Epidemiology of Helicobacter pylori Infection ... - Semantic Scholar

International Journal of Epidemiology © International Epidemiological Association 1997

Vol. 26, No. 4 Printed in Great Britain

Epidemiology of Helicobacter pylori Infection among 4742 Randomly Selected Subjects from Northern Ireland LIAM J MURRAY,* EVELYN E McCRUM,* ALUN E EVANS* AND KATHLEEN B BAMFORD** Murray L J (Department of Epidemiology and Public Health,The Queen’s University of Belfast, Mulhouse Building, Royal Group of Hospitals, Grosvenor Road, Belfast BT12 6BJ, Northern Ireland), McCrum E E, Evans A E and Bamford K B. Epidemiology of Helicobacter pylori infection among 4742 randomly selected subjects from Northern Ireland. International Journal of Epidemiology 1997; 26: 880–887. Background. Despite the widespread prevalence and serious clinical sequelae of infection with Helicobacter pylori, there have been few large population-based studies, using randomly selected subjects, examining the epidemiology of this infection. Aim. To examine the distribution and determinants of H. pylori infection in a developed country. Subjects and Setting. Overall 4742 subjects, aged 12–64, from Northern Ireland were randomly selected. Methods. Helicobacter pylori specific IgG antibodies were measured by enzyme linked immunosorbent assay, using an acid-glycine extract antigen, in stored serum from subjects who had participated in three linked population-based surveys of cardiovascular risk factors performed in 1986 and 1987. Results. The overall prevalence of H. pylori infection was 50.5%. Prevalence increased with age from 23.4% in 12–14 year olds to 72.7% in 60–64 year olds: χ2 for trend 518, P , 10–4. In subjects aged >25, infection was more common in males (60.9%) than females (55.2%): χ2 = 9.53, P , 0.01. This relation remained significant after adjusting for age, and measures of socioeconomic class: odds ratio (OR) for infection, male versus female was 1.19 (95% confidence interval [CI] : 1.02–1.40). Infection was associated with social class: the adjusted OR of infection in subjects from manual social classes relative to those from non-manual classes was 1.7 (95% CI : 1.47–1.98). Infection was significantly more common in current smokers and ex-smokers than in subjects who had never smoked: adjusted OR for infection, ex-smokers versus never smoked was 1.22 (95% CI : 1.01–1.49); for smokers of >20/day versus never smoked OR = 1.33 (95% CI : 1.05–1.67). Infection was not associated with height in adult males but mean height in infected women was lower than in uninfected women after adjusting for age and socioeconomic status: difference in mean height (SE), –0.85 cm (0.32), P , 0.01. There was no demonstrable relationship between H. pylori infection and current alcohol intake. Conclusions. This study demonstrated a high prevalence of infection in a population from a developed country. Previously reported associations between H. pylori infection, age, sex, social class, and reduced height in females were confirmed and smoking was identified as a possible risk factor for H. pylori infection. Keywords: Helicobacter pylori, gender, age, socioeconomic status, height, smoking, alcohol, Northern Ireland

factor for the development of duodenal2 and gastric ulceration.3 It is accepted as a carcinogen, predisposing to gastric cancer,4–6 and is associated with the development of primary gastric B-cell lymphoma.7–9 It has recently been implicated in the development of ischaemic heart disease.10–12 Consequent upon this widespread prevalence and broad clinical significance, infection with H. pylori clearly constitutes an issue of major public health importance.13 Despite this, there have been few large population-based studies of the distribution and determinants of H. pylori infection in developed countries which have examined randomly selected subjects.14–16 As a result the basic epidemiology of

Helicobacter pylori infection is one of the most prevalent bacterial infections in man. Evidence is accumulating that the organism may be unique among bacterial pathogens in causing, or predisposing to, a wide range of pathological states. It causes chronic gastritis1 and increasingly is accepted as the most important risk

* Department of Epidemiology and Public Health, The Queen’s University of Belfast, Mulhouse Building, Royal Group of Hospitals, Grosvenor Road, Belfast BT12 6BJ, Northern Ireland. ** Department of Microbiology and Immunobiology, The Queen’s University of Belfast, Royal Group of Hospitals, Grosvenor Road, Belfast BT12 6BA, Northern Ireland.

