (Langman et al. 1999). The disease burden imposed by NSAID treatment in these patients is nevertheless substantial. However, most treatment is given to patients with muscular and articular symptoms which may be ill-defined or associated with disease requiring simple analgesia, notably osteoarthritis. In such patients there is a strong case for using paracetamol, which is not ulcerogenic, or low doses of the least toxic NSAID, notably ibuprofen. Given current prescribing patterns for individual NSAIDs, such a policy should at least halve the number of ulcer complications attributed to NSAID use.
The position with aspirin cardioprophylaxis is more complex; most evidence showing its value relates to doses of 300 mg or more daily. However, there is clear evidence of substantially raised risks of upper gastrointestinal bleeding. The size of risk may be more than offset by the gain in terms of vascular disease prevented but it does emphasize the need to ensure that aspirin dosage is no greater than it need be. In this context risk appears to persist even when doses of 75 mg daily are in use (Weil et al. 1995).
Although it is plausible that Helicobacter infection might increases risks of ulcer associated with NSAID intake, evidence conflicts (Chan et al. 1997; Cullen et al. 1997).
Once Helicobacter infection and NSAID use have been taken into account, there remains major difficulty in deciding why ulceration develops. Thus, if over half of all those aged 60 and over in the general population have become infected by Helicobacter, and at least 30 per cent are takers of non-steroidal anti-inflammatory drugs or aspirin, then other reasons must explain why the prevalence of ulcer is of the order of 10 per cent. It is therefore reasonable to suggest that to develop an ulcer more than one factor must be operating, and that there are factors other than Helicobacter or drug use which are important.
Social factors
In the past, gastric ulcer was more common in the poor than in the rich, while the reverse was true of duodenal ulcer. However, the social class patterns for gastric and duodenal ulcer are now much the same. Table 5 compares mortality patterns according to social class in the United Kingdom. In general terms, ulcer death seems to be about twice as common in those of low social class (i.e. social classes IV and V).

Table 5 Ulcer mortality (SMR) in men aged 20–64 according to social class in UK

The reasons for these patterns are not easily explained. Ulcer might be expected to cause death in those who are disadvantaged socially and hence less likely to obtain high-quality medical care. However, 50 years ago higher mortality rates for duodenal ulcer obtained in those of social class I than in those of low social class. A common trend of morbidity and mortality patterns suggests that they are describing real differences.
Examination of occupational data does not suggest any link with liability to ulcer. Smokers are in general more prone to peptic ulceration than non-smokers, and the social class distribution of ulcer might simply reflect heavier smoking in those of lower social class. However, when social class was judged by level of educational attainment, peptic ulceration was found to be more common in smokers than non-smokers at each educational level, and Massachusetts physicians who smoked were more liable to peptic ulceration than those who did not smoke (Monson and MacMahon 1969).
Despite many attempts to relate the development, complications, or exacerbations of ulcer to stress, no coherent body of evidence has emerged to support the concept. Polednak (1974) noted that certain somatic symptoms, notably palpitation and sleeplessness, were associated with later development of ulcer, but these were two amongst many, and Piper et al. (1978, 1981a), concluded that no association with stressful life events could be detected for either gastric or duodenal ulcer.
There is clear evidence that smokers are more likely to die from peptic ulceration than non-smokers, that ulcer is likely to be found more often in smokers than non-smokers, and that duodenal ulcers are less likely to heal during histamine H2 antagonist treatment if patients are smokers (Hammond and Horn 1958; Doll and Hill 1964; Friedman et al. 1974; Kratochvil et al. 1982). Examination of the smoking habits of Massachusetts physicians has suggested that those smokers who had an ulcer were likely to have smoked more heavily, and from an earlier age, than those who did not have ulcers (Monson 1970). The association of peptic ulcer with smoking is independent of other associations with consumption of coffee and cola-type soft drinks (Paffenbarger et al. 1974a).
By simple comparison of time trends in smoking habits and ulcer mortality in the United States, Kurata et al. (1986) estimated that most of the changes observed in mortality could be attributed to changed smoking patterns. This evidence has largely ignored the influence of infection by H. pylori. However, it is plausible that smoking is an important factor in converting the infected and non-ulcerated to the infected and ulcerated.
The role of nutritional factors is discussed more fully elsewhere. Despite the likelihood that dietary factors are of significance in affecting liability to upper gastrointestinal disease, there is little supportive evidence. Those who drink coffee and cola-type soft drinks may be relatively unprotected (Paffenbarger et al. 1974b). The buffering capacities of foods vary; proteins are better buffers but are more likely to stimulate gastric secretion, whilst carbohydrate foods are indifferent buffers but promote less acid output. No useful date are available to show whether specific dietary items or major changes in protein, carbohydrate, or fat intake materially alter liability to peptic ulceration. An increase in dietary fibre intake has been claimed to protect against the development of ulcer. This is based upon comparative analyses of ulcer frequency in places where dietary fibre intake is thought to differ markedly. A controlled clinical study suggested that relapse of duodenal ulcer is less likely to occur in those taking dietary fibre supplements (Rydning et al. 1982). Even if true, it is unclear whether it is the quantity or a specific quality of dietary fibre that is important.
A further possibility is that high polyunsaturated fat intake can, by modulating prostaglandin synthesis, reduce liability to peptic ulceration. Convincing evidence in favour of this suggestion is lacking.
Prospects for prevention
It already seems likely that reductions in the frequency of infection with H. pylori, with reductions in smoking, are having major influences on the frequency of gastric and duodenal ulceration. In addition, understanding that NSAID and aspirin actions on cyclo-oxygenase (COX) can be separated into two types, with receptor differentiation, one (COX-1) being constitutive and associated with mucosal protection and antiplatelet effects, and the other (COX-2) being inducible and associated with anti-inflammatory properties (Holtzman et al. 1992), has led to the design of novel NSAIDs devoid of COX-1 activity. These do not appear to induce peptic ulcer or its complications (Langman et al. 1999; Simon et al. 1999), and, if they can be targeted to susceptible patients, should substantially reduce the burden of non-Helicobacter-associated ulcer.
Inflammatory bowel disease
Materially the same problems exist for the examination of changes in the frequency of inflammatory bowel disease as for peptic ulceration. In addition, there are no absolute criteria by which Crohn’s disease and ulcerative colitis can be separated with reliability. In the majority of cases the transmural inflammatory changes with the formation of giant cells and lymphoid aggregates, and the patchy distribution with, frequently, small intestine involvement of Crohn’s disease can be distinguished from the uniform distal mucosal inflammation, particularly involving the rectum, of ulcerative colitis.
Methods of assessment
Mortality rates
Measurement of death rates concentrates upon severe and complicated disease, with postoperative deaths and deaths in the elderly forming particularly important groups. Despite difficulties of interpretation it is worth noting that, in the United Kingdom, death rates and hospital discharge rates for Crohn’s disease have shown a similar upward trend (Miller et al. 1974; Langman 1979).
Hospital admission rates and other indices
It may no longer be true, given the range of medical treatments now available, that all patients with Crohn’s disease are ultimately admitted to hospital. However, it is clear that only a minority of colitics are admitted because most have limited bowel disease. In consequence we are inevitably dependent in epidemiological studies upon data which are collected comprehensively from admission statistics, clinic attendances, and pathological records.
When all data sets are considered it is clear that Crohn’s disease incidence rates rose markedly in the period between the 1960s and 1980s. This does not seem to represent diagnostic transfer, since ulcerative colitis incidence rates seem to have been stable. Crohn’s disease incidence rates themselves may also have now stabilized (Logan 1998), and Table 6 shows representative recent datasets for both diseases.

Table 6 Incidence of inflammatory bowel disease (data for selected studies) (rates per 100 000 per year)

Factors affecting disease patterns
We are largely ignorant of the causes of chronic inflammatory bowel disease. Both Crohn’s disease and ulcerative colitis seem to be associated with Western cultural patterns, but it is difficult to be sure in countries where infective dysenteries are endemic that non-specific disease is not being misclassified as infective. The difficulty is compounded by the tendency for many tropical countries to have less sophisticated medical services than European and North American areas.
Within Western areas there appears to be a tendency, at least in some, for there to be pronounced urban-to-rural gradients for Crohn’s disease, but probably not for colitis. However, there is no clear evidence of social class or occupational differences in disease frequency. Variations in sex ratio are of doubtful significance since data do not generally take into account smoking habits, which have been shown clearly to affect disease occurrence.
There is no conclusive evidence implicating dietary factors in the causation of ulcerative colitis or Crohn’s disease. However, indications have appeared repeatedly that dietary sucrose, as judged by retrospective enquiries, was consistently greater in Crohn’s disease patients than in matched controls (Martini and Brandes 1976; Thornton et al. 1979).
It seems likely, but is not certain, that dietary intakes antedate disease origin. However, a mechanism of action has not been identified, and modulation of dietary sucrose intake does not seem to influence disease behaviour.
Previous suggestions that milk intake could adversely affect disease behaviour were probably confounded by lack of understanding that those with intestinal hypolactasia would be prone to diarrhoea when consuming milk.
Other factors
There are unconfirmed suggestions that dietary sulphate could cause bowel damage—by conversion to sulphide (Gibson et al. 1991; Roediger et al. 1997). Butyrate enemata also seem to be beneficial in treating colitis but whether lack of short-chain fatty acids, a prime mucosal metabolic fuel in the bowel, precipitates disease is unclear.
In 1982, Rhodes and his colleagues noticed that patients with ulcerative colitis were mainly non-smokers (Harries et al. 1982). Subsequent studies have confirmed and extended these observations. Although an increased proportion of disease occurring in non- or ex-smokers could reflect stopping smoking on medical advice after disease onset, that explanation seems untenable.
Ulcerative colitis has been clearly and repeatedly shown to occur not only in non-smokers but in ex-smokers, with disease onset coming after the cessation of smoking (Somerville et al. 1984). Plausibility is increased by evidence that the more patients smoke the greater the protection (Tobin et al. 1987) and that nicotine may be an effective treatment.
In contrast, in Crohn’s disease there is an increased proportion of smokers, and again retrospective enquiry suggests that the habit is established before disease onset (Table 7). The plausibility of a harmful effect of smoking is increased by evidence that progression of Crohn’s disease is less obvious in those who do not smoke and in those persuaded to stop smoking (Cottone et al. 1994; Cosnes et al. 2001).

