Of the potential confounding variables, income and marital status showed the strongest association with social support. For example, 27% of the women in the highest tertile of social support reported incomes greater than $75,000 a year, as compared to 13% of women in the lowest tertile of support. Women in the highest tertile of social support were likely to be married (78%). Of these women reporting higher support, 19% were divorced while out of the women in the lower support tertile, 50% were divorced. Of those in the lowest tertile of support, 10% were black and 5% Hispanic/Latino as compared to 6% and 3%, respectively of the women in the higher tertile of support. Participants in the lowest tertile of support were more likely to have high BMIs—29% of women with BMI higher than 30 reported low support, whereas just 22% of those in the higher tertile of support had BMIs of more than 30. Participants in the higher tertile of support tended to have slightly higher educational attainment and were more likely to have indicated stronger emotional wellbeing. Those in the highest tertile of support scored on average 84 out of 100 points on the emotional wellbeing scale, whereas those in the lower tertile of support scored 72 points (P < 0.0001).
Prediction of outcomes
During the 10.8 years of follow-up in the 92,715 women, there were 2,060 cases of incident CHD (1,200 without prior CVD, and 850 with prior CVD). Total CVD cases were 4,440 (2,730 without prior CVD, and 1,710 with prior CVD). Cases of all-cause mortality were 6,029 (4,030 without prior CVD, and 1,999 with prior CVD).
Cox regression analyses were done separately for women with and without self-reported CVD at baseline (Table 2). Because 1,943 women had missing data on prior CVD, they were excluded from these analyses. Among women without prior CVD, who were the majority of the cohort (n = 73,421), the unadjusted HR for incident CHD per one standard deviation increase in social support was 0.92 (95% confidence interval [CI], 0.87-0.97), indicating an estimated decrease in risk of incident heart disease of 8% for each standard deviation increase in social support. For total CVD, the HR was 0.86 (95% CI, 0.83-0.89), whereas for overall mortality the risk was 0.83 (95% CI, 0.81-0.86). When adjusted for potential confounders, the associations were substantially attenuated. For example, for incident CHD and total CVD, the HRs were 0.99 (95% CI, 0.93-1.06) and 0.96 (95% CI, 0.92-1.00), respectively. For all-cause mortality, the relationship attenuated to 0.95, but remained statistically significant (CI, 0.91-0.98).
Among women with prior CVD, the unadjusted HR for incident CHD per one standard deviation increase in social support was 0.93 (CI, 0.88-1.00), whereas for total CVD the unadjusted HR was 0.91 (95% CI, 0.87, 0.95). The all-cause mortality unadjusted HR was 0.88 (95% CI, 0.85, 0.92). All of these results became nonsignificant when adjusted for confounders—the HR for incident CHD was 1.04 (95% CI, 0.97, 1.12); for total CVD = 1.0 (95% CI, 0.95,1.06); and for all-cause mortality = 1.02 (95% CI, 0.97, 1.07).
In order to explore which of the potential confounders may have accounted for most of the attenuation in social support HRs predicting total CVD, a set of Cox regression analyses was repeated, including for each model a single covariate in the model in addition to social support. Analyses were conducted only in the group that did not report prior CVD at study entry. These data serve as an example of which covariates influenced the results. As shown in Table 3, the greatest attenuation produced by a single covariate occurred in the models with income, marital status, and age, in that order.
In this large diverse cohort of postmenopausal women from across the United States, we found that, in unadjusted analyses, higher levels of perceived social support were associated with a lower risk of incident CHD, total CVD, and all-cause mortality. This conclusion is consistent with other findings in the literature based on other populations.11,23,28,36-38 Furthermore, we found that higher levels of perceived support were related to lower CHD, CVD, and mortality both in women with and without CVD at baseline. However, the associations between perceived support and studied outcomes were not statistically significant after controlling for sociodemographic and lifestyle/behavioral factors, except for mortality in those without prior CVD.
