Data for this study came from a cohort of persons >18 years of age who were living in the New York City metropolitan area on September 11, 2001. The cohort was recruited through a random digit dial telephone survey from March 25 through June 25, 2002. Contact information was obtained and follow-up interviews were conducted from September 25, 2002, through January 31, 2003, and from September 24, 2003, through February 29, 2004. Data regarding knowledge of, worry about, and attitudes toward AIDS and SARS were collected among a randomly selected subset of participants in the second cohort follow-up. The total sample included in this analysis, for whom data about AIDS, SARS, and social support were incomplete, consisted of 914 persons; response rate for eligible participants was 56%. Each participant received a $10 incentive. Additional details on sampling are provided elsewhere (5). The Institutional Review Board of the New York Academy of Medicine reviewed and approved this study.

We collected information about respondent sex, race/ethnicity, age, educational attainment, marital status, and household income at baseline. Social support was assessed by using a 5-item modified version of the Medical Outcomes Study social support scale (6); this abbreviated scale has a Cronbach α of 0.90 (7). Social support was categorized as low, medium, or high, on the basis of tertiles of support reported in the sample. We assessed respondents’ knowledge of, worry about, and attitudes toward AIDS and SARS (Table 1). Further details on measures are provided elsewhere (1,2,5).

We described sociodemographic characteristics and level of social support of the respondents and created multivariable models predicting knowledge of, worry about, and stigmatizing attitudes toward AIDS and SARS, including covariates thought to potentially confound the relationship between social support and each outcome. Analyses were weighted to correct potential selection bias related to the number of household telephones, persons in the household, and oversampling, as well as to make the sample demographically similar to the New York City metropolitan area population according to US Census 2000 estimates. We used SUDAAN software to estimate standard errors and adjust analyses for the weighting (8).

Of the 914 participants, 45.4% were male, 54.1% white, 7.6% Asian or other race, 18.5% black, and 19.8% Hispanic (Table 2). In multivariable models controlling for sociodemographic variables (Appendix Table), persons with low or medium levels of social support were significantly more likely than those with high levels of social support to report being poorly informed about AIDS (p = 0.004 for low support , p = 0.038 for medium support), being very worried about AIDS (p = 0.010 for medium support) and SARS (p = 0.010 for medium support), and to express more stigmatizing attitudes toward SARS (p = 0.020 for low support).

Reporting higher levels of social support was independently associated with greater knowledge of AIDS but not of SARS. Lower levels of social support were associated with more stigmatizing attitudes toward SARS but not AIDS. These patterns may be explained by differences in the epidemiology of AIDS and SARS in New York City. AIDS has been associated with tremendous illness and mortality rates in New York City (9,10). In contrast, the SARS epidemic was declared globally contained by the time of data collection (11). The psychological pathways through which social support influences responses to EIDs may be dependent on the effect of the epidemic on the local population. For example, greater social support may be associated with greater knowledge of AIDS because, relative to SARS, AIDS is more prevalent and persons with greater social support, through their social relationships, may have greater access to information that influences development of knowledge about EIDs.

Social support was inversely associated with worry about both AIDS and SARS, which suggested that association between social support and AIDS and SARS my persist despite differences in the epidemiology of EIDs. These findings are consistent with studies showing that social interactions are associated with perceptions of the risk for HIV/AIDS (3,12). Several biological and psychological mechanisms have been theorized to mediate the relationship between social support and health (4). Self-efficacy, defined as the degree of confidence persons have in their ability to perform specific tasks, is hypothesized to be one of the primary pathways through which social support operates (4). Therefore, social support may act as a buffer against worry about EIDs.

This study has several limitations. First, covariates were treated as time-fixed, which may have resulted in residual confounding. Second, social support is a multidimensional construct; specific dimensions may differentially influence knowledge of, worry about, and stigmatizing attitudes toward EIDs. Finally, for a cross-sectional analysis, it is difficult to establish temporality between levels of social support and knowledge of, worry about, and attitudes toward AIDS and SARS. Observed associations between social support and responses to EIDs may be explained by a common unmeasured covariate.

Emerging and reemerging infectious diseases will continue to profoundly affect human health. The ability to effectively mitigate disease consequences will depend, in part, on minimizing negative public responses. We showed that social support may be central to public responses to EIDs. Further research investigating the pathways linking social support to responses to EIDs may inform interventions that help guide public health and official responses to these diseases.