Table 7 Smoking and inflammatory bowel disease: relative risk for current smokers compared with never smokers at disease diagnosis or onset (data for selected studies)

Perhaps the simplest explanation of this data is that smoking in some way directs the site of action of aetiological factors, or a factor, in inflammatory bowel disease in genetically susceptible individuals.
Oral contraceptives
A relationship, albeit modest, has been suggested between oral contraceptive use and the occurrence of Crohn’s disease (Vessey et al. 1986; Lesko et al. 1985; Logan and Kay 1989). The reasons are unclear, and the findings have been disputed (Lashner et al. 1989), although more recent evidence suggests that current smoking and oral contraceptive use are independent risk factors for relapse (Timmer et al. 1998). An association with smoking through a common relationship with a thrombotic tendency is possible. In this context it is noteworthy that patients with haemophilia and von Willebrand’s disease appear to be protected from both colitis and Crohn’s disease (Thompson et al. 1995).
Non-steroidal anti-inflammatory agents
Though once tested as possible treatments, there is clear evidence of these drugs causing intestinal ulceration, inflammation, perforation, and bleeding (Langman et al. 1985b; Bjarnason et al. 1993). Whether they cause exacerbation of chronic inflammatory bowel disease or precipitate it in the first place is uncertain.
Associated disease
A recent intriguing piece of evidence indicates the rarity with which colitic patients have had appendectomy performed (Rutgeerts et al. 1993). The association appears strong, but not as high as the initial odds ratio suggested (Duggan et al. 1998). The mechanism is unclear but the inverse association is only present when appendectomy is performed in childhood (Andersson et al. 2001).
Childhood factors and infection
Since both Crohn’s disease and ulcerative colitis have maximal incidences in early adult life, it is natural to examine the relevance of childhood factors. Whilst some searches have been almost uniformly negative (Gilat et al. 1987), others have recently suggested that Crohn’s disease, but not ulcerative colitis, is associated with poor domestic hygiene in childhood (Gent et al. 1994). The inevitable associated question would be whether the association reflects an infective cause. Searches for evidence of transmission of infection using mathematical techniques to look for space–time clusters have been uninformative (Miller et al. 1975, 1976). Suggestions that inflammatory bowel disease may be associated with autism as a consequence of measles, mumps, and rubella vaccination do not appear well-supported (Ter Meulen 1998; Committee on Safety of Medicines 1999).
Prospects for prevention
Given the lack of clear evidence about causal agents, advice on preventative methods is hard to construct. The single major exception is that advice to stop smoking should have valuable effects both in preventing Crohn’s disease occurrence and in ameliorating its course. Calculating the actual value is impeded because the risks associated with smoking seem to vary from place to place.
Gallstones (cholelithiasis)
Cholelithiasis is extremely common in many countries, with a prevalence in the elderly of developed countries approaching a third or more (Table 8). As with colonic diverticular disease, the relationship between the presence of gallstones and the development of symptoms is poorly understood. If cholecystectomy rates are any guide, then symptomatic gallstone disease has been increasing in many countries (McPherson et al. 1985). Even so, recent surveys confirm that the numbers with asymptomatic disease exceed those with symptomatic disease, often by a factor of two to one or more (GREPCO 1984; Jorgensen 1987; Attili et al. 1995; Everhart 1999).

Table 8 Prevalence (in per cent) of gallstone disease in large ultrasound surveys

Gallstones contain varying proportions of cholesterol and bile pigments and by tradition have been classified as being cholesterol, pigment, or mixed stones. In Western countries the majority are mixed and by weight consist of 75 per cent or more of cholesterol. Few epidemiological studies distinguish types and the evidence available indicates that any recent increases have been in cholesterol-rich stones. Formation of cholesterol-rich stones is believed to be a four-stage process involving an individual with the genetic propensity or metabolic abnormality that encourages the production of bile saturated with cholesterol, the initiation of stone formation, and stone growth.
Method of assessment
Until the 1980s autopsy surveys were the main source of data on gallstone prevalence. Besides the obvious criticism that autopsied deaths may be unrepresentative, such surveys also depend on consistent examination and recording of the state of the gall bladder (Jorgensen et al. 1994). Abdominal ultrasound is now the method of choice for identifying gallstones and in the last 20 years there have been several large population surveys performed using ultrasound particularly in Europe (Table 8). In a few areas incidence has also been assessed by repeating the ultrasound survey some years after the first one. In one of these studies 5 per cent of those with gallstones identified in the first survey did not have them 5 years later, suggesting that asymptomatic passage or spontaneous dissolution may not be so uncommon (Jensen and Jorgensen 1991).
Hospital admission and operation rates
Measurement of disease using cholecystectomy rates seems logical, but the complex relationship between the presence of gallstones, development of symptoms, and eventual cholecystectomy creates problems. In approximately a third of cases, symptoms continue after cholecystectomy, indicating that gallstones were not the cause. Hospital admission data are further complicated by the fact that several admissions may occur before cholecystectomy is performed, which was normal practice until recently. Furthermore, the introduction of laparoscopic cholecystectomy has been followed by a rapid rise in cholecystectomy rates in several countries (Steiner et al. 1994; Lam et al. 1996). Nevertheless, McPherson et al. (1985) found a strong correlation between cholecystectomy rates and prevalence of cholelithiasis within Britain. In contrast, international comparisons of operation rates show a poor correlation with prevalence data.
Mortality rates
Death from cholelithiasis is now uncommon. In Dundee more patients died from postoperative complications of cholecystectomy than from the complications of untreated gallstones (Bateson and Bouchier 1975). Changes in mortality rates reflect a combination of improvements in medical care, declining case fatality, and the increasing frequency of cholecystectomy in the elderly, and therefore are of little value in assessing disease frequency.
Disease patterns
The autopsy data suggested that there were great variations in gallstone prevalence, being high in Western countries, such as the United Kingdom, United States, and Scandinavia, and low in many African and Third World nations. This pattern has only partly been confirmed by ultrasound surveys (Table 8). While prevalence has been highest in elderly women in Western countries, prevalence has been almost as great in Kashmiri women. Figures for other developing countries are sparse, but in Uganda, Ghana, and Thailand autopsy series indicate that fewer than 5 per cent of elderly women have gallstones (Royal College of Physicians 1980). Gallstone disease was also a rare cause for hospital admission in Africa, India, and Arabia (Burkitt and Tunstall 1975), but cholecystectomy is now one of the most common operations performed in Saudi Arabia (Tamimi et al. 1990).
In the United Kingdom, a twofold variation (9 to 21 per cent) in standardized prevalence was found in a study of nine towns selected to represent various socio-economic conditions (Barker et al. 1979). Measures of affluence within high prevalence areas appeared not to be major determinants of gallstone prevalence.
Secular trends
What little data that exists on time trends suggests that the prevalence in many countries has shown considerable fluctuation over the past century (Bateson and Bouchier 1975; Telium 1990). In areas where gallstones are now common it is generally believed that prevalence has been increasing since 1900, with a more rapid increase since 1945, doubling in a decade in some areas (Holland and Heaton 1972; Fowkes 1980). This may partly reflect more aggressive investigation and surgery.
Age, sex, and race
As shown in Table 8, gallstones are more common in women; the difference is greatest in younger age groups, often being twice that of men. The almost linear increase in prevalence with age has led to suggestions that incidence is largely independent of age, at least between ages 30 and 70 (Opit and Greenhill 1974). Some ultrasound surveys support this hypothesis but one study found that the incidence rate in 30-year-olds was less than half the rate in those over 40 years old (Jensen and Jorgensen 1991). In most countries symptomatic gallstone disease is rare before age 20 years.
Gallstone prevalence is strongly influenced by racial factors, presumably reflecting genetic predisposition. This is well-illustrated by the findings of the large United States NHANES survey in which black men had a significantly lower prevalence of gallbladder disease (5.3 per cent) than white men (8.6 per cent) and Hispanic men (8.9 per cent), whereas in women Hispanics had the highest prevalence (27 per cent), compared with whites (17 per cent) and blacks (14 per cent) (Everhart et al. 1999).
Factors influencing disease frequency
Obesity has long been recognized as a major risk factor for gallstone formation. Obesity is associated with excessive hepatic secretion of cholesterol, which results in obese individuals excreting bile that is more supersaturated with cholesterol than the non-obese. This abnormality reverses with weight reduction. Ultrasound surveys show obesity to be a more important risk factor for women than men (Jorgensen 1989; Everhart et al. 1999). In Copenhagen, for example, the prevalence of gallstone disease was four times greater in the heaviest women ({gt} 30 kg/m2) compared with the lightest ({lt} 20 kg/m2), while in men there was only a 1.6-fold difference. Similarly a sixfold increase in risk of developing symptomatic gallstones was found in the United States Nurses Health study when comparing women weighing greater than 32 kg/m2 with those less than 20 kg/m2 (Maclure et al. 1989). Paradoxically, persons losing weight also have an increased risk of developing gallstones and this is not entirely explained by the confounding effect of obesity. Ultrasound studies of people losing weight rapidly with very low-calorie diets have demonstrated the development of gallstones in up to 12 per cent within 4 months (Everhart 1993).
Although nutritional factors are obviously important in gallstone formation, it has been difficult to establish whether specific nutrients are involved. Total dietary fat and dietary cholesterol are not strongly associated but lack of dietary fibre has been implicated. In one survey of vegetarians, gallstone prevalence was half that of a matched control group of meat eaters (Pixley et al. 1985). Experimental studies have suggested that fibre can reduce the cholesterol saturation of supersaturated bile. Several studies have suggested that an excessive dietary energy intake occurs in patients with gallstones compared with controls (Sarles et al. 1970, 1978; Maclure et al. 1989). Others have disputed this finding, but this may reflect the frailties of dietary measurement.
A moderate alcohol intake has been associated with a modest reduction in frequency of clinical gallbladder disease (Friedman et al. 1966; Scragg et al. 1984; Maclure et al. 1989) and with fewer gallstones in population surveys (Jorgensen 1989; Everhart et al. 1999).
Although gallstone sufferers have been traditionally characterized as being ‘fertile’ as well as being fair, fat, and female, it is only recently that a clear effect of increasing parity has been disentangled from the confounding effects of obesity and age. During pregnancy bile becomes more saturated with cholesterol and the gallbladder tends to increase in volume and contract less.
Although some studies have shown an increasing risk of symptomatic gallbladder disease with parity (Friedman et al. 1966; Layde et al. 1982; Royal College of General Practitioners 1982; Scragg et al. 1984b), no significant relationship was found in the largest cohort (Maclure et al. 1989). However, ultrasound surveys have shown that after adjusting for the effects of age, obesity, and contraceptive use, a single pregnancy is associated with about a 25 per cent increase in overall prevalence of asymptomatic and symptomatic disease (GREPCO 1984; Jorgensen 1988).
Oestrogen therapy
Exogenous oestrogens increase the cholesterol saturation of bile, and oestrogens given as either oral contraceptives, or postmenopausal replacement, or when administered to men have been found to increase cholesterol saturation of bile (Bennion and Grundy 1978). Ultrasound surveys have found overall prevalence of gallbladder disease to be 36 per cent higher in ever compared with never users of oral contraceptives and cohort studies of oral contraceptive users have also found small (15 to 30 per cent) increases in incidence of symptomatic gallbladder disease (Thijs and Knipschild 1993; Murray et al. 1994). It seems likely that hormone replacement therapy will also increase gallstone formation but at present the size of this effect is unclear.
It might be expected that physical activity would be associated with a reduced risk of gallstones because of its relationship to weight. Recent data from the United States have confirmed that physical activity in both men and women is associated with a reduction in symptomatic gallstones independently of obesity and of weight loss. Compared with the least active quintile the most active had a cholecystectomy rate about a third lower (Leitzmann et al. 1998, 1999).
Most but not all studies have found smokers to have small increases in risk of less than twofold for gallbladder disease both asymptomatic and symptomatic. No mechanism is known and at present it is unclear whether this is a direct effect or a marker of some other risk factor (Logan and Skelly 2000).
Associated conditions
As with peptic ulcer, the increased surveillance of patients with gallstones probably accounts for early claims, now discounted, that gallstones are associated with disease such as peptic ulcer and hyperparathyroidism (Bennion and Grundy 1978). Clinical gallbladder disease is more common in diabetics (Maurer et al. 1990) and in population surveys diabetics have had a 50 to 100 per cent increase in prevalence of cholelithiasis (Jorgensen 1989; Everhart et al. 1999). In countries where gallstones are common, gallstones associated with cirrhosis of the liver, haemolytic disease, or resection of the ileum and biliary tract infection make only a small contribution to overall prevalence. In Japan biliary infection with roundworms appears to have been involved in pigment stone formation in the past, but this is no longer so.
Prospects for prevention
Cholesterol gallstone formation is strongly associated with Western lifestyles. Of the risk factors that can be modified, obesity is pre-eminent. Public health measures that reduce the prevalence of obesity might be expected to reduce gallstone formation. Although experimental studies suggest a beneficial effect from an increase in dietary fibre, at present there is insufficient evidence to support more specific dietary measures to reduce gallstone disease.
Appendicectomy for acute appendicitis continues to be one of the commonest emergency operations performed in the United Kingdom and most other Western countries. For example, recent United States operation rates indicate that the lifetime risk of having an appendicectomy for appendicitis is about 9 per cent for 5-year-old boys and 7 per cent for girls (Addiss et al. 1990). Provided that allowance is made for removal of normal appendices and for elective appendicectomy, disease frequency can be monitored using operation rates. Several studies have found that a diagnosis of acute appendicitis can be confirmed in about three-quarters of acute appendicectomies (Donnan and Lambert 1976; Barker and Liggins 1981), the proportion being higher in men than women, who more often have their appendix removed for functional or gynaecological symptoms.
Mortality rates are not useful for disease monitoring, since case fatality rates are low, currently 0.5 per cent in younger age groups. In people aged over 65, case fatality rates rise to over 5 per cent and in 1985 accounted for two-thirds of the 501 deaths from appendicitis in the United States. In England and Wales mortality rates reached a peak in the 1930s and since have fallen steadily, at a time when operation rates were continuing to increase (Donnan and Lambert 1976). In the United States age-adjusted mortality from appendicitis has declined 10-fold from over 2 to less than 0.2 per 100 000 between 1950 and 1985 (Mendeloff and Everhart 1994).
Disease patterns
A striking feature of the epidemiology of the disease has been the rapid changes in incidence evident in Europe and North America during the twentieth century. In England and Wales, mortality rates between 1901 and 1915 rose from 40 to 70 per million at a time when better treatment was reducing case fatality rates, suggesting a severalfold rise in incidence.
The pattern of disease in countries where appendicitis is common is fairly uniform, the peak incidence being between the ages of 5 and 25 years, with incidence being slightly higher in men than women (Table 9). In Western countries appendicitis was initially more common in the more affluent socio-economic groups, but in the United Kingdom this gradient, as measured by mortality and operation rates, had disappeared by 1960 (Donnan and Lambert 1976; Barker and Liggins 1981).

