**Background:** Histoplasma capsulatum, the causative agent of histoplasmosis, is recognized as a priority disease of public health concern in Kenya, yet its burden in the general population is poorly documented. Histoplasmin skin sensitivity surveys in sub-Saharan Africa have reported positivity rates ranging from 0.0% to 35.0%, but data from Kenya are limited. Known risk factors include exposure to bat and bird habitats, soil disturbance, and immunocompromise (especially HIV co-infection), but contextual factors relevant to rural western Kenya remain unexplored.
**Methods:** The study utilized biobanked serum samples and household survey data from the People, Animals and their Zoonoses (PAZ) project, a cross-sectional survey conducted from 2010–2012 in Busia county, western Kenya. A subset of 670 respondents (from 178 households within 102 sub-locations) was selected based on availability of data on bat observation and HIV status. All respondents with HIV-positive status (n=48) and those reporting bat observation (n=348) were included, plus every eighth remaining respondent. Serum samples were tested using the IMMY Latex Agglutination Histoplasma test (LAT), which detects IgM-predominant anti-Histoplasma antibodies. Samples with reaction strength ≥2+ were considered positive. Apparent seroprevalence was calculated directly; true seroprevalence was estimated using published LAT sensitivity (62%) and specificity (97%). Univariable associations were assessed using chi-squared tests, and variables with p<0.20 were entered into a multi-level multivariable logistic regression model with household and sub-location as random effects. A conservative p<0.10 was used for retention in the final model.
**Key Results:** The apparent seroprevalence of anti-Histoplasma antibody was 15.5% (n=104/670, 95% CI 12.9–18.5%). The estimated true seroprevalence was 21.2% (95% CI 16.8–26.2%). Among seropositive samples, 65.4% showed a 2+ reaction strength, 34.6% showed 3+, and none showed 4+. At the household level, 43.3% of households (77/178) had at least one seropositive occupant. On univariable analysis, age category 15–24 years (OR=2.80, 90% CI 1.28–6.15, p=0.03) and observation of rats around the home (OR=2.80, 90% CI 1.17–6.68, p=0.05) showed significant associations. In the final multi-level multivariable model, two variables remained statistically significant: observation of rats within the household in the previous 12 months (OR=2.99, 90% CI 1.04–8.55, p=0.04) and age category 15–24 years (OR=2.70, 90% CI 1.04–6.97, p=0.04; reference: 25–34 years). Two additional variables met the p<0.10 inclusion threshold: age ≥45 years (OR=2.60, 90% CI 0.98–6.89, p=0.06) and mud wall construction (OR=2.50, 90% CI 0.85–7.36, p=0.097). Clustering by household (variance=0.02, SE 0.18) and sub-location (variance=0.18, SE 0.17) accounted for only 0.6% and 5.2% of variance, respectively. The Hosmer-Lemeshow test indicated good fit (χ²=2.091, p=0.970).
**Clinical Implications:** The 15.5% apparent seroprevalence indicates that H. capsulatum exposure is common in this rural Kenyan population, providing a baseline for future epidemiological studies and sample size calculations. The association with rat observation suggests that peridomestic rodent populations may serve as environmental reservoirs or indicators of contaminated soil, though direct isolation of H. capsulatum from rats or burrows in this region is needed. The elevated odds in adolescents and young adults (15–24 years) may reflect age-related differences in occupational or recreational soil exposure. The lack of significant clustering at household or sub-location level suggests exposure is widespread rather than focal. Limitations include the cross-sectional design (precluding temporal inference), the moderate sensitivity of the LAT (62%), potential cross-reactivity with other fungi (e.g., Aspergillus), and possible underestimation of seropositivity among HIV-positive individuals due to impaired antibody responses. The study did not assess clinical outcomes, so the proportion of seropositive individuals who develop symptomatic disease remains unknown. Prospective longitudinal studies incorporating molecular detection, clinical follow-up, and detailed environmental sampling are warranted to clarify transmission dynamics and disease burden.