The first example is a cross-sectional study of social relationships and postpartum depression.


A stratified random sample of 412 African-American, Hispanic & White women who had recently given birth was obtained from patient lists of urban health centers serving low and moderate income patients.

Relationships of interest:

  • Social support and postpartum depressive symptoms
  • Social networks and postpartum depressive symptoms


To investigate these relationships, researchers administered questionnaires to gather information about the following:

  • Social Support
  • Social Networks
  • Sociodemographics
  • Postpartum depression

In this study, researchers used previously developed instruments to measure each variable.

Four types of social support were measured using the 20 item Medical Outcomes Study (MOS) Social Support Survey (1991)&ask;

  • Emotional/info support
  • Tangible support
  • Positive interaction
  • Affection

Social networks were determined using questions for a previous study. Participants were asked questions such as how many friends and relatives she felt close to.

Potential confounders (variables that may distort existing relationships) were assessed using items from the Breast Cancer Core Questionnaire.

  • Time since pregnancy
  • Number of children under age 5
  • Stable housing
  • Discrimination experiences
  • Crowding

&ask;For more on the MOS Social Support Survey see Sherbourne, C., & Stewart, A.L. (1991). The MOS Social Support Inventory. Social Science Medicine, 32(6), 705-714.

The outcome in this study, postpartum depressive symptoms, measured by a widely used instrument, the Center for Epidemiologic Studies of Depression Scale (CES-D).

Additionally, instrument translations were available when necessary.

Data Analysis & Findings

Linear regression models were created to examine relationships statistically. Findings from the study showed:

  • Increases in social support scores were associated with decreases in scores on the depressive symptoms scale.
  • Women with 2 or more friends or family members had fewer depressive symptoms (9 point lower CES-D score) compared to those without supportive friends or family members and those with only one.


Although a cross-sectional study design cannot provide insight about causality, and findings cannot be generalized to all women, this study shows one way the social networks and social support model may be used to explain a specific health outcome like postpartum depressive symptoms in a particular group of women.

Note: Although it is not explicitly stated, it appears that pathway 3 is under investigation.