Background: There are theoretical and empiric concerns about how socioeconomic status/position (SES) is commonly measured in health literature. We aimed to provide conceptual and empiric bases for selecting and interpreting SES measures in studies of low birthweight, delayed prenatal care, unintended birth, and breastfeeding.
Methods: Data are from 2 large statewide-representative surveys (1994; 1999) of California's maternity population and census estimates and projections. Overall and by ethnicity, we: (1) examined correlations among multiple measures of income and education at individual/household and neighborhood (census tract, zip) levels of aggregation; (2) constructed logistic regression models for unadjusted associations between each SES measure and outcome; and (3) developed multivariate models using different SES measures/combinations to explore the associations of ethnicity with these outcomes while adjusting for SES.
Results: Correlations between income and education measures at different aggregation levels were modest and varied by ethnicity. Unadjusted associations between SES and outcomes varied by SES measure and outcome, and these associations varied by ethnicity. Conclusions regarding independent associations between ethnicity and outcomes depended upon scale (continuous vs. categorical), aggregation level, and number of SES measures.
Conclusions: Education, an important SES measure in its own right, is not an acceptable income proxy, and neighborhood measures are not proxies for individual-level ones, overall or among different ethnic groups. Neither education nor income alone sufficiently represents "SES". Selecting and interpretating SES measures must be outcome- and population-specific. For some of these outcomes, measurement scale seemed less important than aggregation level and SES factor (e.g., education or income) represented.
Learning Objectives: 1. Provide conceptual and empiric bases for selecting and interpreting SES measures in studies of low birthweight, delayed prenatal care, unintended birth, and breastfeeding. 2. Provide conceptual and empiric bases for interpreting associations of race or ethnic group in studies of low birthweight, delayed prenatal care, unintended birth, and breastfeeding when adjusting for SES.
Keywords: Measuring Social Inequality, MCH Epidemiology
Presenting author's disclosure statement:
Organization/institution whose products or services will be discussed: None
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.