Methods

 \(\)
All data was collected from the Demographic and Health Surveys (DHS) public website.  Data from 73 separate DHS Surveys, conducted over the time period of 1986-2016 was initially used, and was eventually parsed down based on specifications. Each DHS survey was conducted in a separate year and country.  DHS surveys from Sub-Saharan Africa were chosen for this analysis for multiple reasons; they provided the most complete data on both fertility and HIV status, and Sub-Saharan Africa is the geographical area in which most previous studies on the effect of HIV seropositive status on fertility have been conducted.
The two indicators used to compare the effect of HIV status on fertility were:
1. “HIV Status Among Women Age 15-49”, which the DHS defined as the percentage of women aged 15-49 in the survey who tested positive to HIV.
2. “Fertility Rate”, which the DHS defined as the total fertility rate for the three years preceding the survey for age group 15-49 expressed per woman.
 Additionally, the third variable, “Ever Use of Injections”, which was used as a possible confounding variable in the multiple linear regression analysis, was defined by the DHS as the percentage of women surveyed who reported ever having used injectable contraceptives.
Restrictions were placed on the data set in order to select for relevant data points.  Only surveys which contained values in both the “HIV Status Among Women Age 15-49” and “Fertility Rate” were selected for the linear regression, which resulted in a data set containing 55 surveys among 29 countries in Sub-Saharan Africa. A linear regression analysis was then conducted using R-studio. All statistical tests were conducted with a significance level of α=0.05. 
    For the multiple linear regression of HIV and Fertility rate + Injectable Contraceptive Use, the data were again restricted to include only surveys which had values for all three indicators. This resulted in a data set of 23 DHS Surveys
    Studies of the effect of HIV on fertility are difficult to quantify due to HIV seropositive status likely having multiple interacting effects on fertility, including possible biological factors which may lead to decreased fertility, confounding effects of contraceptive choice, and behavioral factors which may cause a women to choose not to seek pregnancy.  In order to further investigate the true effect of HIV seropositive status as a biological predictor of fertility, I chose an additional confounding variable of “Ever Use of Injections” for use in a multiple linear-regression