Longitudinal analyses suggest a higher burden of self-reported influenza-like illness among women and girls in a rural community in western Kenya

• Published in: Global Epidemiology Vol. 3; p. 100059
• Main Authors: Chweya, R.N., Mambo, S.N., Gachohi, J.M.
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 01.11.2021; Elsevier
• Subjects: Illness; Influenza-like; Syndrome; Western Kenya; Influenza-like; Western Kenya; Syndrome; Illness
• ISSN: 2590-1133; 2590-1133
• DOI: 10.1016/j.gloepi.2021.100059

Longitudinal studies that quantify influenza-like illness (ILI) burden while determining associated risk factors and intra-cluster correlation coefficients (ICCs) are scarce in Africa. This study aimed to identify longitudinal patterns of ILI and the associated independent variables while quantifying the burden by age and sex. We enrolled and followed 390 subjects nested within 92 households for 12 weeks, collecting weekly data on prespecified self-reported symptoms in a rural community in western Kenya. Using a standard case definition, reporting at least any four prespecified symptoms attained an ILI syndrome. Analyses related relationship between attaining ILI syndrome and independent variables using multilevel logistic regression models adjusting for age and sex. ICCs quantified clustering of ILI syndrome computed at different levels of data aggregation. Of the 4261 reports received, 592 (14%) registered an illness, with 278 (48%) of these attaining ILI syndrome. The 278 reports belonged to 176 (41%) subjects dwelling in 76 (83%) households. On average, each study subject and household experienced 1.6 and 3.7 episodes of ILI syndrome, respectively. Making a visit outside the local area of residence (Odds ratio (OR) =2.7, 95% CI 1.8, 4.1) and living in a cement-floored house (OR = 1.9, 95% CI 1.1, 3.3) independently predicted attaining ILI syndrome. Correcting for the population structure revealed an excess risk of 16% in attaining the syndrome among the study women and girls. Clustering of ILI syndrome was lowest at the individual level (ICC = 0.04) and slightly higher within households (ICC = 0.14), suggesting moderate household contextual influences. Regardless of the cause, we provide evidence of a potentially higher ILI syndrome burden among women and girls in our study area. Studies are needed to establish granular exposures defining the increasing risk of ILI syndrome associated with gender, making a visit outside the local sub-county of residence and living in a cement-floored house.