Annals of Global Health (Jan 2020)
Prevalence of Type 2 Diabetes Mellitus in Adult Population of Pakistan: A Meta-Analysis of Prospective Cross-Sectional Surveys
Abstract
Background: The clinical and methodological diversity observed in national and regional diabetes surveys, emphasized on the need of the weighted average prevalence of diabetes. Objective: To measure the pooled prevalence of type 2 diabetes mellitus in the adult population of Pakistan. Methods: The prospective cross-sectional studies reporting adult diabetes in Pakistan and published on any date were retrieved from PubMed, ScienceDirect and PakMediNet databases. In the meta-analysis, PRISMA guidelines were used for reporting; the AXIS tool for assessing quality and risk of bias within studies; I2 statistics for measuring heterogeneity between studies and subgroups; and Tableau Public 10.4 for geographic mapping of included studies. Using Meta-Analyst 3.13 βeta, overall and subgroup pooled estimates were measured by random effects model. Results: The pooled sample of twelve studies included 42,051 adults (≥20 years) comprised of both sexes from urban and rural Pakistan. The pooled prevalence of diabetes was 13.7% (95% CI, 10.7–17.3). None of the twelve studies was of poor quality (<10 scores). Ten studies were published in ISI indexed journals, and nine of them were indexed for Medline. The level of heterogeneity observed across studies and between subgroups was moderate (<50%). The subgroup analysis revealed a higher pooled estimate of diabetes in males than in females (13.1 vs. 12.4%). It was also higher in urban than in rural patients (15.1 vs. 1.6%), and in HbA1c than in OGTT tests (23.9 vs. 14.4%). However, pooled estimates of the WHO and the ADA criteria were similar (13.8 vs. 13.5%). Conclusions: The prevalence of diabetes is on the rise in the adult population of Pakistan. The heterogeneity across studies observed in the meta-analysis suggested that the design of future diabetes surveys should be efficient and purposeful, and that valid tools and methods should be used to generate more precise data. Moreover, harmony between the stakeholders is much needed to seek a true picture of the diabetes burden in the country.