880

HELICOBACTER PYLORI INFECTION IN NORTHERN IRELAND

this infection is unclear. The aim of this seroprevalence study of subjects, randomly selected from the general population in a developed country, is to determine the prevalence of H. pylori infection and the relation of infection with sociodemographic characteristics (e.g. age, sex and social class) and lifestyle factors such as smoking and alcohol intake.

METHODS Study Population Subjects studied were those recruited in Northern Ireland into the World Health Organization (WHO) MONICA (MONItoring of trends and determinants in CArdiovascular disease) Project’s second survey, the WHO co-ordinated CINDI (Countrywide Integrated Non-communicable Diseases Intervention) Programme and the Change of Heart Baseline Survey. These three population surveys were performed in 1986 and 1987 using the same protocol.17 All subjects were recruited randomly. A single-stage random sampling procedure (sampling frame—the general practitioners [GP] register held by Central Services Agency, Belfast) with stratification by age and sex was utilized for the MONICA project and the CINDI project. Subjects recruited into these two surveys resided in Belfast and neighbouring districts (the ‘MONICA area’) and were aged 25–64 years and 12–25 years, respectively. A two-stage sampling scheme was used in the Change of Heart Baseline Survey with GPs as primary sampling units: subjects recruited resided in the remainder of Northern Ireland and were aged 12–64 years. Survey Methods At screening, information was collected on medical history, diet, alcohol consumption, smoking history, education and socioeconomic status. Among other parameters, height and weight were measured and a blood sample was taken: centrifugation and separation took place within 3 hours of venepuncture. Serum was aliquoted and stored at –20ºC. The serum used for the determination of H. pylori status had been thawed on one previous occasion. Determination of Helicobacter pylori Status Helicobacter pylori specific IgG antibodies were measured by enzyme linked immunosorbent assay using an acid-glycine extract18 of H. pylori NCTC 12686. Sera were diluted 1/200 and assayed in duplicate and the mean optical density expressed as a percentage of the optical density of pooled positive control sera assayed on the same plate. The results obtained were bimodally distributed. A log-normal distribution

881

which closely approximated to the first peak of the bimodal distribution was calculated and a threshold was determined which allocated 95% of the results in this peak as negative for H. pylori antibodies. Within a test population of 75 dyspeptic patients undergoing endoscopy and antral biopsy the sensitivity and specificity of the system, when compared with culture and polymerase chain reaction, were 94.7% (54/57) and 94.4% (17/18) respectively. Positive predictive value was 98.2% (54/55) and negative predictive value 85% (17/20). Statistical Analysis Age was used in the analyses as a continuous untransformed variable. Social class was coded according to the Registrar General’s Classification of Occupations19 and was grouped into manual (IIIM, IV and V) and nonmanual (I, II and IIINM) classes. Housing tenure was categorized into rented and owned. Occupation of the head of the household and housing tenure were available for 91% and 95% of subjects aged >25, respectively, but for only 34% and 16% of younger subjects. Highest educational level achieved was categorized into four groups for subjects aged >25; elementary or secondary school, grammar school, technical college and tertiary education. In the under 25s, current or highest educational establishment attended was dichotomized into elementary or secondary school or technical college and grammar school or tertiary education. Highest or current educational establishment attended was available for 98.6% of the sample, irrespective of age group. Smoking habit was grouped into four categories: current smoker of >20 cigarettes/ day, current smoker ,20 cigarettes/day, ex-smoker and never smoked. Alcohol intake was categorized into lifetime abstinence, current abstinence, current alcohol intake less than the recommended weekly intake (14 units for women and 21 for men), 1–2 times the recommended weekly intake and more than twice the recommended weekly intake. Body mass index and height were analysed as continuous untransformed variables. Both univariate (independent samples t-test and χ2) and multivariate analyses (multiple regression and logistic regression) were performed using SPSS for Windows. The majority of analyses were performed on two groups of subjects, those aged >25 years and those ,25 years old. In multivariate analyses involving the older subjects, three measures of socioeconomic status were used; social class, housing tenure and educational level achieved. In the younger subjects current or highest educational establishment attended was used as a proxy measure of socioeconomic status.

