Abstract
INTRODUCTION: Helicobacter pylori (H. pylori) is predominantly acquired in childhood and persists as a chronic infection in the stomach. H. pylori has been linked with household crowding and is an important causal factor which is considered necessary for non-cardia stomach cancer. Māori and Pacific peoples in New Zealand experience greater household crowding, H. pylori infection and stomach cancer incidence.
OBJECTIVES: 1) To summarise the association between household crowding and H. pylori infection in the literature; and estimate the contribution of household crowding to Māori, Pacific and European H. pylori seroprevalence. 2) To estimate the excess Māori and Pacific stomach cancer incidence that is attributable to H. pylori exposure and smoking.
METHODS: A systematic literature review and meta-analysis was conducted to summarise the evidence on the association between household crowding density and H. pylori, and investigate heterogeneity.
Pooled serology data were regionally weighted and adjusted to estimate the Māori, Pacific and European seroprevalence of H. pylori by 1926-40, 1941-55, 1956-70 and 1971-85 birth cohorts. In the latter cohort, the meta-analysis odds ratio and household crowding prevalence from the 1986 census were used to estimate the contribution of crowding to H. pylori seroprevalence.
For the 1926-40, 1941-55 and 1956-70 birth cohorts, age-standardised Māori and Pacific incidence rates of stomach cancer were compared to European/Other. Rate ratios (RRs) were adjusted for H. pylori by restricting to the proportion of each ethnic group who were seropositive, assuming that H. pylori is a necessary causal factor for stomach cancer. RRs were adjusted for smoking by probabilistic bias analysis. Adjusted RRs were compared with observed RRs to calculate an ‘excess rate ratio proportion’ which estimates the contribution of H. pylori and smoking to excess Māori and Pacific stomach cancer incidence. In order to better meet the necessary factor proposition, analysis was then restricted to non-cardia stomach cancer cases.
RESULTS: Meta-analysis showed that persons experiencing the greatest vs. the least household crowding had 1.73 (95% CI 1.48-2.03, n=28, I2=87%) times greater odds of H. pylori infection. Children exposed to household crowding had significantly greater risk of infection (OR 2.06, CI: 1.53-2.77, n=19, I2=86%). The average pooled H. pylori seroprevalence was greatest for Pacific (62%), followed by Māori (35%) and European (18%). Seroprevalence declined in subsequent birth cohorts for all ethnic groups but relative ethnic differences in seroprevalence increased. Household crowding among children born 1971-85 contributed to 44% of Pacific (95% CI: 32-54%), 36% of Māori (95% CI: 25-47%), and 14% of European (95% CI: 9-20%) H. Pylori seroprevalence.
For men born in the 1926-40, 1941-55 and 1956-70 birth cohorts, H. pylori and smoking (to a lesser degree) contributed to more than half of the excess non-cardia stomach cancer among Māori (53%, 65%, 63% respectively) and approximately eight-tenths of the excess non-cardia stomach cancer among Pacific (74%, 83%, 90% respectively).
CONCLUSION: This analysis is distinctive because it quantifies a two-step process from household crowding to H. pylori infection, and H. pylori infection to stomach cancer; and then distinguishes how this process differs by ethnicity. Household crowding is a major contributing factor to Māori and Pacific H. pylori seroprevalence; and the primary driver of excess non-cardia stomach cancer incidence among Māori and Pacific. Household crowding reduction interventions that focus on children are recommended. H. pylori ‘screen and treat’ approaches for asymptomatic Pacific and Māori men in NZ require evaluation. Improved sensitivity in H. pylori detection measures and better stomach cancer subsite classification will improve future evaluation of the H. pylori contribution.