Table 9 Hospital discharge rate per 100 000 for appendicitis (ICD-9, 540–543), England, 1985

In contrast, Burkitt (1971) has drawn attention to the rarity of acute appendicitis among natives of rural Africa and in other rural areas of the Third World. A survey of 17 000 16- to 18-year-olds in South Africa found the prevalence of appendicectomy to be 0.6 per cent in rural African, 0.7 per cent in urban African, 2.9 per cent in Indian, and 10.5 per cent in white populations (Walker et al. 1982). In countries such as Nigeria and Kenya, where the disease is now occurring in the native population, it is first being reported in the more affluent natives of urban areas (Burkitt 1971).
In many Western countries including the United Kingdom hospital discharges for appendicitis have been declining steadily since the 1950s (Barker 1985); in England discharge rates have almost halved since 1968 (Fig. 2). In the United States hospitalization data show appendicitis incidence to have fallen from 1 per 400 a year to 1 per 1000 over the past 30 years (Mendeloff and Everhart 1994a). Some of the early decline was the result of diagnostic transfer from appendicitis to abdominal pain, but there is no evidence that diagnostic transfer is continuing. By contrast, in Hong Kong discharge rates for appendicitis have doubled in 20 years (Donnan 1986).

Fig. 2 Hospital discharge rate per 100 000 for appendicitis (ICD-8 and ICD-9 540–543), England and Wales, 1967–1985. (Data from OPCS 1985.)

Aetiological factors
Appendicitis is thought to result from obstruction of the appendix lumen with subsequent bacterial invasion of the ischaemic tissue. Obstruction is found in about one-third of all cases and may be due to faecoliths, which are possibly formed more readily on a low-fibre diet or to swelling of the lymphatic tissue surrounding the neck of the appendix in response to infection (Heaton 1987).
The low dietary fibre hypothesis remained largely untested until recently. Several case–control studies have now examined diet and although some have found slightly lower dietary fibre intakes in appendicitis cases compared with controls (Arnbjornsson 1983; Brender et al. 1985), others have found no significant differences (Nelson et al. 1984; Donnan 1986; Nelson et al. 1986). Furthermore, appendicitis rates in different parts of England and Wales have shown little correlation with cereal or total dietary fibre consumption but have been positively correlated with potato consumption and negatively with green vegetables and tomato consumption (Barker et al. 1986).
Barker (1985) has suggested that dietary changes correlate poorly with the time-trends for appendicitis in the United Kingdom and that the trends fit better with an infectious aetiology. On the basis of United Kingdom data correlating appendicitis rates with the provision of household amenities such as fixed baths, inside toilets, and running hot water, he attributes the rise and subsequent decline in appendicitis to improvements in domestic hygiene. With better hygiene early in life exposure to enteric infection is reduced, which in turn alters the response to later infection. This triggers appendicitis either by causing a greater lymphatic response around the appendix or by appendiceal mucosal damage allowing bacterial invasion. Further improvement in domestic hygiene results in even less enteric infection in childhood and adolescence and appendicitis rates then decline (Heaton 1987; Barker and Morris 1988; Barker et al. 1988).
Prospects for prevention
Although these rapid changes in appendicitis incidence underline the importance of environmental factors in its aetiology, primary prevention is clearly not possible on the basis of current knowledge. Indeed, if Barker’s sanitation hypothesis holds true epidemics of appendicitis may be anticipated as a temporary consequence of improvements in domestic hygiene in some developing countries. While dietary factors are also likely to be involved no such factors have been consistently demonstrated.
Diverticular disease of the colon
In Western countries colonic diverticulosis is an increasingly common condition. In the United States almost 1 per cent of the population reported having diverticular disease in the National Health Interview Survey (Mendeloff and Everhart 1994b). Possibly only a tenth of cases will develop significant symptoms and for the remainder the condition is little more than an anatomical curiosity. It is not possible to predict if and when diverticulosis will produce symptoms.
Most estimates of prevalence have been obtained from either autopsy data or routine barium enema examinations. Both are usually performed on people with varying degrees of ill-health and are likely to overestimate prevalence (Mendeloff 1986). Accurate autopsy examination also requires cleaning and fixing of the colon before examination of the luminal surface with a hand lens. A radiographic technique suitable for population surveys, using oral barium and a single film at 48 h, has been described, but as yet has been used in only a few studies (Manousos et al. 1967, 1985; Foster et al. 1978).
Mortality data for diverticular disease are of limited value in assessing disease frequency since fatality rates are low, and changes in mortality are likely to reflect improvements in medical care and differences in diagnostic coding practices. Nevertheless, about 3000 deaths a year are attributed to diverticular disease in the United States, five times as many as for inflammatory bowel disease (Mendeloff and Everhart 1994b). Hospital discharge data are similarly difficult to interpret, particularly as the indications for admission to hospital and colectomy are ill-defined. In Scotland, depending on age, only 4 to 14 per cent of cases discharged from hospital with diverticular disease, even from surgical units, have had surgery performed (Chalmers et al. 1983).
Disease patterns
Like appendicitis, the striking feature of the epidemiology of diverticular disease is the great increase in prevalence that has taken place in Western countries in the twentieth century and the continuing low prevalence in less-developed countries. Between 1923 and 1963, the mortality from diverticular disease in England and Wales rose from 2 to 25 per million with a plateau during and immediately after the Second World War and food rationing (Painter and Burkitt 1971).
In countries where diverticular disease is common, the prevalence is similar in men and women, being rare before the age of 30 years and rising sharply at 60 years and over (Table 10). Although no relationship between presence of diverticular disease and socio-economic status has been established in the United States, self-reported disease is greater in those with higher incomes and twice as commonly reported by women despite hospitalization rates for United States women being only 20 per cent higher than men.

Table 10 Diverticular disease: prevalence by age in Oxford, England

Although many of the estimates are anecdotal, diverticular disease is almost non-existent in rural Africans (Painter and Burkitt 1971). In these areas diverticulosis has been detected in less than one in 1000 autopsies or barium enemas. In urban-living Africans in South Africa and Kenya, diverticular disease is now being reported with increased frequency (Calder 1979; Segal and Walker 1982). Prevalence also appears low in the Oriental populations such as in Japan and Hong Kong, and when present tends to involve the right colon (Coode et al. 1985; Chia et al. 1991; Nakada et al. 1995).
Factors influencing disease frequency
To account for the geographical distribution, attention has focused mainly on dietary factors. Painter and Burkitt (1971) were the first to suggest that diverticular disease is the result of a fibre-deficient diet. In support of the importance of a low-fibre diet, a radiographic survey of English vegetarians with a high dietary fibre intake (41.5 g/day) found that the prevalence of asymptomatic diverticular disease was only 12 per cent, compared with 33 per cent in a matched control population (mean dietary fibre intake 21.4 g/day) (Gear et al. 1979). In case–control studies patients with symptomatic diverticular disease have been found to have lower dietary fibre intakes than controls and in the only prospective study United States male health professionals with a low fibre intake developed more symptomatic diverticular disease than those with a high intake (Brodribb and Humphreys 1976; Segal and Walker 1982; Manousos et al. 1985; Aldoori et al. 1994). Furthermore, in animal models rats have developed diverticulosis in inverse proportion to the amount of fibre in their feed (Fisher et al. 1985).
The role of other dietary components is not yet established; meat-eating has been implicated in diverticular disease in two studies (Manousos et al. 1985; Aldoori et al. 1994). Ingestion of non-steroidal anti-inflammatory drugs has also been implicated in the development of complications such as diverticulitis and diverticular bleeding in several British studies (Langman et al. 1985; Campbell and Steele 1991; Aldoori et al. 1998). In the United States health professionals study, physical activity, especially vigorous activity, appeared to protect from symptomatic disease (Aldoori et al. 1995). Other factors such as obesity, arteriosclerosis, and weakness of the colonic muscle, once considered responsible, are now regarded as secondary associations.
Prospects for intervention
The evidence for a central role for dietary fibre is stronger than for any other disease where it has been implicated. At least one randomized controlled trial has shown benefit in symptomatic diverticular disease (Brodribb and Humphreys 1976). On these grounds recommendations to increase dietary fibre intake can be supported.
Coeliac disease
Compared with other gastrointestinal diseases the aetiology of coeliac disease is well understood, as is indicated by its alternative name of gluten-sensitive enteropathy. Dietary gluten, the protein fraction of wheat, rye, and barley, is essential to the development of the disease. Oats also used to be proscribed but have now been shown not to be harmful. With avoidance of dietary gluten the characteristic intestinal villous atrophy recovers, often completely, and the disease remits. Gluten intolerance as demonstrated by the development of villous atrophy is believed to be lifelong, and in most coeliacs reintroduction of dietary gluten leads to recurrence of villous atrophy within 6 months.
A genetic predisposition to coeliac disease is well established. Family studies have consistently found that about 10 per cent of first-degree relatives have villous atrophy, which is often asymptomatic (Logan 1992). Strong associations with several genes within the HLA complex have been shown. Nevertheless, the difference in concordance rates between monozygotic twins and HLA identical siblings (70 per cent versus 30 per cent) implicates non-HLA genes as well.
Besides the appropriate genetic predisposition and dietary gluten, the importance of other unidentified environmental factors is underlined by the discordance of some monozygotic twin pairs (Polanco et al. 1981), and by the variable delay in recurrence of villous atrophy after gluten re-exposure (Bardella et al. 2000).
Methods of assessment
Mortality data are of no value in assessing disease frequency since the mortality rate of coeliacs is no more than twice that of the general population and death is rarely certified as being due to coeliac disease itself (Logan et al. 1989). Autopsy surveys are also not possible because of the effects of postmortem autolysis on the small intestinal mucosa. Investigation and diagnosis do not normally require hospital admission. Thus figures for incidence and prevalence were, until recently, derived from hospital-based case series and the steadily increasing prevalence reported for adults was generally assumed to reflect increased investigation rather than increased incidence (Logan 1992). Serological screening for coeliac disease is now possible using tests for the presence of antibodies to either gliadin, a cereal protein fraction, or to endomysium, a tissue antigen. Serological screening is now widely used in clinical practice and increasingly in population surveys (Catassi et al. 1994; Johnston et al. 1997; Meloni et al. 1999).
Patterns of disease
Depending on the measure used, the occurrence of clinical coeliac disease diagnosed in childhood shows a 10-fold variation across Europe (Fig. 3). Particularly striking is the recent doubling in incidence of childhood coeliac disease in Sweden where 3.5 of every 1000 births were being diagnosed as coeliac (Cavell et al. 1992). In contrast over a similar period the incidence in Denmark was constant at around 0.09 per 1,000 births (Weile and Krasilnikoff 1993). How much of this difference can be explained by differing proportions of subclinical undiagnosed disease is not known.