882

INTERNATIONAL JOURNAL OF EPIDEMIOLOGY

TABLE 1 Age and social class distributions of subjects for whom serum was and was not available and of the Northern Ireland population NI populationa (%)

Subjects for whom serum not available (%)

Subjects for whom serum available (%)

Social class I II IIINM IIIM IV V

6.0 18.0 22.9 28.1 16.1 9.0

3.5 21.2 25.1 27.1 14.9 8.2

3.1 23.6 21.1 25.7 19.2 7.2

Age group (years) ,15 15–24 25–34 35–44 45–54 .55

7.2 23.2 15.8 16.2 17.8 19.9

7.1 18.9 16.6 19.6 19.4 18.4

11.7 23.2 19.4 17.0 14.8 14.0

TABLE 2 Age-specific prevalence of Helicobacter pylori infection Age group (years) 12–14 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64

From 1991 Northern Ireland Census.

RESULTS Response Rate An initial sample of 10 992 subjects aged 12–64 years was drawn between the three surveys. Some 2043 of these were age ineligible (age on contact was ,12 or .64 years), deceased, resided outside the ‘MONICA area’ or were untraced after repeated attempts. Of the 8949 eligible and traceable subjects, 5230 were screened by trained observers adhering to a standardized WHO protocol between September 1986 and July 1988 giving a response rate of 58.4%. Serum was available for anti-H. pylori antibody detection in 4742 subjects (90.7% of all recruits). Subjects for whom blood was available were similar with respect to age, sex and social class to those for whom blood was not available (Table 1). Prevalence of Infection The overall prevalence of infection was 50.5%. Prevalence increased with age (Table 2) from 23.4% in 12–14 year olds to 72.7% in 60–64 year olds (χ2 for trend 518, P , 10–4). Sex and Infection Subjects aged >25 years. Infection was more common in males (60.9% infected) than females (55.7% infected) (χ2 = 9.53, P , 0.01). This association remained after adjusting (logistic regression) for age and

No. seropositive (%)

258 403 219 264 205 213 211 183 132 141 117

79 128 144 159 161 227 277 315 291 303 312

(23.4) (24.1) (39.7) (37.6) (44.0) (51.6) (56.8) (63.3) (68.8) (68.2) (72.7)

TABLE 3 Social class distribution of Helicobacter pylori seropositivity in subjects aged >25 years Social class

a

No. seronegative

I II IIINM IIIM IV V

No. seronegative 69 355 363 299 163 82

No. seropositive 52 359 407 546 298 175

Percentage seropositive 43.0 50.3 52.9 64.6 64.6 68.1

measures of socioeconomic class (social class, tenure and educational level achieved): odds ratio (OR) for infection male versus female was 1.19 (95% confidence interval [CI] : 1.02–1.40). When this relation was further adjusted for smoking status it no longer remained statistically significant (OR = 1.13, 95% CI : 0.97–1.33). Subjects ,25 years. There was no significant difference in prevalence of infection between the sexes in this age group; 29.65% of males and 27.4% of females were seropositive for anti-H. pylori IgG. Following adjustment for current or highest educational establishment attended, the OR for infection male versus female was 1.12 (95% CI : 0.87–1.45). Socioeconomic Class and Infection Subjects aged >25 years. The social class distribution of seropositivity in this age group is shown in Table 3. There was a statistically significant trend across the social classes with the lowest rate of infection in social class I (43.0%) and the highest rate of infection in social class V (68.1%) (χ2 for trend 62.3, P , 10–6.

883


TABLE 4 Smoking and alcohol and Helicobacter pylori infection in subjects of both sexes aged >25 years No. seronegative

No. seropositive (%)

Unadjusted odds ratio of infectiona (95% CI)b

Adjusted odds ratio of infectionc (95% CI)b

Smoking status Never smoked Ex-smoker Current smoker ,20/day Current smoker >20/day

693 300 256 206

792 545 343 361

(53.3) (64.5) (57.3) (63.6)

1.58 (1.33–1.90) 1.17 (0.96–1.43) 1.53 (1.25–1.88)

1.22 (1.00–1.49) 0.97 (0.78–1.21) 1.33 (1.05–1.67)

