A meta-analysis of the worldwide prevalence of pica during pregnancy and the postpartum period.

Abstract

Although pica has long been associated with pregnancy, the exact prevalence in this population remains unknown. To estimate the prevalence of pica during pregnancy and the postpartum period, and to explain variations in prevalence estimates by examining potential moderating variables. PsycARTICLES, PsycINFO, PubMed, and Google Scholar were searched from inception to February 2014 using the keywords pica, prevalence, and epidemiology. Articles estimating pica prevalence during pregnancy and/or the postpartum period using a self-report questionnaire or interview were included. Study characteristics, pica prevalence, and eight potential moderating variables were recorded (parity, anemia, duration of pregnancy, mean maternal age, education, sampling method employed, region, and publication date). Random-effects models were employed. In total, 70 studies were included, producing an aggregate prevalence estimate of 27.8% (95% confidence interval 22.8-33.3). In light of substantial heterogeneity within the study model, the primary focus was identifying moderator variables. Pica prevalence was higher in Africa compared with elsewhere in the world, increased as the prevalence of anemia increased, and decreased as educational attainment increased. Geographical region, anemia, and education were found to moderate pica prevalence, partially explaining the heterogeneity in prevalence estimates across the literature.