RESUMEN
INTRODUCTION: Long-term persistent Helicobacter pylori infection has been associated with ulceropeptic disease and gastric cancer. Although H. pylori is predominantly acquired early in life, a clear understanding of infection dynamics during childhood has been obfuscated by the diversity of populations evaluated, study designs, and methods used. AIM: Update understanding of true prevalence of H. pylori infection during childhood, based on a critical analysis of the literature published in the past 5 years. METHODS: Comprehensive review and meta-analysis of original studies published from 2011 to 2016. RESULTS: A MEDLINE® /PubMed® search on May 1, 2016, using the terms pylori and children, and subsequent exclusion, based on abstract review using predefined criteria, resulted in 261 citations. An Embase® search with the same criteria added an additional 8 citations. In healthy children, meta-analysis estimated an overall seroprevalence rate of 33% (95% CI: 27%-38%). Seven healthy cohort studies using noninvasive direct detection methods showed infection prevalence estimates ranging from 20% to 50% in children ≤5 and 38% to 79% in children >5 years. The probability of infection persistence after a first positive sample ranged from 49% to 95%. Model estimates of cross-sectional direct detection studies in asymptomatic children indicated a prevalence of 37% (95% CI: 30%-44%). Seroprevalence, but not direct detection rates increased with age; both decreased with increasing income. The model estimate based on cross-sectional studies in symptomatic children was 39% (95% CI: 35%-43%). CONCLUSIONS: The prevalence of H. pylori infection varied widely in the studies included here; nevertheless, model estimates by detection type were similar, suggesting that overall, one-third of children worldwide are or have been infected. The few cohort and longitudinal studies available show variability, but most studies, show infection rates over 30%. Rather surprisingly, overall infection prevalence in symptomatic children was only slightly higher, around 40%. Studies including only one positive stool sample should be interpreted with caution as spontaneous clearance can occur.