Weekly alcohol intake Lifetime abstinence None ,Recommended limit 25 years. Helicobacter pylori infection was significantly more common in current smokers of .20 cigarettes/day (63.6% seropositive) and in exsmokers (64.5% seropositive) than in subjects whom had never smoked (53.5% seropositive). Infection was slightly more common, but not significantly so, in current smokers of ,20 cigarettes/day. These associations

held on adjusting for age, sex and measures of socioeconomic status (Table 4). Subjects aged 16–24 years. Helicobacter pylori infection was significantly more common in current smokers of >20 cigarettes/day (44.1% seropositive) than in subjects who had never smoked (30% seropositive). However, there was no association between a history of smoking or current smoking of ,20 cigarettes a day and infection and the relation between smoking >20 cigarettes/day and infection did not hold after adjustment for the available measure of socioeconomic status: the OR fell from 1.84 (95% CI : 1.06–3.18), unadjusted, to 1.39 (95% CI : 0.77–2.28), adjusted. Infection, Height and Body Mass Index (BMI) in Subjects Aged >25 Years Infection was not associated with height in adult males (Table 5) but mean height in infected women was lower than in uninfected women (159.3 cm versus 160.8 cm; t-test value 5.29, P , 0.01). This association remained after adjustment for age and socioeconomic class (multiple regression): difference in mean height (SE), –0.85 cm (0.32), P = 0.01. There was no significant difference in BMI in adult females or adult males between sero-positives and negatives after controlling for age and measures of socioeconomic class. Infection and Alcohol Intake Subjects aged >25 years. There was no significant difference in percentage seropositivity between current abstainers from alcohol, or subjects whose weekly alcohol consumption was equal to or greater than the

884


TABLE 5 Height and body mass index in subjects aged >25 years: relation with Helicobacter pylori infection Body mass index (kg/m2)

Height (cm)

Mean in uninfected (SE) Mean in infected(SE) Unadjusted differences in means (SE) Adjusted differences in means (SE)a

Males

Females

Males

Females

174.5 (0.26) 172.5 (0.21) –2.00 (0.33) –0.62 (0.35)*

160.9 (0.22) 159.3 (0.19) –1.60 (0.29) –0.85 (0.32)**

25.3 (0.14) 25.9 (0.11) 0.60 (0.18) 0.28 (0.19)†

24.9 (0.15) 25.9 (0.15) 1.00 (0.2) 0.32 (0.24)†

a

= Adjusted for social class, tenure, and educational level achieved. * P between 0.05 and 0.1. ** P , 0.01. † P . 0.1.

recommended limit for their sex, and lifetime abstainers. However, H. pylori infection was less common in subjects with a weekly alcohol intake less than the recommended intake for their sex (55.3% seropositive) than in lifetime abstainers (61.4% seropositive) (χ2 = 6.05, P = 0.01, but this relation did not hold on adjustment for age and measures of socioeconomic class (Table 4). Subjects aged 16–24 years. Helicobacter pylori infection was more common in subjects with a weekly alcohol intake of 1–2 times the recommended intake for their sex (41.7% seropositive) than in lifetime abstainers (26.8% seropositive) (χ2 = 6.8, P = 0.01), but this relation did not hold on adjustment for age and the available measure of socioeconomic class. In this age group, there were no significant differences between percentage seropositivities for H. pylori in the other categories of alcohol intake compared to lifetime abstinence.

DISCUSSION There have been very few surveys of H. pylori infection which have studied large numbers of subjects selected at random from the general population and we are aware of only one such study which has been performed in the UK.16 The current study, which is the largest seroprevalence survey of H. pylori infection that has been reported to date, utilizes randomly selected subjects and therefore avoids the inherent bias that may result from studies which have involved volunteers, blood donors or attendees at health clinics. The low overall response rate may have introduced selection bias if the non-responders were consistently different from responders in terms of characteristics which influence the acquisition of H. pylori infection, particularly age

