Here we used recent secondary cross-sectional data from Bangladesh Demographic and Health Survey (BDHS), 2014. This data were available at DHS programmed website (https://dhsprogram.com/data/available-datasets.cfm). The data is nationally representative and ethically approved (NIPORT, 2016). Our sample included all ever married women age of 15-49 years.
In the DHS project researchers and policymakers were interested in constructing a measure of economic status that would be independent of other demographic characteristics. In this survey direct estimates of household income and expenditures are desirable but not practical (Rutstein et al., 2014). This index is a compound measure of a household’s cumulative living standard. This index is a proxy measure of using data on a household’s ownership of selected assets (e.g. TV, bicycles, radio) and materials used for housing construction such as types of sanitation and water access. This index is generated with a popular statistical procedure known as principle component analysis. (NIPORT, 2016). The lowest scores and highest scores illustrating poorest and richest households respectively. The wealth index was valuable in those countries that lack reliable on income and expenditures, whose were the traditional measures of household economic status.
Multivariate statistical analyses such as, Multifactor Analysis (MFA) and Multiple Correspondence Analysis (MCA) has been performed to conduct the study. In MFA, we divided the all-important risk factors into two groups i.e., Reproductive Group (Group 1) and the second group is Demographic Factors group (Group 2). In first group, the variables are Age, Age at first birth, Menopausal Status, Marital Status, Contraceptive use, No of child, BMI. In second group the variables are division, Educational status, TV. Watching, Household member, Husband education, Residence, Current working status, and wealth index is a supplementary variable. The group separation table is given in the following Table