We used a subset of questions from the MOS-SSS, which measures perceived functional support, and evaluated this subset's association with CVD and mortality. Other studies have used measures of functional support,36-38 but this very large cohort study is one of the largest to date to specifically study the relationship of perceived social support questions to CVD. Previous research has broadly showcased social support's association with CVD, but many studies have focused not on perceived social support but instead on structural measures of support, such as social network ties and integration.19,58 Perception of support is hypothesized to influence an individual's neuroendocrine system and buffer stress.49 Therefore, an individual who perceives she has support may be affected by physiological changes, which would protect her from CVD. However, in this longitudinal study of 10.8 years, we did not find an association between higher support and less CVD. A previous study in which adolescent Finns were surveyed on psychosocial factors and evaluated 27 years later demonstrated that a positive social environment as an adolescent contributes to better cardiovascular health.59 We may have captured social support later in the lifetime of a woman than is relevant to development of CVD.
Some studies suggest that there may be a minimum threshold of social support,20,60 but our analyses showed a largely linear association of the scale we used with outcome HRs. This finding in this large cohort contributes to the theory regarding perceived support's influence on CVD and all-cause mortality.
The potential mechanisms by which perceived social support influences CVD and other adverse outcomes is complex. There may be many influences on the outcomes that are difficult to extrapolate, so it is important to control for variables that may influence social support, CVD, or both. Social support can influence behaviors,26 and we controlled for the confounders of smoking, physical activity, alcohol use, and diet. However, the relationships on behaviors and support may be bidirectional, which makes the relationship complex. Although we found that social support was associated with a reduced risk of CVD outcomes and death, the findings were nearly completely eliminated when adding in confounding variables. In addition, while we controlled for a large number of cardiovascular risk factors, we acknowledge that not all potential confounders affecting CVD were controlled for. For example, since the study was designed and completed, inflammatory markers such as C-reactive protein and interleukin-6 are more commonly assessed in studies on CVD. We controlled for family income while other studies did not.36-38 The large number of covariates included in our analyses provided more rigorous control for confounding than in other studies. Other research has demonstrated social support's association with MI,37 coronary morbidity,36 and CHD,38 but none of these studies controlled for income, arguably our strongest confounding variable. Given our ability to control for a full range of confounders, our findings call into question prior results.
Women who had prior CVD were analyzed separately from those without prior CVD. Contrary to earlier reports, we did not find an association between social support and incident CHD or total CVD in women with or without prior CVD, but we did find a small inverse association between perceived support and all-cause mortality in women without prior CVD, who were in the majority. This may suggest that social support's influence on mortality differs between those with CVD versus those who are free of CVD; this issue needs further exploration. In addition, we acknowledge that social support could be confounded with self-care, in that women with higher social support have extra support to get to medical appointments and perform healthy behaviors such as exercise. Messina et al55 reported that women with higher support were more likely to access breast cancer screening.
Our study contributes to the literature by showcasing use of functional support, using perception of support as the measure. Other studies assessed the relationship of social support (social ties) with mortality.23,28 Our findings suggest that the association between perceived social support and CVD in postmenopausal women is modest and that the confounding had a small effect on this association. One study evaluating the association of the psychosocial variables of anger, depression, and perceived social support found that persons with anger proneness and with depression had higher chances of developing peripheral artery disease while followed over 9.7 years.61 However, this same study did not show an association between low social support and peripheral artery disease. We found similar in that low social support did not appear to influence cardiovascular disease, like peripheral artery disease, another vascular disease. Many studies that show the association between low support and CVD indicate a trend for this relationship; however, continued exploration of this topic is of interest since study findings are conflicting.