and social class. However, the age and social class structures of the study population and that of the Northern Ireland population20 corresponded quite closely apart from an underrepresentation of 16–24 year olds in the study population. The results obtained in the study may, therefore, be applicable to the Northern Ireland population as a whole. Overall 50.5% of the study population were infected with H. pylori. This figure corresponds very closely with the prevalence of infection detected in a large population-based study performed in San Marino14 (51% infected) but the different age ranges surveyed in the two studies makes a direct comparison of the overall prevalence meaningless. Age-specific infection rates for corresponding age groups may, however, be compared between the two studies and are surprisingly similar except for a higher rate of infection in the younger Northern Irish subjects (30% of 15–24 year olds infected in Northern Ireland, 23% of similarly aged subjects in San Marino). The age-specific rates of infection observed in Northern Irish males can also be compared to the rates found in a population-based study performed in South Wales16 involving randomly selected men and using an almost identical ELISA for the detection of anti-H. pylori IgG antibodies. Although the overall prevalence in the Welsh study was higher than that found in Northern Ireland (an older age range was surveyed in Wales); for the corresponding age range (30–64 years) the prevalence of infection in Northern Irish males was higher than that of Welsh males (64.2% versus 53%). The difference was even more marked in the younger adults (54.2% of 30–44 year olds infected versus 39%). This indicates that, as in San Marino, infection with H. pylori is very prevalent in Northern Irish adults and more common than in Wales. A high proportion of teenagers and adolescents studied were also infected (e.g. 23% of 12–14 year olds). There are


no directly comparable studies which include this age band and studies involving subjects attending hospital for elective surgery in developed countries have shown both higher21 and lower22 prevalences of infection in similar age groups. However, the prevalence of infection in 12–14 year olds in this study was more than twice that observed in 11 year olds attending 30 randomly selected primary schools in Edinburgh.23 Why this should be the case is unclear. Acquisition of infection has been related to deprivation in childhood but there is little evidence that material deprivation is more profound or widespread in Northern Ireland than in the rest of the UK. Average household size in Northern Ireland is, however, larger than in the rest of the UK24 and may contribute to the high prevalence of infection by causing overcrowding within the home and fostering person-to-person spread of infection. A recent meta-analysis of seroprevalence studies of H. pylori25 revealed a sex difference in the prevalence of infection with a common OR of infection in males of 1.2 (95% CI : 1.02–1.40). The same group also demonstrated this sex difference in a study of adults aged 20–39 in Northern California.25 Our findings confirm a sex difference in H. pylori infection in adults (but not in 12–24 year olds) and the magnitude of OR we observed was very similar to that seen in the meta-analysis. Why this sex difference should exist is not known. Relative immunodeficiency in males has been postulated as an explanation for higher incidences of other infections in male children26 but, as Replogle et al.25 point out, failure to observe a sex difference in H. pylori infection in children argues against immunodeficiency in males underlying this finding. It is possible that differences in smoking habit between the sexes (32.2% of male and 52.6% of female subjects in this study aged >25 had never smoked) may alter the rate of colonization given similar exposure to the organism. Interestingly, when the relation between sex and infection in the subjects in this study (aged >25 years) was adjusted for smoking habit the association between male sex and infection was no longer statistically significant. Lack of association between infection and male sex in the subjects aged 16–25 may reflect equalization of smoking rates between the sexes in younger age cohorts: in this study; 31% and 30% of 16–25 year old males and females, respectively, were smokers. Whatever the explanation for the sex difference in H. pylori infection, the higher prevalence of infection in males may explain, at least in part, the predominance of H. pyloriassociated disease, e.g peptic ulcer disease and gastric cancer, in males. Population-based studies examining the relation between smoking and infection with H. pylori have

