Application of Geographically Weighted Regression Analysis to Assess Predictors of Early Sexual Initiation in Ethiopia, EDHS 2016

ABSTRACT Background and Aims Early sexual initiation was defined as the experience of first intercourse before 18 years of age. It has negative health, social, and economic consequences with an increased lifetime prevalence of sexual partners, thereby increasing the risk of exposure to sexually transmitted diseases. Thus, this study aimed to assess the spatial relationship of factors associated with early sexual initiation and identify its predictors in Ethiopia. Methods The data were obtained from the Measure demographic and health survey website and extracted through SPSS and Stata software. To account for disproportionate sampling and non‐response, sample weights were applied. A total weighted samples of 15,683 women were selected and the DHS used a stratified, two‐stage cluster sampling technique. Six hundred forty‐five enumeration areas with latitude and longitude coordinates were used. ArcGIS version 10.5 software was used to analyse spatial statistics. The geographically weighted regression analysis technique was used to identify predictors of early sexual initiation. Result The result of geographically weighted regression analysis has identified different variable coefficients through different geographical locations. Accordingly, higher coefficients for rural residence were detected in Dire Dawa, Harari region, north‐eastern part of Oromia and northern part of Somali, The other spatial predictor was poorest wealth index and its higher coefficients were detected in Addis Ababa, Benishangul Gumuz, western part of Oromia, northern part of SNNP and Amhara region, and being from female headed household was also an important predictor with higher coefficients detected in Afar, all parts of Tigray region, the northern part of Amhara region, Benishangul Gumuz region and Wellga zone of Oromia. Conclusion The study revealed significant spatial variation in predictors of early sexual initiation, with rural residence, poorest wealth index, and female‐headed households showing stronger effects in specific regions. High‐risk areas include Dire Dawa, Harari, parts of Oromia, Addis Ababa, Benishangul Gumuz, and Tigray. Targeted, region‐specific interventions focusing on poverty reduction, rural youth support, and empowerment of female‐headed households are essential to address early sexual initiation effectively.