Fig. 3 Coeliac disease in children in Europe: cumulative incidence by age 15 (per 1000 live births). (Data from Greco et al. 1992.)

The increasing use of serological screening has shown that the prevalence of asymptomatic or subclinical (minor symptoms only) coeliac disease, first recognized in family studies, considerably exceeds previous estimates. In recent population surveys in children the prevalence has been as high as 1 per cent in Sardinia and 0.5 per cent in The Netherlands (Csizmadia et al. 1999; Meloni et al. 1999). In adults prevalence has been of a similar order at 0.8 per cent in Northern Ireland and 0.5 per cent in northern Sweden (Johnston et al. 1997; Ivarsson et al. 1999). What is clear is that the prevalence of undiagnosed coeliac disease in these areas is much greater than that already diagnosed, giving credence to the concept of a coeliac disease ‘iceberg’. What is not known is the extent to which the hazards traditionally associated with untreated coeliac disease (osteoporosis and malignancy) apply to subclinical disease.
There are few figures on the occurrence of coeliac disease outside Europe. In North America the disease has been thought to be less common but preliminary serological surveys indicate that the prevalence as assessed by serology may be similar to Europe (Not et al. 1998).
Factors influencing disease frequency
In children the development of coeliac disease has been related to infant feeding practices. Case–control data implicate both bottle feeding and early introduction of dietary gluten (Greco et al. 1988). In the United Kingdom in the mid-1970s there was a sudden halving in incidence in children associated with a marked increase in mothers’ breast feeding and delaying the introduction of dietary gluten (Littlewood et al. 1980; Dossetor et al. 1981; Logan 1992). The recent increase in Sweden, where there is already a high prevalence of breast feeding, occurred when there was an increase in the amount of gluten introduced at weaning, suggesting that relatively small changes in infant feeding either prevent disease developing or delay its development for many years (Juto et al. 1994; Ivarsson et al. 2000).
If breast feeding and the amount and timing of gluten introduction are critical then measures to encourage the right pattern of infant feeding should reduce occurrence in childhood. Whether this will prevent the disease or merely postpone the onset of overt disease into adulthood remains to be determined.
The frequency of pancreatic disease is difficult to determine. Acute illness associated with grossly raised serum amylase levels is readily detected, but lesser varieties of acute disease may have no clear markers of occurrence, and the diagnosis of chronic disease in the absence of gross pancreatic calcification is dependent upon the use of sophisticated examination procedures of doubtful sensitivity and specificity. The few sets of data available give incidence rates for acute pancreatitis of between 5 and 25 per 100 000 per year with case-fatality rates of between 10 and 20 per cent. Thus, in the United States about 2700 deaths were certified annually to pancreatitis, mainly acute disease (Go and Everhart 1994). Data from non-European countries show that chronic disease may be common in tropical areas, such as South India, where gross calcific pancreatitis is regularly seen and is believed to be secondary to malnutrition.
Factors influencing disease frequency
Alcohol intake
In those areas where alcohol consumption is heavy, pancreatitis and liver cirrhosis tend to be common. But when clinical series from different areas are compared, the frequency with which disease is attributed to alcohol intake can vary from less than 5 per cent of acute disease in the United Kingdom to almost 90 per cent of chronic disease in France (Sarles et al. 1965). Chronic pancreatitis has also been associated with smoking independently of alcohol intake in several studies (Bourliere et al. 1991).
Gallstone migration along the common bile duct is an important cause of acute pancreatitis, accounting for one-third to one-half of all cases in Westernized countries depending on the prevalence of gallstones and alcohol abuse (Corfield et al. 1985). Biliary sludge or microlithiasis may also account for as many as one-half of otherwise idiopathic acute disease (Lee et al. 1992). There is no evidence that there are any special peculiarities about gallstone disease associated with pancreatitis.
No specific nutritional factors have been associated with liability to pancreatitis. Claims in Western countries of an association with protein or fat intake are inconsistent with the suggestion from tropical areas that malnutrition is of significance.
Drug-induced disease
Acute pancreatitis has been said to occur following treatment with, inter alia, corticosteroid drugs, analgesic agents, oral contraceptives, and diuretics. Almost all these claims rest upon anecdotal evidence, and taken overall it seems unlikely that drug-induced disease accounts for more than a very small proportion of the total. A single case–control study supports an association with diuretic use (Bourke et al. 1978).
Other factors
If there are other important environmental factors, and the general lack of aetiological information suggests that there must be, then they are unknown. It has been persistently suggested that viral infections can cause pancreatitis. Mumps virus could account for a very small proportion of cases of idiopathic non-gallstone non-alcoholic pancreatitis, but little is known about other possible causes.
Liver disease
Throughout the world morbidity and mortality from liver disease varies considerably. Comparative morbidity data are not available but in countries where liver disease is common, mortality can account for up to 5 per cent of all deaths, whereas in the United Kingdom, mortality from liver disease is less than 0.5 per cent of all deaths. Nevertheless, even in the United Kingdom liver disease ranks second only to peptic ulcer as a cause of digestive disease mortality, other than digestive cancers. In contrast in the United States around 1.5 per cent of all deaths are certified to chronic liver disease (32 000 in 1985) (Everhart 1994).
In the main, mortality results from infective liver disease including the acute viral hepatitides, hepatocellular cancer, and cirrhosis of the liver. In African countries such as Mozambique and Uganda, and also in developed countries like Greece and Japan, mortality is predominantly the result of hepatitis B virus infection and hepatocellular cancer. In Western countries hepatitis C infection has emerged as a major public health problem with 1.8 per cent of the United States population showing evidence of infection (Alter et al. 1999). Globally, prevalence of chronic infection is estimated to average 3 per cent and to account for 70 per cent of chronic hepatitis, 40 per cent of cirrhosis, and 60 per cent of hepatocellular cancer in Western Europe and North America (Proceedings of the European Association for the Study of the Liver 1999).
Of the other acute liver diseases, Reye’s syndrome, first described in 1963, was estimated to account for almost a thousand cases per year in the United States in the 1970s (Sullivan-Bolyai and Corey 1981) (Fig. 4). The condition, which had a 50 per cent fatality rate, mainly affects children and produces an acute encephalopathy with cerebral oedema and liver failure due to hepatic steatosis. Case–control studies showing a strong association with aspirin ingestion led authorities in many countries to advise caution in the use of aspirin in children with viral illnesses. The incidence has subsequently shown a dramatic decline in both the United States and United Kingdom (Belay et al. 1999).

Fig. 4 Number of cases of Reye’s syndrome reported in the United States.

Cirrhosis of the liver
The term ‘cirrhosis’ indicates an endstage histopathological state resulting from a wide range of causes. Death usually results from liver failure, often precipitated by gastrointestinal bleeding, or from development of an hepatocellular carcinoma once liver failure is established. Fewer than 15 per cent of cirrhotics survive 5 years, making mortality rates a reasonable measure of disease frequency. Nevertheless, liver failure is not an inevitable consequence of cirrhosis, as demonstrated in a British survey in which 11 per cent of cases were an incidental finding at autopsy (Saunders et al. 1981).
In Western countries alcohol abuse is the most common cause of cirrhosis and mortality from liver disease tends to correlate with levels of alcohol consumption (Fig. 5). The proportion of cirrhosis that is due to alcohol varies with the level of alcohol consumption in the population but even in populations where consumption has been low around a half of cases are alcohol-related.

Fig. 5 Comparison of alcohol consumption and deaths from cirrhosis of the liver in 12 countries. (Data from Popham et al. 1975.)