885

not shown any association15,16,27 and studies performed on symptomatic subjects have variously reported a positive,28,29 negative30 or no association.31 However, the positive association demonstrated in symptomatic subjects may be a spurious finding because both smoking and peptic ulceration are independently related and because adjustment for the possible confounding effect of social class was not performed. This study is the first population-based study to demonstrate an association between smoking and H. pylori infection which is independent of age, sex and socioeconomic status. It is interesting that ex-smokers had similar rates of infection to current heavy smokers suggesting that smoking may exert its effect early in life, possibly around the time of exposure and colonization e.g. in adolescence. If this were so, one might expect the relation between H. pylori infection and smoking to be strongest in adolescents. We, therefore, examined this relation in 16–25 year olds (children aged ,16 were not included in the analysis because this age group do not have legal access to cigarettes and may deny smoking irrespective of their true habit). The OR of infection in smokers of >20 cigarettes/day in this age group were higher than for older subjects but the association did not hold with adjustment for socioeconomic status, although the OR remained elevated. These data suggest that smoking may be a risk factor for H. pylori infection. This effect is unlikely to be exerted through altered exposure to infection but the effects of nicotine on gastric mucosal blood flow, mucus secretion, and epidermal growth factor secretion32 may facilitate colonization following exposure to the organism. Infection with H. pylori may, in part, explain the association between smoking and peptic ulceration and gastric cancer. On the other hand the demonstrated association between H. pylori infection and smoking may be due to inadequate adjustment for the effects of social class, although we minimized this risk by using three well recognized measures of socioeconomic status (occupational class, tenure and educational achievement). In this study an association was evident between infection with H. pylori and adult height: infected women were shorter than uninfected women. Infected males were of similar height to uninfected males. This finding agrees with that of Patel et al.23 who observed significantly reduced growth in H. pylori infected girls (but not boys) between the ages of seven and 11. However, the association between infection and reduced height may be the product of residual confounding. We were unable to control for nutritional intake before, at the time of, or after acquisition of infection. Nutritional intake during childhood is obviously a determinant of adult height and it is possible that inadequate childhood

886


nutrition may also affect susceptibility to infection by H. pylori perhaps by reducing immune competence. Alternatively, the association may be true and H. pylori infection, if symptomatic, may result in reduced nutritional intake during a critical period for growth or, because it is a chronic infection, may exert non-specific systemic effects which result in a reduction in adult height attained. However, this would not explain an effect in the females only. Perhaps, as Patel suggests,23 cytokines released by H. pylori disturb endocrine ovarian function. However, the findings of this study require replication, preferably in studies where data on nutritional intake around the critical period of acquisition of H. pylori infection are available.

CONCLUSION This study, the largest seroprevalence study of H. pylori infection performed to date, showed a high prevalence of infection in a population from a developed country. Previously reported associations between H. pylori infection, age and social class were confirmed. Infection was significantly more common in males and was also associated with reduced adult height in females and smoking was identified as a possible risk factor for H. pylori infection. REFERENCES 1

Veldhuyzen van Zanten S J O, Sherman P M. Helicobacter pylori infection as a cause of gastritis, duodenal ulcer, gastric cancer and non-ulcer dyspepsia: a systematic overview. Can Med Assoc J 1994; 150: 177–85. 2 NIH Consensus Development Panel on Helicobacter pylori in peptic ulcer disease. Helicobacter pylori in peptic ulcer disease. JAMA 1994; 272: 65–69. 3 Labenz J, Borsch G. Evidence for the essential role of Helicobacter pylori in gastric ulcer disease. Gut 1994; 35: 19–22. 4 Nomura A, Stemmerman G N, Chyou P et al. Helicobacter pylori infection and gastric carcinoma among Japanese Americans in Hawaii. N Engl J Med 1991; 325: 1132–36. 5 Parsonnet J, Friedman G D, Vandersteen D P et al. Helicobacter pylori infection and the risk of gastric carcinoma. N Engl J Med 1991; 325: 1127–31. 6 Forman D, Newell D G, Fullerton F et al. Association between infection with Helicobacter pylori and risk of gastric cancer: evidence from a prospective investigation. Br Med J 1991; 302: 1302–05. 7 Stolte M. Helicobacter pylori gastritis and gastric MALT lymphoma. Lancet 1992; 339: 745–46. 8 Hussell T, Isaacson P G, Crabtree J E, Spencer J. The response of cells from low-grade B-cell gastric lymphomas of mucosal-associated lymphoid tissue to Helicobacter pylori. Lancet 1993; 342: 571–74. 9 Wotherspoon A, Doglioni C, Diss T C et al. Regression of primary low-grade B-cell gastric lymphoma of mucosaassociated lymphoid tissue type after eradication of Helicobacter pylori. Lancet 1993; 342: 575–77.