Early studies on social support's effect on CVD were performed in cohorts that were solely male36,38 while others investigated both sexes but reported different responses to social support. In a large Japanese cohort study,43 (n = 44,152) low perception of social support was strongly associated with stroke mortality in men (HR = 1.59; 95% CI, 1.01-2.51), but the association was nonsignificant in women. Another study showed the opposite37—low social support was associated with increased risk of MI (HR = 2.72; 95% CI, 1.42-5.22) and stroke (HR = 1.80, 95% CI, 1.05-3.10) in women but results were nonsignificant in men. In another large Japanese cohort study,62 researchers studied the psychosocial factor of living in multigenerational households as it relates to CHD and mortality. Although the Japanese study did not measure social support, but social interaction, it found that women living in multigenerational households had two to three times higher risk of CHD than did women who lived only with their spouses. Early on, researchers suspected that there may be sex differences related to the health effects of social support.21,63 However, many of these studies including ours were completed more than 10 years ago. We speculate that when early research was done, men and women had more traditional roles, and men might have had access to more support from women than women did from men. In recent years, the balance has been shifting and men may be giving support equal to that given by women. It is difficult to project whether the findings would be true today as traditional roles of men and women are everchanging. In addition, future studies should look at support levels in same-sex couples.
The WHI-OS cohort was interviewed at baseline 20 years ago, and the outcomes were assessed 10 years ago. Because CVD rates have increased in women in the last few decades,1 a similar study done today may have different results. In addition, since the initial assessment of social support in this cohort, online and social media support outlets such as Facebook have become commonplace, perhaps changing the way that support is gleaned and assessed by individuals. According to the Pew Research Center, 34% of those older than 65 years use social media such as Facebook or Twitter.64 However, recent research shows that, converse to what might be expected, Facebook may decrease perceived support.65 And in a study on a Facebook measure of social support, researchers found that access to Facebook did not decrease depression and did not improve quality of life in study participants.66 Further studies my help determine whether online support is a valid type of social support.
Strengths of our study include the large sample size, the adjudicated outcomes, and the relative racial/ethnic diversity of the study sample. Follow-up was extensive, and attempts to capture missing data were prioritized.51 We controlled for a large number of confounders which can be seen as a strength of our study or a limitation as there could have been a flaw in our theory that the chosen covariates influenced the relationship.
All of these results, including ours, point to the need for further research on social support and its relation to CVD. Our large prospective cohort study in women contributes to the literature by showing that social support's relationship to CVD in women needs further clarification, and that there may be a small, but significant association between social support and mortality in women who do not have prior CVD.
Participation bias could have influenced the outcomes in that volunteers in the WHI could have been more healthy or different than the general population. The WHI researchers made a great effort to recruit a diverse cohort from across the United States, yet women who had the time and freedom to participate might have different characteristics than other women.
Although many potential confounding variables were controlled statistically, some residual confounding is likely. In particular, measures of inflammation, such as C-reactive protein and interleukin-6 were not yet seen as risk factors for CVD when the study was designed.
Social support was captured only at baseline of this study in which participants were followed for up to 10.8 years. We were thus unable to account for support at time of outcome or evaluate any relationship between change in social support, lifestyle or clinical variables, or outcomes.
In addition, there is no criterion standard for measuring social support in relation to health outcomes. This study used a common approach by measuring perceived social support but did not measure social ties, another common measure of support.
Most studies relating social support to CVD have assessed “emotional” or “attachment” support,38-40,45 a type of functional perceived support. We assessed emotional/informational support as well as tangible support, positive social interaction, and affection using a composite perceived support scale. We used nine questions from the MOS-SSS which was developed to assess perceived support in chronically ill persons,54 and as such might not have been the ideal measure for our healthy cohort. At least one study which demonstrated a relationship between social support and CVD assessed support in a work setting,37 which may indicate that support received at work is especially protective against CVD.
We acknowledge that aspects of social well-being such as having children or being religious could overlap with social support. We did not include assessment of those measures in our model.
This is the largest observational study to date in women to evaluate perceived social support's association with CVD and all-cause mortality. The major finding of our current study is that although perceived social support may be associated with fewer cardiovascular outcomes and decreased all-cause mortality, after controlling for potential confounders, these associations either become much smaller in magnitude or disappear.