In areas such as China, South East Asia and sub-Saharan Africa where hepatitis B infection is hyperendemic, hepatitis B infection is the major cause of cirrhosis. In these countries most adults have evidence of past infection and chronic carrier rates exceed 10 per cent (compared with {lt} 0.2 per cent in blood donors in the United States and United Kingdom). Estimates as to what proportion of chronic carriers have a chronic hepatitis vary from 10 to 30 per cent, with higher rates for early ages of infection. Cirrhosis tends to develop only in those with more severe chronic infection (chronic active hepatitis) and in this group annual incidence has been found to be between 1.5 and 2.5 per cent. It has been estimated that worldwide there are around 300 million hepatitis B carriers (Dusheiko and Hoofnagle 1991). In comparison there are estimated to be around 150 million chronic hepatitis C carriers, of whom it is thought about 20 per cent will eventually develop cirrhosis 10 to 20 years after becoming infected (European Association for the Study of the Liver 1999).
Of the other causes of cirrhosis, a substantial number (half of non-alcoholic cirrhosis) are labelled cryptogenic and assumed to be the result of an autoimmune process such as chronic active hepatitis and primary biliary cirrhosis. Little is known of the aetiology of these autoimmune chronic liver diseases, which accounted for 70 per cent of non-alcoholic cirrhosis in Birmingham, United Kingdom (Saunders et al. 1981). Primary biliary cirrhosis is often asymptomatic but can be detected by testing for antimitochondrial antibodies and raised alkaline phosphatase levels in blood. In Europe, the prevalence of clinically diagnosed disease is 7 to 24 per 100 000, with incidence of around a tenth of prevalence (Lofgren et al. 1985; Metcalf et al. 1997).
Prospects for prevention
Despite the limited data available it is abundantly clear that alcohol abuse accounts for the majority of chronic liver disease in countries where drinking alcohol is popular. Measures that either reduce overall alcohol consumption or selectively reduce extreme consumption will reduce the incidence of chronic liver disease.
Infection due to hepatitis B is a prime contributor to morbidity from cirrhosis and to mortality from hepatocellular cancer. The free availability of genetically engineered vaccines against the virus could make a crucial contribution to reducing the impacts of disease. The incidence of hepatitis C infection is already declining as transmission by blood products has virtually ceased. Increased awareness of the risks from intravenous drug use and needle exchange programs would reduce incidence further.
Prospects for the control of gastrointestinal disease in the new millennium
The rapid rise and fall in frequency of various digestive diseases worldwide is ample evidence of the importance of lifestyle and environmental factors in their causation. This suggests that the prevention of much gastrointestinal disease is a realistic objective.
For acute infective diseases which in many countries account for much of the burden of gastrointestinal disease, the prime requirement for control is the reliable provision of safe drinking water. In many developed countries modern methods of food production and preparation coupled with public complacency has lead to a resurgence of foodborne infections, emphasizing the need for continued surveillance and adequate legislation.
For chronic gastrointestinal disease the greatest rewards are likely to come from the control and prevention of H. pylori and hepatitis B and C virus infections. Recognition of the vital role of H. pylori in peptic ulceration and probably also in gastric carcinogenesis has thrown up the real possibility of producing substantial reductions in incidence of both conditions by the prevention of infection. However, public health action awaits a better understanding of the epidemiology of the infection, in particular the circumstances of its transmission. In contrast, the ability to prevent hepatitis B infection by screening and vaccination already exists and further public health action awaits the development of suitable strategies and resources.
Prospects for control of those diseases in which infection is not thought to be involved are less promising, mainly because understanding of their causation is rudimentary. Measures that reduce overall alcohol consumption will reduce the levels of liver and pancreatic disease, but reductions in tobacco consumption are likely to have little effect as smoking is only weakly related to most gastrointestinal diseases (the outstanding exception being oesophageal cancer). Measures that promote diets high in cereal or vegetable fibre and low in animal fat are likely to have a favourable impact on colonic disease as well as on vascular disease. On the debit side the rise in levels of inflammatory bowel disease may be a consequence of the lower rates of enteric infection that ‘hygienic’ Western lifestyles promote.
Chapter References
Addiss, D.G., Shaffer, N., Fowler, B.S., and Tauxe, R.V. (1990). The epidemiology of appendicitis and appendectomy in the United States. American Journal of Epidemiology, 132, 910–25.
Aldoori, W.H., Giovannucci, E.L., Rimm, E.B., Wing, A.L., Trichopoulos, D.V., and Willett, W.C. (1994). A prospective study of diet and the risk of symptomatic diverticular disease in men. American Journal of Clinical Nutrition, 60, 757–64.
Aldoori, W.H., Giovannucci, E.L., Rimm, E.B., et al. (1995). Prospective study of physical activity and the risk of symptomatic diverticular disease in men. Gut, 36, 276.
Aldoori, W.H., Giovannucci, E.L., Rimm, E.B., Wing, AL., and Willett, W.C. (1998). Use of acetaminophen and nonsteroidal anti-inflammatory drugs: a prospective study and the risk of symptomatic diverticular disease in men. Archives of Family Medicine, 7, 255–60.
Alter, M.J., Kruszon-Moran, D., Nainan, O.V., et al. (1999). The prevalence of hepatitis C virus infection in the United States, 1988 through 1994. New England Journal of Medicine, 341, 556–62.
Andersson, R.E., Olaison, G., Tysk, C., and Ekbom, A. (2001). Appendectomy and protection against ulcerative colitis. New England Journal of Medicine, 344, 808–14.
Arnbjornsson, E. (1983). Acute appendicitis and dietary fibre. Archives of Surgery, 118, 868.
Atherton, J.C., Peck, Jr, R., Tham, K.T., Cover, T.L., and Blaser, M.J. (1997). Clinical and pathological importance of heterogeneity in vac A, the vacuolating cytotoxin gene of Helicobacter pylori. Gastroenterology, 112, 92–9.
Attili, A.F., Carulli, N., Roda, E., et al. (1995). Epidemiology of gallstone disease in Italy: prevalence data of the Multicenter Italian Study on Cholelithiasis (MICOL). American Journal of Epidemiology, 141, 158–65.
Baird-Parker, A.C. (1990). Foodborne salmonellosis. Lancet, 336, 1231–5.
Bardella, M.T., Fredella, C., Prampolini, Marino, R., and Conte, O. (2000). Gluten sensitivity in monozygous twins: a long-term follow-up study. American Journal of Gastroenterology, 95, 1503–5.
Barker, D.J.P. (1985). Acute appendicitis and dietary fibre: an alternative hypothesis. British Medical Journal, 290, 1125.
Barker, D.J.P. and Liggins, A. (1981). Acute appendicitis in nine British towns. British Medical Journal, 283, 1083.
Barker, D.J.P. and Morris, J. (1988). Acute appendicitis, bathrooms and diet in Britain and Ireland. British Medical Journal, 296, 953.
Barker, D.J.P., Gardner, M.J., Power, C., and Hutt, M.S.R. (1979). Prevalence of gallstones at necropsy in British towns. British Medical Journal, 11, 1389.
Barker, D.J.P., Morris, J., and Nelson, M. (1986). Vegetable consumption and acute appendicitis in 59 areas in England and Wales. British Medical Journal, 292, 927.
Barker, D.J.P., Osmond, C., Golding, J., and Wadsworth, M.E.J. (1988). Acute appendicitis and bathrooms in three samples of British children. British Medical Journal, 296, 956.
Bateson, M.C. and Bouchier, I.A.D. (1975). Prevalence of gallstones in Dundee: a necropsy study. British Medical Journal, iv, 425.
Belay, E.D., Bresee, J.S., Holman, R.C., Khan, A.S., Shahriari, A., and Schonberger, L.B. (1999). Reye’s syndrome in the United States from 1981 through 1997. New England Journal of Medicine, 340, 1377–82.
Bennion, L.J. and Grundy, S.M. (1978). Risk factors for the development of cholelithiasis in man. New England Journal of Medicine, 299, 1221.
Bjarnason, I., Hayllar, J., MacPherson, A.J., and Russell, A.S. (1993). Side-effects of non steroidal drugs on the small and large intestine in humans. Gastroenterology, 104, 1832.
Bourke, J.B., Mead, G.M., McIllmurray, M.B., and Langman, M.J.S. (1978). Drug-associated primary acute pancreatitis. Lancet, i, 706–8.
Bourliere, M., Barthet, M., Berthezene, P., Durbec, J.P., and Sarles, H. (1991). Is tobacco a risk factor for chronic pancreatitis and alcoholic cirrhosis? Gut, 32, 1392–5.
Brender, J.D., Weiss, N.S., Koepsell, T.D., and Marcuse, E.K. (1985). Fiber intake and childhood appendicitis. American Journal of Public Health, 75, 399.
Brodribb, A.J.M. and Humphreys, D.M. (1976). Diverticular disease: three studies. British Medical Journal, i, 424.
Burkitt, D.P. (1971). The aetiology of appendicitis. British Journal of Surgery, 58, 695.
Burkitt, D.P. and Tunstall, M. (1975). Gallstones: geographical and chronological features. Journal of Tropical Medicine and Hygiene, 78, 140.
Calder, J.F. (1979). Diverticular disease of the colon in Africans. British Medical Journal, i, 1465.
Calkins, B., Lilienfeld, A., Mendeloff, A., Garland, C.F., Monk, M., and Garland, F. (1984). Smoking factors in ulcerative colitis and Crohn’s disease in Baltimore. American Journal of Epidemiology, 120, 498.
Campbell, K. and Steele, R.J.C. (1991). Non-steroidal anti-inflammatory drugs and complicated diverticular disease: a case–control study. British Journal of Surgery, 78, 190–1.
Catassi, C., Rätsch, I.-M., Fabiani, E., et al. (1994). Coeliac disease in the year 2000: exploring the iceberg. Lancet, 343, 200.
Cavell, B., Stenhammar, L., Ascher, H., Danielsson, L., Dannaeus, A., Lindberg, T., and Lindquist, B. (1992). Increasing incidence of childhood coeliac disease in Sweden. Results of a national study. Acta Paediatrica, 81, 589–92.
Chalmers, K., Wilson, J.M.G., Smith, A.N., and Eastwood, M.A. (1983). Diverticular disease of the colon in Scottish hospitals over a decade. Health Bulletin, 41, 32–41.
Chan, F.K.L., Sung, J.J.Y., Chung S.C.S., et al. (1997). Randomised controlled trial of eradication of Helicobacter pylori before non-steroidal anti-inflammatory drug therapy to prevent peptic ulcers. Lancet, 350, 975–9.
