Data on gestational diabetes (GDM) in sub‑Saharan Africa, particularly differentiating early and late pregnancy onset and examining underlying pathophysiology, are limited. The ORCHID cohort study in South Africa aimed to assess the prevalence of GDM among pregnant women with and without HIV, characterize the timing of GDM (early versus late), and investigate measures of insulin sensitivity and beta‑cell function to better understand glucose metabolism in pregnancy.
The study enrolled 1,573 pregnant women at or before 18 weeks' gestation. Among these participants, 668 were women with HIV (WWH). The median age across the cohort was 28 years and the median gestational age at enrollment was 13 weeks. Those basic demographic descriptors set the context for prevalence and physiologic comparisons by HIV status and by timing of GDM diagnosis.
Participants underwent a 75 g oral glucose tolerance test (OGTT) at enrollment (early pregnancy) and again at 32–36 weeks' gestation (late pregnancy). GDM diagnosis followed WHO criteria as applied to the 75 g OGTT at each time point, enabling classification of cases as early or late GDM depending on when diagnostic thresholds were met.
To explore underlying glucose metabolism, the study calculated several indices from OGTT and model outputs: the Matsuda index and Stumvoll index as measures of insulin sensitivity; the oral disposition index as an integrated measure of insulin secretion adjusted for sensitivity; and glucose sensitivity parameters derived from Mari modeling to quantify beta‑cell responsiveness to glucose.
Logistic regression models were used to evaluate the association between HIV status and the presence of GDM. Models were adjusted for covariates, though the abstract does not list which covariates were included. The analysis compared odds of GDM in WWH versus HIV‑seronegative women.
Across the full cohort, the overall prevalence of GDM was 7.9%. By HIV status, prevalence was 6.7% among WWH and 9.1% among HIV‑seronegative women. Among all participants diagnosed with GDM, 65% were identified at the early assessment (enrollment), indicating that a substantial proportion of GDM cases in this population were present in early pregnancy.
Women who met criteria for early GDM demonstrated the most pronounced impairments in the physiologic indices measured at enrollment. Specifically, early GDM cases had the lowest Matsuda and Stumvoll indices (reflecting reduced insulin sensitivity), the lowest oral disposition index (indicating impaired insulin secretion relative to sensitivity), and the lowest glucose sensitivity from Mari models compared with those who developed late GDM and those who did not develop GDM. These findings suggest that early GDM in this cohort is characterized by combined defects in insulin action and beta‑cell responsiveness.
In adjusted logistic regression analyses, women with HIV had lower odds of GDM than HIV‑seronegative women, as reported in the abstract. The abstract states this direction of association but does not provide the numeric adjusted odds ratio, confidence interval, or the list of covariates included in adjustment. These specific effect estimates and model details were not reported in the abstract and would require consultation of the full text for confirmation.
This study provides prevalence data for early and late GDM in a South African cohort and indicates that early GDM is associated with marked impairments in insulin sensitivity and beta‑cell function. The lower observed prevalence among WWH in adjusted models raises questions about interactions between HIV status, antiretroviral therapy, and glucose metabolism in pregnancy; however, the abstract does not present the adjusted effect estimates or model covariates needed to interpret causality. Future work should report full model results, explore mechanisms linking HIV and GDM, and consider implications for screening timing and management strategies in similar settings.
The abstract supplies cohort size, basic demographics, prevalence estimates, timing of GDM diagnoses, and comparative physiologic findings. However, several details are not reported in the abstract and are therefore not available here: the numeric adjusted odds ratio and confidence intervals for the association between HIV and GDM; the covariates included in adjusted models; detailed numeric results for the physiologic indices with measures of variance; and clinical management or outcomes for women diagnosed with GDM. These items would require access to the full article to extract.