Practicing What They Preach: Health Behaviors of Those Who Provide Health Advice to Extensive Social Networks

Abstract

As a way of identifying a conduit to disseminate health information, this study aims to explore health behaviors and attitudes of a unique group of extensively socially-networked individuals who regularly are asked for their health advice. Respondents from a population-based consumer opinion panel (n = 2,639) were categorized as “extensively socially-networked” (75+ friends and acquaintances, and almost daily giving friends advice on general issues) vs. “non-networked.” The networked respondents were further divided into “health-networked” (regularly asked for health advice) versus “only-socially-networked” groups (asked for general advice, not health). Chi-square analyses, ANOVA tests, and multivariate regressions controlling for sociodemographic variables compared health behaviors and attitudes between groups. Results indicated that health-networked individuals reported more positive health behaviors (e.g., fruit and vegetable consumption) and attitudes than only-socially-networked and non-networked individuals. Future research is warranted to elucidate how providing health advice to a large network contributes to the positive health of health-networked individuals. Exploratory analyses revealed that doctors and health/fitness magazines were main sources of health and nutrition information for health-networked respondents. Through their advice and word-of-mouth, health-networked individuals have the potential to influence the health information of large groups of people and, therefore, may serve as valuable change agents to disseminate health and nutrition information.

Notes

¹Porter Novelli, 1909 K Street NW, Washington, DC 20006.

²Synovate, Inc., 222 South Riverside Plaza, Chicago, IL 60606.

^a P values are based on ANOVA or Pearson X² tests for continuous variables or categorical variables, respectively.

^b Includes individuals who were never married or are widowed, separated, or divorced.

^a Responses were poststratified and weighted according to U.S. Census benchmarks on age, sex, race/ethnicity, income, and household size.

^b P values are based on ANOVA or Pearson X² tests for continuous variables or categorical variables, respectively.

^c ANOVA tests indicate which means are significantly different from each other (p < 0.05).

^d Indicates which means are not significantly different from each other (p > 0.05).

†Percent that comply with 30 minutes/day or more.

^£Values are 5-pt scale mean ± SD or percentages. 5-pt scale from “not at all” to “a lot.”

^a Models controlled for gender, age, education, income, location of residency, ethnicity, and marital status. All variables were included in the model simultaneously.

Values are statistically significant (p < 0.05).

^a Models controlled for gender, age, education, income, location of residency, ethnicity, and marital status. All variables were included in the model simultaneously.

†Percent that comply with 30 minutes/day or more.

∗Values are statistically significant (p < 0.05).

†Values are 5-pt scale mean ± SD or percentages. 5-pt scale from “not at all” to “a lot.”