Chia, J.G., Chiantana, C.W., Ngoi, S.S., Goh, P.M., and Ong, C.L. (1991). Trends of diverticular disease of the large bowel in a newly developed country. Diseases of the Colon and Rectum, 34, 498.
Committee on Safety of Medicines (1999). The safety of MMR vaccine. Current problems in pharmacovigilance, 25, 9–10.
Coode, P.E., Chan, K.W., and Chan, Y.T. (1985). Polyps and diverticula of the large intestine: a necropsy survey in Hong Kong. Gut, 26, 1045–8.
Cook, G.C. (1985). Tuberculosis—certainly not a disease of the past! Quarterly Journal of Medicine, 56, 519–21.
Corfield, A.P., Cooper, M.J., and Williamson, R.C.N. (1985). Acute pancreatitis: a lethal disease of increasing incidence. Gut, 26, 724–9.
Cosnes, J., Beaugerie, L., Carbonnel, F., and Gendre, J.P. (2001). Smoking cessation and the course of Crohn’s disease—an intervention study. Gastroenterology, 120, 1093–9.
Cottone, M., Rosselli, M., Orlando, A., et al. (1994). Smoking habits and recurrence in Crohn’s disease. Gastroenterology, 106, 643–8.
Covacci, A., Censini, S., Bugnoli, M., et al. (1993). Molecular characterization of the 120 kDa immunodominant antigen of Helicobacter pylori associated with cytotoxicity and duodenal ulcer. Proceedings of the National Academy of Sciences of the United States of America, 90, 5791–5.
Csizmadia, C.G.D.S., Mearin, M.L., von Blomberg, B.M.E., Brand, R., and Verloove-Vanhorick, S.P. (1999). An iceberg of childhood coeliac disease in the Netherlands. Lancet, 353, 813–14.
Cukor, G. and Blacklow, N.R. (1984). Human viral gastroenteritis. Microbiology Reviews, 48, 157–79.
Cullen, D.J.E., Collins, B.J., Christiansen, K.J., Epis, J., Warren, J.R., Surveyor, I., and Cullen, K.J. (1993). When is Helicobacter pylori infection acquired? Gut, 34, 1681.
Cullen, D.J.E., Hawkey, G.M., Greenwood, D.C., Humphreys, H., Shepherd, V., Logan, R.F.A., and Hawkey, C.J. (1997). Peptic ulcer bleeding in the elderly: relative roles of Helicobacter pylori and non-steroidal anti-inflammatory drugs. Gut, 41, 459–62.
De Pancorbo, C.M., Carballo, F., Horcajo, P., et al. (1997). Prevalence and associated factors for gallstone disease: results of a population survey in Spain. Journal of Clinical Epidemiology, 50, 1347–55.
Devesa, S.S., Blot, W.J., and Fraumeni, J.F. (1998). Changing patterns in the incidence of esophageal and gastric carcinoma in the United States. Cancer, 83, 2049–53.
Doll, R. and Hill, A.B. (1964). Mortality in relation to smoking: 10 years observation of British doctors. British Medical Journal, i, 1399.
Donnan, S.P.B. (1986). Appendicitis in Hong Kong in the aetiology of acute appendicitis, pp. 16–19. Scientific Report 7, Southampton MRC Environmental Epidemiology.
Donnan, S.P.B. and Lambert, P.M. (1976). Appendicitis: incidence and mortality. Population Trends, 5, 26.
Dossetor, J.F.B., Gibson, A.A.M., and McNeish, A.S. (1981). Childhood coeliac disease is disappearing. Lancet, i, 322–3.
Duggan, A.E., Usmani, I., Neal, K.R., and Logan, R.F.A. (1998). Appendicectomy, childhood hygiene, Helicobacter pylori status, and risk of inflammatory bowel disease: a case control study. Gut, 43, 494–8.
Dusheiko, G. and Hoofnagle, J.H. (1991). Viral hepatitis. In Oxford textbook of clinical hepatology (ed. N. MacIntyre, J.P. Benhamou, J. Bircher, M. Rizzetto, and J. Rodes), Vol.1, pp. 571–92. Oxford University Press.
El Serag, H. and Sonnenberg, A. (1998). Opposing time trends of peptic ulcer and reflux disease. Gut, 43, 327–33.
Everhart, J.E. (1993). Contributions of obesity and weight loss to gallstone disease. Annals of Internal Medicine, 119, 1029–35.
Everhart, J.E. (1994). Overview. In Digestive diseases in the United States: epidemiology and impact (ed. J.E. Everhart), pp. 1–53. NIH Publication 94–1447, US Government Printing Office, Washington, DC.
Everhart, J.E., Khare, M., Hill, M., and Maurer, K.R. (1999). Prevalence and ethnic differences in gallbladder disease in the United States. Gastroenterology, 117, 632–9.
Farthing, M.J.G. (1993). Pathophysiology of infective diarrhoea. European Journal of Gastroenterology and Hepatology, 5, 796–809.
Farthing, M.J.G. (1994). Oral rehydration therapy. Pharmacology and Therapeutics, 64, 477–92.
Farthing, M.J.G., Du Pont, H.L., Guandalini, S., Keusch, G.T., and Steffen, R. (1992). Treatment and prevention of travellers’ diarrhoea. Gastroenterology International, 5, 162–75.
Farthing, M.J.G., Katelaris, P.H., Dias, J., Munzer, D., and Popovic, O. (1993). Bacterial and parasitic intestinal infections in Europe. Gastroenterology International, 6, 149–66.
Fisher, N., Berry, C.S., Fearn, T., Gregory, J.A., and Hardy, J. (1985). Cereal dietary fibre consumption and diverticular disease: a life-span study in rats. American Journal of Clinical Nutrition, 42, 788.
Foster, K.J., Holdstock, G., Whorwell, P.J., Guyer, P., and Wright, R. (1978). Prevalence of diverticular disease of the colon in patients with ischaemic heart disease. Gut, 19, 1054.
Fowkes, F.G.R. (1980). Cholecystectomy and surgical resources in Scotland. Health Bulletin, 38, 126.
Franceschi, A., Panza, E., La Vecchia, S., Parazzini, F., Decarh, A., and Bianchi Porro, G. (1987). Non-specific inflammatory bowel disease and smoking. American Journal of Epidemiology, 125, 445.
Friedman, G.D., Kannel, W.B., and Dawber, T.R. (1966). The epidemiology of gallbladder disease: observations on the Framingham study. Journal of Chronic Diseases, 19, 273.
Friedman, G.D., Siegelaub, A.B., and Seltzer, C.C. (1974). Cigarettes, alcohol, coffee and peptic ulcer. New England Journal of Medicine, 290, 469.
Gear, J.S.S., Ware, A., Fursdon, P., Mann, J.I., Nolan, D.J., Brodribb, A.J.M., and Vessey, M.P. (1979). Symptomless diverticular disease and intake of dietary fibre. Lancet, i, 511.
Gent, A.E., Hellier, M.D., Grace, R.H., Swarbrick, E.T., and Coggon, D. (1994). Inflammatory bowel disease and domestic hygene in infancy. Lancet, 343, 766.
Gibson, G.R., Cummings, J.H., and Macfarlane G.T. (1991). Growth and activities of sulphate-reducing bacteria in gut contents of healthy subjects and patients with ulcerative colitis. FEMS Microbiol Ecol, 86, 103–12.
Gilat, T., Hacohen, D., Lilos, P., and Langman, M.J.S. (1987). Childhood factors in ulcerative colitus and Crohn’s disease: an international cooperative study. Scandinavian Journal of Gastroenterology, 22, 1009.
Glambek, I., Kvaale, G., Arnesjö, B., Søreide, O. (1987). Prevalence of gallstones in a Norwegian population. Scandinavian Journal of Gastroenterology, 22, 1089–94.
Go, V.L.W. and Everhart, J.E. (1994). Pancreatitis. In Digestive diseases in the United States: epidemiology and impact (ed. J.E. Everhart), pp. 691–712. NIH Publication 94–1447, US Government Printing Office, Washington, DC.
Gollop, J.H., Phillips, S.F., Melton III, L.J., and Zinsmeister, A.R. (1988). Epidemiological aspects of Crohn’s disease; a population-based study in Olmsted County, Minnesota 1943–82. Gut, 29, 49.
Gower-Rousseau, C., Salomez, J.L., Dupas, J.L., et al. (1994). Incidence of inflammatory bowel disease in Northern France (1988–1990). Gut, 35, 1433.
Greco, L., Auricchio, S., Mayer, M., and Grimaldi, M. (1988). Case control study on nutritional risk factors in celiac disease. Journal of Paediatric Gastroenterology and Nutrition, 7, 395–9.
Greco, L., Maki, M., Di Donato, F., and Visakorpi, J.K. (1992). Epidemiology of coeliac disease in Europe and the Mediterranean Area. In Common food intolerances. Vol. 1, Epidemiology of coeliac disease (ed. S. Auricchio and J.K. Visakorpi), pp. 25–44. Karger, Basel.
GREPCO (1984). Prevalence of gallstone disease in an Italian adult female population. American Journal of Epidemiology, 119, 796–805.
GREPCO (1988). The epidemiology of gallstone disease in Rome, Italy. Part I. Prevalence data in men. Hepatology, 8, 904–6.
Hammond, E.C. and Horn, D. (1958). Smoking and death rates-report on forty-four months of follow-up of 187,783 men. I. Journal of the American Medical Association, 166, 1294.
Harries, A.D., Baird, A., and Rhodes, J. (1982). Non-smoking: a feature of ulcerative colitis. British Medical Journal, 284, 706.
Heaton, K.W. (1987). Aetiology of acute appendicitis. British Medical Journal, 294, 1632–3.
Heaton, K.W., Braddon, F.E.M., Mountford, R.A., Hughes, A.O., and Emmett, P.M. (1991). Symptomatic and silent gallstones in the community. Gut, 32, 316–20.
Henry, D., Dobson, A., and Turner, C. (1993). Variability in the risk of major gastrointestinal complications from non-aspirin non-steroidal anti-inflammatory drugs. Gastroenterology, 105, 1978.
Holland, C. and Heaton, K.W. (1972). Increasing frequency of gallbladder operations in the Bristol clinical area. British Medical Journal, iii, 627.
Holtzman, M.J., Turk, J., and Shornick, L.P. (1992). Identification of a pharmacologically distinct prostaglandin H synthase in cultured epithelial cells. Journal of Biological Chemistry, 267, 21438–45.
Ivarsson, A., Persson, L.A., Juto, P., Peltonen, M., Suhr, O., and Hernell, O. (1999). High prevalence of undiagnosed coeliac disease in adults: a Swedish population-based study. Journal of Internal Medicine, 245, 63–8.
Ivarsson, A., Persson, LA., Nyström, L., Ascher, H., Cavell, B., Danielsson, L., et al. (2000). Depidemic of coeliac disease in Swedish children. Acta Paediatrica, 89, 165–71.
Jensen K.H. and Jorgensen, T. (1991). Incidence of gallstones in a Danish population. Gastroenterology, 100, 790–4.
Johnston, S.D., Watson, R.G.P., McMillan, S.A., Sloan, J., and Love, A.H.G. (1997). Prevalence of coeliac disease in Northern Ireland. Lancet, 350, 1370.
Jorgensen, T. (1987). Prevalence of gallstones in a Danish population. American Journal of Epidemiology, 126, 912.
Jorgensen, T. (1988). Gallstones in a Danish population: fertility period, pregnancies and exogenous female sex hormones. Gut, 29, 433–9.
Jorgensen, T. (1989). Gallstones in a Danish population: relation to weight, physical activity, smoking, coffee consumption, and diabetes mellitus. Gut, 30, 528–34.
Jorgensen, T., Rossen, K., and Thorvaldsen, P. (1994). Are autopsy studies reliable in assessing gallstone prevalence in the community? International Journal of Epidemiology, 23, 566–9.
Juto, P., Meeuwisse, G., and Mincheva-Nilsson, L. (1994). Why has coeliac disease increased in Swedish children? Lancet, 343, 1372.
Katelaris, P.H., Salam, I., and Farthing M.J.G. (1995). Lactobacilli to prevent traveller’s diarrhea. New England Journal of Medicine, 333, 1360–1.
Khuroo, M.S., Mahajan, R., Zargar, S.A., Javid, G., and Sapru, S. (1989). Prevalence of biliary tract disease in India: a sonographic study in adult population in Kashmir. Gut, 30, 201–5.
Klein, P.D., Graham, D.Y., and Gaillour, A. (1991). Watersource as a risk factor for Helicobacter pylori infection in Peruvian Children. Lancet, 337, 1503.
