Diabetes comorbidities in low- and middle-income countries: An umbrella review

Background: Diabetes mellitus, particularly type 2 diabetes, is a major public health burden globally. Diabetes is known to be associated with several comorbidities in high-income countries. However, our understanding of these associations in low- and middle-income countries (LMICs), where the epidemiological transition is leading to a growing dual burden of non-communicable and communicable disease, is less clear. We therefore conducted an umbrella review to systematically identify, appraise and synthesise reviews reporting the association between diabetes and multiple key comorbidities in LMICs.

Methods: We searched Medline, Embase, Global Health, and Global Index Medicus from inception to 14 November 2020 for systematic reviews, with or without meta-analyses, of cohort, case-control or cross-sectional studies investigating the associations between diabetes and cardiovascular disease (CVD), chronic kidney disease (CKD), depression, dengue, pneumonia, and tuberculosis within LMICs. We sought reviews of studies focused on LMICs, but also included reviews with a mixture of high-income and at least two LMIC studies, extracting data from LMIC studies only. We conducted quality assessment of identified reviews using an adapted AMSTAR 2 checklist. Where appropriate, we re-ran meta-analyses to pool LMIC study estimates and conduct subgroup analyses.

Results: From 11 001 articles, we identified 14 systematic reviews on the association between diabetes and CVD, CKD, depression, or tuberculosis. We did not identify any eligible systematic reviews on diabetes and pneumonia or dengue. We included 269 studies from 29 LMICs representing over 3 943 083 participants. Diabetes was positively associated with all comorbidities, with tuberculosis having the most robust evidence (16 of 26 cohort studies identified in total) and depression being the most studied (186 of 269 studies). The majority (81%) of studies included were cross-sectional. Heterogeneity was substantial for almost all secondary meta-analyses conducted, and there were too few studies for many subgroup analyses.

Conclusions: Diabetes has been shown to be associated with several comorbidities in LMICs, but the nature of the associations is uncertain because of the large proportion of cross-sectional study designs. This demonstrates the need to conduct further primary research in LMICs, to improve, and address current gaps in, our understanding of diabetes comorbidities and complications in LMICs.