Our conclusion that perceived social support, a type of functional support, may be mildly inversely associated with mortality is important in that it suggests that this is a trend that needs further clarification and investigation. Future research should focus on refining measurement of psychological risk factors such as social support, determining whether there are sex differences in psychological influences on health, particularly cardiovascular health, and assessing how psychological factors can be modified to positively affect health. In addition, research using, and clarifying between, multiple measures of social support with longitudinal assessment, are needed before drawing stronger conclusions about the causal role of social support in the development or progression of CVD.
1. Benjamin EJ, Virvani SS, Callaway CW, et al. Heart disease and stroke statistics-2018-update: a report from the American Heart Association. Circulation
2. Thompson DR, Ski CF. Psychosocial interventions in cardiovascular disease
—what are they? Eur J Prev Cardiol
3. Linden W, Stossel C, Maurice J. Psychosocial interventions for patients with coronary artery disease: a meta-analysis. Arch Intern Med
4. Hegeman A, Schutter N, Comijs H, et al. Loneliness and cardiovascular disease
and the role of late-life depression. Int J Geriatr Psychiatry
5. Valtorta NK, Kanaan M, Gilbody S, Ronzi S, Hanratty B. Loneliness and social isolation as risk factors for coronary heart disease and stroke: systematic review and meta-analysis of longitudinal observational studies. Heart
6. Chen SJ, Chang CH, Chen KC, Liu CY. Association between depressive disorders and risk of breast cancer recurrence after curative surgery. Medicine (Baltimore)
7. Creed F, Morgan R, Fiddler M, Marshall S, Guthrie E, House A. Depression and anxiety impair health-related quality of life and are associated with increased costs in general medical inpatients. Psychosomatics
8. Jaremka LM, Andridge RR, Fagundes CP, et al. Pain, depression, and fatigue: loneliness as a longitudinal risk factor. Health Psychol
9. Cauley JA, Smagula SF, Hovey KM, et al. Optimism, cynical hostility, falls, and fractures: the Women's Health Initiative observational study (WHI-OS). J Bone Miner Res
10. Stickley A, Koyanagi A, Koposov R, Schwab-Stone M, Ruchkin V. Loneliness and health risk behaviours among Russian and U.S. adolescents: a cross-sectional study. BMC Public Health
11. Berkman LF, Leo-Summers L, Horowitz RI. Emotional support and survival after myocardial infarction: a prospective, population-based study of the elderly. Ann Intern Med
12. Kim ES, Hagan KA, Grodstein F, DeMeo DL, De Vivo I, Kubzansky LD. Optimism and cause-specific mortality: a prospective cohort study. Am J Epidemiol
13. Shipley BA, Weiss A, Der G, Taylor MD, Deary IJ. Neuroticism, extraversion, and mortality in the UK Health and Lifestyle Survey: a 21-year prospective cohort study. Psychosom Med
14. Giltay EJ, Geleijnse JM, Zitman FG, Hoekstra T, Schouten EG. Dispositional optimism and all-cause and cardiovascular mortality in a prospective cohort of elderly Dutch men and women. Arch Gen Psychiatry
15. Smoller JW, Pollack MH, Wassertheil-Smoller S, et al. Panic attacks and risk of incident cardiovascular events among postmenopausal
women in the Women's Health Initiative Observational Study. Arch Gen Psychiatry
16. Arbelaez JJ, Ariyo AA, Crum RM, Fried LP, Ford DE. Depressive symptoms, inflammation, and ischemic stroke in older adults: a prospective analysis in the cardiovascular health study. J Am Geriatr Soc
17. Shah BM, Shah S, Kandula NR, Gadgil MD, Kanaya AM. Psychosocial factors associated with subclinical atherosclerosis in South Asians: the MASALA study. J Immigr Minor Health
18. McCurley JL, Penedo F, Roesch SC, et al. Psychosocial factors in the relationship between socioeconomic status and cardiometabolic risk: the HCHS/SOL sociocultural ancillary study. Ann Behav Med
19. Ramsay S, Ebrahim S, Whincup P, et al. Social engagement and the risk of cardiovascular disease
mortality: results of a prospective population-based study of older men. Ann Epidemiol
20. Cohen S, Underwood LG, Gottlieb BH. Oxford University Press, Social Support
Measurement and Intervention: A Guide for Social Scientists. New York: 2000.