Kratochvil, P., Gugler, R., and Rohner, H.G. (1982). Effect of smoking on duodenal ulcer healing with cimetidine and oxmetidine. Gut, 23, 866.
Kurata, J.H., Elashoff, J.D., Nogawa, A.N., and Haile, D.M. (1986). Sex and smoking differences in duodenal ulcer mortality. American Journal of Public Health, 76, 700.
Labenz, J. and Borsch, G. (1994). Evidence for the essential role of Helicobacter pylori in gastric ulcer disease. Gut, 35, 19.
Lagergren, J., Bergstrom, R., and Lindgren, A. (1999). Symptomatic gastroesophageal reflux as a risk factor for esophageal adenocarcinoma. New England Journal of Medicine, 340, 825–31.
Laine, L., Schoenfield, P., and Fennerty, B. (2001). Therapy for helicobacter pylori in patients with non-ulcer dyspepsia: a meta-analysis for randomised controlled trials. Annals of Internal Medicine, 134, 361–9.
Lam, C.-M., Murray, F.E., and Cuschieri, A. (1996). Increased cholecystectomy rate after the introduction of laparoscopic cholecystectomy in Scotland. Gut, 38, 282–4.
Langman, M.J.S. (1979). The epidemiology of chronic digestive disease. Edward Arnold, London.
Langman, M.J.S., McConnell, T.H., Spiegelhalter, D.J., and McConnell, R.B. (1985a). Changing patterns of coeliac disease frequency: an analysis of Coeliac Society membership records. Gut, 26, 175–8.
Langman, M.J.S., Morgan, L., and Worrall, A. (1985b). Use of anti-inflammatory drugs by patients admitted with small or large bowel perforations or haemorrhage. British Medical Journal, 290, 347–9.
Langman, M.J.S., Weil, J., Wainwright, P., et al. (1994). Risks of bleeding peptic ulcer associated with individual non steroidal anti-inflammatory drugs. Lancet, 343, 1075.
Langman, M.J., Jensen, D.M., Watson, D.J., et al. (1999). Adverse upper gastrointestinal effects of rofecoxib compared with NSAIDs. Journal of the American Medical Association, 282, 1929–33.
Laporte, J.R., Carne, X., Vidal, X., Morena, M., and Juan, J. (1991). Upper gastrointestinal bleeding in relation to previous use of analgesics and non-steroidal anti-inflammatory drugs. Lancet, 337, 85.
Lashner, B.A., Kane, S.V., and Hanauer, S.B. (1989). Lack of association between oral contraceptive use and Crohn’s disease: a community-based matched case–control study. Gastroenterology, 97, 1442.
Layde, P.M., Vessey, M.P., and Yeates, D. (1982). Risk factors for gall-bladder disease: a cohort study of young women attending family planning clinics. Journal of Epidemiology and Community Health, 36, 274.
Lee, S.P., Nicholls, J.F., and Park, H.Z. (1992). Biliary sludge as a cause of acute pancreatitis. New England Journal of Medicine, 326, 589–93.
Leitzmann, M.F., Giovannucci, E., Rimm, E.B., Stampfer, M.J., Spiegelman, D., Wing, A.L., and Willett, W.C. (1998). The relation of physical activity to risk for symptomatic gallstone disease in men. Annals of Internal Medicine, 128, 417–25.
Leitzmann, M.F., Rimm, E.B., Willett, W.C., et al. (1999). Recreational physical activity and the risk of cholecystectomy in women. New England Journal of Medicine, 341, 777–84.
Lesko, S.M.N., Kaufman, D.W., Rosenberg, L., Helmrich, S.P., Miller, D.R., Stolley, P.O., and Shapiro, S. (1985). Evidence for an increased risk of Crohn’s disease in oral contraceptive users. Gastroenterology, 89, 1046–9.
Levine, M.M. (1990). Modern vaccines: enteric infections. Lancet, 335, 958–61.
Lindberg, E., Tysk, C., Andersson, K., and Jarnerot, G. (1988). Smoking and inflammatory bowel disease: a case–control study. Gut 29, 352.
Littlewood, J.M., Crollick, A.J., and Richards, I.D.G. (1980). Childhood coeliac disease is disappearing. Lancet, ii, 1359.
Lofgren, J., Jarnerot, G., Danielsson, D., and Hemdal, I. (1985). Incidence and prevalence of primary biliary cirrhosis in a defined population in Sweden. Scandinavian Journal of Gastroenterology, 20, 647–50.
Logan, R.F.A. (1998). Inflammatory bowel disease incidence: up, down or unchanged? Gut, 42, 309–11.
Logan, R.F.A. (1992). The epidemiology of coeliac disease. In Coeliac disease (ed. M.N. Marsh), pp. 192–214. Blackwell Scientific, Oxford.
Logan, R.F.A. and Kay, C.R. (1989). Oral contraception, smoking and inflammatory bowel disease—findings in the Royal College of General Practitioners Oral Contraception Study. International Journal of Epidemiology, 18, 105.
Logan, R.F.A. and Skelly, M.M. (2000). Smoking and hepato-biliary disease. European Journal of Gastroenterology and Hepatology, 12, 863–7.
Logan, R.F.A., Edmond, M., Somerville, K.W., and Langman, M.J.S. (1984). Smoking and ulcerative colitis. British Medical Journal, 288, 751.
Logan, R.F.A., Rifkind, E.A., Turner, I.D., and Ferguson, A. (1989). Mortality in coeliac disease. Gastroenterology, 97, 265–71.
Maclure, K.M., Hayes, K.C., Colditz, G.A., Stampfer, M.J., Speizer, F.E., and Willett, W.C. (1989). Weight, diet, and the risk of symptomatic gallstones in middle-aged women. New England Journal of Medicine, 321, 563–9.
McPherson, K., Strong, P.M., Jones, L., and Britton, B.J. (1985). Do cholecystectomy rates correlate with geographic variations in the prevalence of gallstones? Journal of Epidemiology and Community Health, 39, 179.
Manousos, O.N., Truelove, S.C., and Lumsden, K. (1967). Transit times of food inpatients with diverticulosis or irritable colon syndrome and normal subjects. British Medical Journal, iii, 760.
Manousos, O.N., Day, N.E., Tzonou, A., et al. (1985). Diet and other factors in the aetiology of diverticulosis: an epidemiologic study in Greece. Gut, 26, 544.
Martini, G.A. and Brandes, J.W. (1976). Increased consumption of refined carbohydrates in patients with Crohn’s disease. Klinische Wochenschrift, 54, 367.
Maurer, K.R., Everhart, J.E., Knowler, W.C., Shawker, T.H., and Roth, H.P. (1990). Risk factors for gallstone disease in the Hispanic populations of the United States. American Journal of Epidemiology, 131, 836–44.
Mayberry, J.F. and Rhodes, J. (1986). The changing incidence of Crohn’s disease in Wales and the role of heredity in its aetiology. In The genetics and epidemiology of inflammatory bowel disease (ed. R.B. McConnell, P. Rozen, M. Langman, and T. Gilat), pp. 114–17. Karger, Basel.
Meloni, G., Dore, A., Fanciulli, G., Tanda, F., and Bottazzo, G.F. (1999). Subclinical coeliac disease in schoolchildren from northern Sardinia. Lancet, 353, 37.
Mendall, M.A., Goggin, P.M., and Molineaux, N. (1992). Childhood living conditions and Helicobacter seropositivity in adult life. Lancet, 339, 896.
Mendeloff, A.I. (1986). Thoughts on the epidemiology of diverticular disease. Clinics in Gastroenterology, 15, 855.
Mendeloff, A.I. and Everhart, J.E. (1994). Acute appendicitis. In Digestive diseases in the United States: epidemiology and impact (ed. J.E. Everhart), pp. 457–67. NIH Publication 94–1447, US Government Printing Office, Washington, DC.
Mendeloff, A.I. and Everhart, J.E. (1994). Diverticular disease of the colon. In Digestive diseases in the United States: epidemiology and impact (ed. J.E. Everhart), pp. 551–65. NIH Publication 94–1447, US Government Printing Office, Washington, DC.
Metcalf, J.V., Bhopal, R.S., Gray, J., Howel, D., and James, O.F.W. (1997). Incidence and prevalence of primary biliary cirrhosis in the City of Newcastle upon Tyne, England. International Journal of Epidemiology, 26, 830–6.
Miller, D.S., Keighley, A.C., and Langman, M.J.S. (1974). Changing patterns in epidemiology of Crohn’s disease. Lancet, ii, 691.
Miller, D.S., Keighley, A., Smith, P.G., Hughes, A.O., and Langman, M.J.S. (1975). Crohn’s disease in Nottingham: a search for time–space clustering. Gut, 16, 454.
Miller, D.S., Keighley, A., Smith, P.G., Hughes, A.O., and Langman, M.J.S. (1976). A case control method for seeking evidence of contagion in Crohn’s disease. Gastroenterology, 71, 385.
Moayyedi, P., Feltbower, R., Brown, J., et al. (2000). Effect of population screening and treatment for Helicobacter pylori on dyspepsia and quality of life: a randomised controlled trial. Lancet, 355, 1665–71.
Monson, R.R. (1970). Cigarette smoking and body form in peptic ulcer. Gastroenterology, 58, 337.
Monson, R.R. and MacMahon, B. (1969). Peptic ulcer in Massachusetts physicians. New England Journal of Medicine, 281, 11.
Morris, T. and Rhodes, J. (1984). Incidence of ulcerative colitis in the Cardiff region 1968–77. Gut, 25, 846.
Murray, F.E., Logan, R.F.A., Hannaford, P.C., and Kay, C.R. (1994). Cigarette smoking and parity as risk factors for the development of symptomatic gall bladder disease in women: results of the Royal College of General Practitioners’ oral contraception study. Gut, 35, 107–11.
Nakada, I., Ubukata, H., Goto, Y., et al. (1995). Diverticular disease of the colon at a regional general hospital in Japan. Diseases of the Colon and Rectum, 38, 755–9.
Neal, K.R., Brij, S.O., Slack, R.C.B., Hawkey, C.J., and Logan, R.F.A. (1994). Recent treatment with H2 antagonists and antibiotics and gastric surgery as risk factors for Salmonella infection. British Medical Journal, 308, 176.
Nelson, M., Barker, D.J.P., and Winter, P.D. (1984). Dietary fibre and acute appendicitis: a case–control study. Human Nutrition: Applied Nutrition, 38A, 126.
Nelson, M., Morris, J., Barker, D.J.P., and Simmonds, S. (1986). A case–control study of acute appendicitis and diet in children. Journal of Epidemiology and Community Health, 41, 316.
Nomura, A., Kashiwagi, S., Hayashi, J., et al. (1988). Prevalence of gallstone disease in a general population of Okinawa, Japan. American Journal of Epidemiology, 128, 598–605.
Nomura, A., Stemmermann, G.N., Chyou, P., Perez-Perez, G.I., and Blaser, M.J. (1994). Helicobacter pylori infection and the risk for duodenal and gastric ulceration. Annals of Internal Medicine, 120, 997.
Not, T., Horvath, K., Hill, I.D., Partanen, J., Hammed, A., Magazzu, G., and Fasano, A. (1998). Celiac disease risk in the USA: high prevalence of antiendomysium antibodies in healthy blood donors. Scandinavian Journal of Gastroenterology, 33, 494–8.
Oksanen, P.J., Salminen, S., Saxelin, M., et al. (1990). Prevention of travellers’ diarrhea by Lactobacillus. Annals of Medicine, 22, 53–6.
OPCS (Office of Population Censuses and Surveys) (1985). Hospital activity analysis. OPCS, London.
OPCS (Office of Population Censuses and Surveys) (1986). Decennial supplement occupational mortality 1979–80, 1982–3. OPCS, London.
OPCS (Office of Population Censuses and Surveys) (1993). Mortality statistics, cause 1991. Series DH2, No. 18. HMSO, London.
Opit, L.J. and Greenhill, S. (1974). Prevalence of gallstones in relation to differing treatment rates for biliary disease. British Journal of Preventive and Social Medicine, 28, 268.