10

Mendall M A, Goggin P M, Molineaux N et al. Relation of Helicobacter pylori and coronary heart disease. Br Heart J 1994; 71: 437–39. 11 Patel P, Mendall M A, Carrington D et al. Association of Helicobacter pylori and Chlamydia pneumoniae infections with coronary heart disease and cardiovascular risk factors. Br Med J 1995; 311: 711–14. 12 Murray L J, Bamford K B, O’Reilly D P J, McCrum E E, Evans A E. Helicobacter pylori infection: relation with cardiovascular risk factors, ischaemic heart disease and social class. Br Heart J 1995; 74: 497–501. 13 Graham D Y. Evolution of concepts regarding Helicobacter pylori: from a cause of gastritis to a public health problem. Am J Gastroenterol 1994; 89: 470–71. 14 Gasbarrini G, Pretolani S, Bonvincini F et al. A population based study of Helicobacter pylori infection in a European country: the San Marino Study. Relations with gastrointestinal diseases. Gut 1995; 36: 838–44. 15 The EUROGAST Study Group. Epidemiology of, and risk factors for, Helicobacter pylori infection among 3194 asymptomatic subjects in 17 populations. Gut 1993; 34: 1672–76. 16 Sitas F, Forman D, Yarnell J W G et al. Helicobacter pylori infection rates in relation to Welsh men. Gut 1991; 32: 25–28. 17 Evans A E, Kerr M Mc F, McCrum E E et al. Coronary risk factor prevalence in a high incidence area: results from the Belfast Monica Project. Ulster Med J 1989; 58: 60–68. 18 Newell D G, Johnston B J, Ali M H, Reed P I. An enzyme-linked immunosorbent assay for the serodiagnosis of Campylobacter pylori-associated Gastritis. Scan J Gastroenterol 1988; 23 (Suppl. 142): 53–57. 19 Office of Population Censuses and Surveys. Classification of Occupations. London: Government Statistical Service, 1980. 20 Department of Health and Social Services, Registrar General Northern Ireland. The Northern Ireland Census 1991. Belfast: HMSO, 1992. 21 Fiedorek S C, Malaty H M, Evans D L et al. Factors influencing the epidemiology of Helicobacter pylori infection in children. Pediatrics 1991; 88: 578–82. 22 Blecker U, Hauser B, Lanciers S, Peeters S, Suys B, Vandeplas Y. The prevalence of Helicobacter pylori positive serology in asymptomatic children. J Pediatr Gastroenterol Nutr 1993; 16: 252–56. 23 Patel P, Mendall M, Khulusi S, Northfield T C, Strachan D P. Helicobacter pylori infection in childhood: risk factors and effect on growth. Br Med J 1994; 309: 1119–23. 24 Central Statistics Office. Social Trends 25. London: HMSO, 1995. 25 Replogle M L, Glaser S L, Hiatt R A, Parsonnet J. Biological sex as a risk factor for Helicobacter pylori infection in healthy young adults. Am J Epidemiol 1995; 142: 856–63. 26 Green M S. The male predominance in the incidence of infectious diseases in children: a postulated explanation for disparities in the literature. Int J Epidemiol 1992; 21: 381–86. 27 Graham D Y, Malaty H M, Evans D G et al. Epidemiology of Helicobacter pylori in an asymptomatic population in the United States. Effect of age, race and socioeconomic status. Gastroenterol 1991; 100: 1495–501. 28 Bateson M. Cigarette smoking and Helicobacter pylori infection. Postgrad Med J 1993; 69: 41–44.

HELICOBACTER PYLORI INFECTION IN NORTHERN IRELAND 29

Fontham E T, Ruiz B, Perez A, Hunter F, Correa P. Determinants of Helicobacter pylori infection and chronic gastritis. Am J Gastroenterol 1995; 90: 1094–101. 30 Lindell G, Hesselvik M, Schalen C, Wikander M, Graffner H. Helicobacter pylori, smoking and gastroduodenitis. Digestion 1991; 49: 192–97. 31 Battaglia G, Di-Mario F, Pasini M et al. Helicobacter pylori infection, cigarette smoking and alcohol consumption.

887

A histological and clinical study on 286 subjects. Ital J Gastroenterol 1993; 25: 419–24. 32 Endoh K, Leung F W. Effects of smoking and nicotine on the gastric mucosa: a review of clinical and experimental evidence. Gastroenterol 1994; 107: 864–78.

(Revised version received November 1996)