21. Shumaker SA, Hill DR. Gender differences in social support
and physical health. Health Psychol
22. Perkins JM, Subramanian SV, Christakis NA. Social networks and health: a systematic review of sociocentric network studies in low- and middle-income countries. Soc Sci Med
23. Becofsky KM, Shook RP, Sui X, Wilcox S, Lavie CJ, Blair SN. Influence of the source of social support
and size of social network on all-cause mortality
. Mayo Clin Proc
24. Shaya FT, Chirikov VV, Mullins DC, et al. E. Social networks help control hypertension. J Clin Hypertens
25. Loucks EB, Berkman LF, Gruenewald TL, Seeman TE. Relation of social integration to inflammatory marker concentrations in men and women 70-79 years. Am J Cardiol
26. Ford EC, Ahluwalia IB, Galuska DA. Social relationships and cardiovascular disease
risk factors: findings from the third national health and nutrition exam survey. Prev Med
27. Rutledge T, Linke SE, Olson MB, et al. Social networks and incident stroke among women with suspected myocardial ischemia. Psychosom Med
28. Kawachi I, Colditz GA, Ascherio A, et al. A prospective study of social networks in relation to total mortality and cardiovascular disease
in men in the USA. J Epidemiol Community Health
29. Rutledge T, Kenkre TS, Thompson DV, et al. Psychosocial predictors of long-term mortality among women with suspected myocardial ischemia: the NHLBI-sponsored Women's Ischemia Syndrome Evaluation. J Behav Med
30. Cobb S. Presidential Address-1976. Social support
as a moderator of life stress. Psychosom Med
31. Cohen S, Wills TA. Stress, social support
, and the buffering hypothesis. Psychol Bull
32. Kaplan BH, Cassel JC, Gore S. Social support
and health. Med Care SUPP
33. Lakey B, Cassady P. Cognitive processes in perceived social support
. J Pers Soc Psychol
34. Lakey B, Drew J. Pierce GR, Lakey B, Sarason IB, Sarason BR. A social-cognitive perspective on social support
. Sourcebook of Social Support and Personality
. New York: Springer Science and Business Media; 1991. 107–140.
35. Rodriguez C, Burg M, Di Tullio M, et al. Effect of social support
on nocturnal blood pressure dipping. Psychosom Med
36. Rosengren A, Wilhelmsen L, Orth-Gomer K. Coronary disease in relation to social support
and social class in Swedish men. A 15 year follow-up in the study of men born in. Eur Heart J
37. Andre-Petersson L, Engstrom G, Hedblad B, Janzon L, Rosvall M. Social support
at work and the risk of myocardial infarction and stroke in women and men. Soc Sci Med
38. Orth-Gomer K, Rosengren A, Wilhelmsen L. Lack of social support
and incidence of coronary heart disease in middle-aged Swedish men. Psychosom Med
39. Rackow P, Scholz U, Hornung R. Received social support
and exercising: an intervention study to test the enabling hypothesis. Br J Health Psychol
40. Maunder RG, Nolan RP, Park JS, James R, Newton G. Social support
and the consequences of heart failure compared with other cardiac diseases: the contribution of support received within an attachment relationship. Arch Cardiovasc
41. Fontana A, Diegnan T, Villeneuve A, Lepore S. Nonevaluative social support
reduces cardiovascular reactivity in young women during acutely stressful performance situations. J Behav Med
42. Rose G, Kumlin L, Dimberg L, Bengtsson C, Orth-Gomer K, Cai X. Work-related life events, psychological well-being and cardiovascular risk factors in male Swedish automotive workers. Occup Med (Lond)
43. Ikeda A, Iso H, Kawachi I, Yamagishi K, Inoue M, Tsugane S. for the JPHC Study Group. Social support
and stroke and coronary disease: the JPHC study cohorts II. Stroke
44. Kopp M, Skrabski A, Szanto Z, Siegrist J. Psychosocial determinants of premature cardiovascular mortality differences within Hungary. J Epidemiol Community Health
45. Wang HX, Mittleman MM, Orth-Gomer K. Influence of social support
on progression of coronary disease in women. Soc Sci Med
46. Krumholz HM, Butler J, Miller M, et al. Prognostic importance of emotional support for elderly patients hospitalized with heart failure. Circulation
47. Lurie I, Myers V, Goldbourt U, Gerber Y. Perceived social support
following myocardial infarction and long-term development of frailty. Eur J Prev Cardiol
48. Weiss-Faratci N, Lurie I, Neumark Y, et al. Perceived social support
at different times after myocardial infarction and long-term mortality risk: a prospective cohort study. Ann Epidemiol
49. Mezuk B, Diez Roux AV, Seeman T. Evaluating the buffering vs. direct effects hypotheses of emotional social support
on inflammatory markers: the multi-ethnic study of atherosclerosis. Brain Behav Immun
50. The Women's Health Initiative Study Group. Design of the Women's Health Initiative Clinical Trial and Observational Study. Control Clin Trials
51. Curb JD, McTiernan A, Heckbert SR, Kooperberg C, Stanford J, Nevitt M, et al. Outcomes ascertainment and adjudication methods in the women's health initiative. Ann of Epidemiol
52. Langer RD, White E, Lewis CE, Kotchen JM, Henrdrix SL, Trevisan M. The Women's Health Initiative Observational Study: baseline characteristics of participants and reliability of baseline measures. Ann Epidemiol
53. Hay J, Hunt JR, Hubbell FA, et al. The Women's Health Initiative recruitment methods and results. Ann of Epidemiol
54. Sherbourne CD, Stewart AL. The MOS social support
survey. Soc Med
55. Messina CR, Lane DS, Glanz K, et al. Relationship of social support
and social burden to repeated breast cancer screening in the Women's Health Initiative. Health Psychol
56. Kroenke CH, Michael Y, Tindle H, et al. Social networks, social support
and burden in relationships, mortality after breast cancer diagnosis. Breast Cancer Res Treat
57. Patterson R, Kristal A, Tinker L, Carter R, Bolton M, Agurs-Collins T. Measurement characteristics of the Women's Health Initiative food frequency questionnaire. Ann Epidemiol
58. Shumaker S, Czajkowski SM. Springer Science and Social Media, Social Support
and Cardiovascular Disease
. New York: 2013.
59. Pulkki-Raback L, Elovainio M, Hakulinen C, et al. Cumulative effect of psychosocial factors in youth on ideal cardiovascular health in adulthood: the Cardiovascular Risk in Young Finns Study. Circulation
60. Mookadam F, Arthur H. Social support
and its relationship to morbidity and mortality after acute myocardial infarction: systematic overview. Arch Intern Med
61. Wattanakit K, Williams JE, Schreiner PJ, Hirsch AT, Folsom AR. Association of anger proneness, depression and low social support
with peripheral arterial disease: the Atherosclerosis Risk in Communities Study. Vasc Med
62. Ikeda A, Iso H, Kawachi I, et al. Living arrangement and coronary heart disease: the JPHC study. Heart
63. Bartley M, Martikainen P, Shipley M, Marmot M. Gender differences in the relationship of 6 partner's social class to behavioural risk factors and social support
in the Whitehall II study. Soc Sci Med
65. Lee E-J, Cho E. When using Facebook to avoid isolation reduces perceived social support
. Cyberpsychol Behav Soc Netw
66. McCloskey W, Iwanicki S, Lauterbach D, Giammittorio DM, Maxwell K. Are Facebook “friends” helpful? Development of a Facebook-based measure of social support
and examination of relationships among depression, quality of life, and social support
. Cyberpsychol Behav Soc Netw
All-cause mortality; Cardiovascular disease; Postmenopausal; Social support
Supplemental Digital Content
© 2019 by The North American Menopause Society.