Paffenbarger, R.S., Wing, AL., and Hyde, R.T. (1974a). Coffee, cigarettes and peptic ulcer. New England Journal of Medicine, 290, 1091.
Paffenbarger, R.S., Wing, A.L., and Hyde, R.T. (1974b). Chronic disease in former college students. XIII. Early precursors of peptic ulcer. American Journal of Epidemiology, 100, 307.
Painter, N.S. and Burkitt, D.P. (1971). Diverticular disease of the colon: a deficiency disease of Western civilisation. British Medical Journal, ii, 450.
Parsonnet, J., Blaser, M.J., Perez-Perez, G.I., Hargrett-bean, N., and Tauxe, R.V. (1992). Symptoms and risk factors of Helicobacter pylori infection in a cohort of epidemiologists. Gastroenterology, 102, 41–6.
Peltola, H., Sitonen, A., Kyronseppa, H., et al. (1991). Prevention of travellers’ diarrhoea by oral B-subunit/whole-cell cholera vaccine. Lancet, 338, 1285–9.
Pera, M., Cameron, A.J., Trastek, V.F., Carpenter, H.A., and Zinsmeister, A.R. (1993). Increasing incidence of adenocarcinoma of the esophagus and esophagogastric junction. Gastroenterology, 104, 510–13.
Piper, D.W., Greig, M., Shinners, J., Thomas, J., and Crawford, J. (1978). Chronic gastric ulcer and stress. Digestion, 18, 303.
Piper, D.W., McIntosh, J.H., Ariotti, D.E., Calgiuri, J.V., Brown, R.W., and Shy, C.M. (1981). Life events and chronic duodenal ulcer: a case–control study. Gut, 22, 1011.
Pixley, F. and Mann, J. (1988). Dietary factors in the aetiology of gallstones: a case control study. Gut, 29, 1511.
Pixley, F., Wilson, D., McPherson, K., and Mann, J. (1985). Effect of vegetarianism on development of gallstones in women. British Medical Journal, 291, 11.
Polednak, A.P. (1974). Some early characteristics of peptic ulcer descendants. Gastroenterology, 67, 1094.
Popham, R.E., Schmidt, W., and de Lint, L. (1975). The prevention of alcoholism: epidemiological studies of the effect of government control measures. British Journal of Addiction, 70, 125.
Porter, J.D., Ragazzoni, H.P., Buchanon, J.D., Waskin, H.A., Juranek, D.D., and Parkin, W.E. (1988). Giardia transmission in a swimming pool. American Journal of Public Health, 78, 659–62.
Primatesta, P., Goldacre, M.J., and Seagroatt, V. (1994). Changing patterns in the epidemiology and hospital care of peptic ulcer. International Journal of Epidemiology, 23, 1206.
European Association for the Study of the Liver (1999). Proceedings of the International Consensus Conference on Hepatitis C, Paris, France, 26–27 February 1999. Journal of Hepatology, 31 (Supplement 1), 1–268.
Registrar General (1971). Decennial supplement. Occupational mortality tables for 1959–63. HMSO, London.
Reid, J.A., Caul, E.O., White, D.G., and Palmer, S.R. (1988). Role of infected food handler in hotel outbreak of Norwalk-like viral gastroenteritis: implications for control. Lancet, ii, 321–3.
Roediger, W.E.W., Moore, J., and Babidge, W. (1997). Colonic sulfide in pathogenesis and treatment of ulcerative colitis. Digestive Diseases and Sciences, 42, 1571–9.
Royal College of General Practitioners Oral Contraceptive Study (1982). Oral contraceptives and gallbladder disease. Lancet, ii, 957.
Royal College of Physicians (1980). Medical aspects of dietary fibre. Pitman, London.
Rutgeerts, P., D’Haens, G., Heile, M., Geboesk, and Vantrappen, G. (1994). Appendectomy protects against ulcerative colitis. Gastroenterology, 106, 1251–3.
Rydning, A., Berstad, A., Aadland, E., and Odegaard, B. (1982). Prophylactic effect of dietary fibre in duodenal ulcer disease. Lancet, ii, 736.
Sarles, H., Sarles, J.C., Camatte, R., et al. (1965). Observations in 205 confirmed cases of acute pancreatitis, recurring pancreatitis and chronic pancreatitis. Gut, 6, 545.
Sarles, H., Hawton, J., Planche, N.E., Lafont, H., and Gerolami, A. (1970). Diet, cholesterol gallstones and composition of bile. American Journal of Digestive Diseases, 15, 251.
Sarles, H., Gerolami, A., and Bord, A. (1978). Diet and cholesterol gallstones. Digestion, 17, 128.
Saunders, J.B., Walters, J.R.F., Davies, P., and Paton, A. (1981). A 20-year prospective study of cirrhosis. British Medical Journal, 282, 263.
Savage, R.L., Moller, P.W., Ballantyne, C.L., and Wells, J.E. (1993). Variation in the risk of peptic ulcer complications with non-steroidal anti-inflammatory drug therapy. Arthritis and Rheumatism, 36, 84.
Schubert, T.T., Schubert, M.A., and Ma, C.K. (1992). Symptoms, gastritis and Helicobacter pylori in patients referred for endoscopy. Gastrointestinal Endoscopy, 38, 357.
Scragg, R.K.R., McMichael, A.J., and Baghurst, P.A. (1984a). Diet, alcohol, and relative weight in gallstone disease: a case–control study. British Medical Journal, 288, 1113.
Scragg, R.K.R., McMichael, A.J., and Seamark, R.F. (1984b). Oral contraceptives, pregnancy and endogenous oestrogens in gall-stone disease: a case–control study. British Medical Journal, 288, 1795.
Segal, I. and Walker, A.R.P. (1982). Diverticular disease in urban Africans in South Africa. Digestion, 24, 42.
Shivananda, S., Pena, A.S., Mayberry, J.F., Ruiten, E.J., and Hoedemaeter, P.J. (1987a). Epidemiology of proctocolitis in the region of Leiden, the Netherlands. Scandinavian Journal of Gastroenterology, 22, 993.
Shivananda, S., Pena, A.S., Mayberry, J.F., Ruiten, E.J., and Hoedemaeter, P.J. (1987b). Epidemiology of Crohn’s disease in region Leiden, the Netherlands. Gastroenterology, 93, 966.
Simon, L.S., Weaver, A.L., Graham, Dy., et al. (1999). Anti-inflammatory and upper gastrointestinal effects of celecoxib in rheumatoid arthritis: a randomised controlled trial. Journal of the American Medical Association, 282, 1921–8.
Somerville, K.W., Logan, R.F.A., Edmond, M., and Langman, M.J.S. (1984). Smoking and Crohn’s disease. British Medical Journal, 289, 954.
Sonnenberg, A. (1994). Peptic ulcer. In Digestive diseases in the United States: epidemiology and impact (ed. J.E. Everhart), pp. 357–408. NIH Publication 94–1447, US Government Printing Office, Washington, DC.
Sorvillo, F.J., Fujoka, K., Nahlen, B., Tormey, M.P., Kebabjian, R., and Mascola, L. (1992). Swimming-associated cryptosporidiosis. American Journal of Public Health, 82, 742–4.
Steiner, C., Bass, E.B., Talamini, M.A., Pitt, H.A., and Steinberg, E.P. (1994). Surgical rates and operative mortality for open and laparoscopic cholecystectomy in Maryland. New England Journal of Medicine, 330, 403–8.
Stonnington, C.M., Phillips, S.F., Melton, III, L.J., and Zinsmeister, A.R. (1987). Chronic ulcerative colitis: incidence and prevalence in a community. Gut, 28, 402.
Suharyono, S., Simanjuntak, C., Witham, N., et al. (1992). Safety and immunogenicity of single-dose live oral cholera vaccine CVD 103-HgR in 5–9 year old Indonesian children. Lancet, 340, 689–94.
Sullivan-Bolyai, J.Z. and Corey, L. (1981). Epidemiology of Reyes’ syndrome. Epidemiologic Review, 3, 1.
Tamimi, T.M., Wosornu, L., Al-Khozaim, A., and Abdul-Ghani, A. (1990). Increased cholecystectomy rates in Saudi Arabia. Lancet, 336, 1235–7.
Taylor, D.N. and Blaser, M.J. (1991). The epidemiology of Helicobacter pylori infection. Epidemiologic Reviews, 13, 42.
Telium, D. (1990). Prevalence of gallstones at autopsy at the Institutes of Forensic Medicine in Aarhus and Copenhagen, Denmark, in 1944–1985. Scandinavian Journal of Gastroenterology, 25, 401–4.
Telzak, E.E., Budnick, L.D., Greenberg, M.S.L., Blum, S., Shayegani, M., Benson, C.E., and Schultz, S. (1990). A nosocomial outbreak of Salmonella enteritidis infection due to the consumption of raw eggs. New England Journal of Medicine, 323, 394–7.
Ter Meulen, V. (1998). Measles virus and Crohn’s disease: view of a medical virologist. Gut, 43, 733–4.
Thijs, C. and Knipschild, P. (1993). Oral contraceptives and the risk of gallbladder disease: a meta-analysis. American Journal of Public Health, 83, 1113–20.
Thomas, J.G., Gibson, G.R., and Darboe, M.K. (1992). Isolation of Helicobacter pylori from human faeces. Lancet, 340, 1194.
Thompson, N.P., Wakefield, A.J., and Pounder, R.E. (1995). Inherited disorders of coagulation appear to protect against inflammatory bowel disease. Gastroenterology, 108, 1011–15.
Thornton, J.R., Emmett, P.M., and Heaton, K.W. (1979). Diet and Crohn’s disease: characteristics of the pre-illness diet. British Medical Journal, ii, 762.
Timmer, A., Sutherland, L.R., Martin, F., and the Canadian Mesalamine for the Remission of Crohn’s Disease Study Group (1998). Smoking, use of oral contraceptives, and medical induction of remission were risk factors for relapse in Crohn’s disease. Gastroenterology, 114, 1143–50.
Tobin, M.V., Logan, R.F.A., Langman, M.J.S., McConnell, U.R.B., and Gilmore, I.T. (1987). Cigarette smoking and inflammatory bowel disease. Gastroenterology, 93, 316.
Tsianos, E.V., Masalas, C.N., Merkouropoulos, M., Kalekos, G.N., and Logan, R.F.A. (1994). Incidence of inflammatory bowel disease in north west Greece: rarity of Crohn’s disease in an area where ulcerative colitis is common. Gut, 35, 369.
Vessey, M., Jewel, D., Smith, A., Yeates, D., and McPherson, K. (1986). Chronic inflammatory bowel disease, cigarette smoking and use of oral contraceptives: findings in a large cohort study of women of childbearing age. British Medical Journal, 292, 1101–3.
Walker, A. (1992). Swimming—the hazards of taking a dip. British Medical Journal, 304, 242–5.
Walker, A.R.P., Walker, B.F, Duvenhage, A., Jones, J., Neongwane, J., and Segal, I. (1982). Appendicectomy prevalences in South African adolescents. Digestion, 23, 274.
Walt, R., Katschinski, B., Logan, R., Ashley, J., and Langman, M.J.S. (1986). Rising frequency of ulcer perforation in elderly people in the United Kingdom. Lancet, i, 489.
Wheeler, J.G., Sethi, D., Cowden, J.M., et al. (1999). Study of infectious intestinal disease in England: rates in the community, presenting to general practice, and reported to national surveillance. British Medical Journal, 318, 1046–50.
Weile, B. and Krasilnikoff, P.A. (1993). Extremely low incidence rates of celiac disease in the Danish population of children. Journal of Clinical Epidemiology, 46, 661–4.
Weil, J., Colin Jones, D., Langman, M.J.S., et al. (1995). Prophylactic aspirin and risk of peptic ulcer bleeding. British Medical Journal, 310, 827–30.

3 comments on “9.12 Gastrointestinal disease: public health aspects

  1. Couldn’t have said it better myself.

  2. […] Non-Steroidal Anti-Inflammatory Drugs (NSAIDs)9.12 Gastrointestinal disease: public health aspects var _gaq = _gaq || []; _gaq.push(['_setAccount', 'UA-3856133-4']); _gaq.push(['_setDomainName', […]

  3. Very good post. Looking forth to the next one.

    Provided that you continue this quality. I’m positive you will have a lot of visitors in no time.
    All the best, and also looking for your up coming content.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: