Non-albicans Candida in sub-Saharan vaginal infections

The reported random-effects estimate for non-albicans Candida among all women in the included studies was 12.56%; the estimate among confirmed Candida isolates was 34.98%. These are two different model estimates, not crude observed counts. They are specific to studies in sub-Saharan Africa, and inconsistent table totals limit interpretation.

It brings regional findings on non-albicans Candida together across studies of women in sub-Saharan Africa. The results offer context for the species mix reported in those settings, but do not estimate prevalence elsewhere or guide an individual decision.

The authors searched for primary studies published from 2015 through 2025 and included 14 cross-sectional studies with 3,587 women ages 15–49 in nine sub-Saharan African countries. The studies used culture-based phenotypic or biochemical species identification; two also used molecular confirmation.

The review estimated non-albicans Candida as a proportion of all enrolled women and as a proportion of confirmed Candida isolates, and pooled proportions for four named species. Antifungal susceptibility was discussed narratively rather than pooled.

Among all women, the reported random-effects estimate was 12.56% (95% CI 8.73–17.83%; 14 studies, 3,587 women; I²=66.40%). The 14 row-level non-albicans Candida case counts in Table 1 add to 448, while its pooled row prints 507; the paper does not explain the difference. Neither count is the model estimate.

Among confirmed Candida isolates, the reported random-effects estimate was 34.98% (95% CI 25.29–46.11%; 14 studies; the paper states N=1,508; I²=68.77%). Table 2's study denominators add to 1,509, rather than the stated 1,508.

For species proportions among isolates, the reported random-effects estimates were C. glabrata 16.38% (95% CI 10.72–24.27%; 10 studies, N=1,186; I²=54.94%), C. krusei 4.52% (2.07–9.56%; 10 studies, N=1,186; I²=70.39%), C. tropicalis 4.15% (2.12–7.91%; 11 studies, N=1,370; I²=62.34%), and C. parapsilosis 1.46% (0.61–3.46%; 9 studies, N=1,031; I²=50.72%). The isolate estimates use a different denominator from the estimate among all women.

These estimates describe women ages 15–49 in the included sub-Saharan African studies. They should not be generalized as prevalence estimates for the United States or Poland, or used to predict an individual's result. The review does not test diagnosis, recurrence, treatment response, or screening benefit.

All included studies were cross-sectional, and the pooled estimates had high heterogeneity. Settings and species-identification methods varied. The 14 row-level non-albicans Candida case counts in Table 1 total 448 although its pooled row says 507; Table 2 denominators total 1,509 although the text states 1,508. Neither discrepancy is resolved in the article.

Resistance was not pooled: five studies reported susceptibility findings using varied or unstated methods and breakpoints. No complete species-specific MIC data interpreted against current EUCAST or CLSI breakpoints were available. The review also did not pool recurrence, treatment response, or patient-reported outcomes, and conventional identification methods may misclassify some species.

Contextual. The review adds sub-Saharan African prevalence and species-distribution estimates. It does not establish a global or locale-specific prevalence, individual risk, or treatment effect; the inconsistent printed totals further limit reuse of the aggregate counts.

No practical change. This regional, cross-sectional review does not change clinical guidance or establish an individual diagnosis or treatment choice.

The authors reported no external funding and said the work relied on collaboration among the co-authors. They declared no competing interests.

Read the original publication (published 24 July 2026).