Influence of social support on progression of coronary artery disease in women

ARTICLE IN PRESS

Social Science & Medicine 60 (2005) 599–607

Influence of social support on progression of coronary artery disease in women Hui-Xin Wanga, Murray A. Mittlemanb, Kristina Orth-Gomera,* b

a Department of Public Health Sciences, Karolinska Institute, P.O. Box 220, 171 77 Stockholm, Sweden Department of Epidemiology, Cardiovascular Division, Beth Israel Deaconess Medical Center, Harvard Medical School, 1 Autumn Street, Boston, MA 02215, USA

Available online 19 July 2004

Abstract There is strong and consistent evidence across numerous studies that social isolation or lack of social support is an independent risk factor for incident coronary heart disease. However, the impact of social isolation or lack of social support on the progression of coronary atherosclerosis in women has not been well documented. Among 292 women, aged 30–65 years, consecutively hospitalized with acute myocardial infarction or unstable angina between 1991 and 1994 enrolled in the female coronary risk study, 102 women were available to study disease progression over an average of 3.2 years. Three aspects of social support were studied: emotional support, social integration, and interpersonal social relations. Quantitative coronary angiography was performed 3–6 months following index hospitalization and repeated 3 years later. Progression of coronary atherosclerosis was evaluated as the change in mean luminal diameter from first to second measurements of 10 pre-defined coronary segments. Mixed model ANOVA was used to analyze the impact of social support on progression of coronary atherosclerosis. Significantly greater coronary atherosclerosis progression was found among women who lacked emotional support (mean coronary artery luminal diameter narrowing by 0.15 mm), with social isolation (0.14 mm), and lack of interpersonal social relations (0.13 mm), whereas women with high levels of support progressed less. It was 0.05 mm in women with high levels of emotional support, 0.07 mm for socially integrated women, and 0.04 mm in women with adequate interpersonal social relations. These associations were independent of conventional clinical and lifestyle factors such as age, smoking history, body mass index, menopausal status, and diagnosis of index event of acute myocardial infarction. The results of our study suggest that lack of emotional support, social isolation, and lack of interpersonal social relations are important risk factors for accelerated progression of coronary atherosclerosis in middle-aged women. r 2004 Elsevier Ltd. All rights reserved. Keywords: Atherosclerosis; Coronary heart disease; Social support; Quantitative coronary angiography; Women’s health; Sweden

Introduction Social networks and social supports have been related to a wide variety of health outcomes. These range from a beneficial effect on longevity in large populations (Eng, *Corresponding author. Tel.: +46-87286012; fax: +4687286013. E-mail addresses: [email protected] (H.-X. Wang), [email protected] (M.A. Mittleman), [email protected] (K. Orth-Gomer).

Rimm, Fitzmaurice, & Kawachi, 2002; House, Landis, & Umberson, 1988; Welin, Lappas, & Wilhelmsen, 2000) to more recently observed effects of delaying onset of dementia (Fratiglioni, Wang, Ericsson, Maytan, & Winblad, 2000), and to enhancing healthy lifestyles such as the capacity of heavy smokers to quit their habit (Pirie, Rooney, Pechacek, Lando, & Schmid, 1997). The most convincing results are derived from the studies of social networks and social supports on cardiovascular disease. There is strong and consistent evidence across the literature that social isolation or lack

0277-9536/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.socscimed.2004.05.021

ARTICLE IN PRESS 600

H.-X. Wang et al. / Social Science & Medicine 60 (2005) 599–607

of social support is an independent risk factor for coronary heart disease (Bunker et al., 2003; Eriksen, 1994; Rozanski, Blumenthal, & Kaplan, 1999). Social support has been linked to incidence of coronary heart disease (Eng et al., 2002; Orth-Gome´r, Rosengren, & Wilhelmsen, 1993; Rozanski et al., 1999), including nonfatal acute myocardial infarction (AMI) (Bunker et al., 2003; Rozanski et al., 1999), as well as to cardiovascular mortality (Eng et al., 2002; Frasure-Smith et al., 2000; Johnson, Stewart, Hall, Fredlund, & Heorell, 1996; Kawachi et al., 1996; Orth-Gome´r, Unde´n, & Edwards, 1988; Rosengren, Orth-Gomer, & Wilhelmsen, 1998; Ruberman, Weinblatt, & Goldberg, 1984). The beneficial effects of social support on patient prognosis, once coronary disease is present, and on the risk of recurrent CHD events have been persuasively demonstrated in men (Hedblad et al., 1992; Orth-Gome´r et al., 1993; Ruberman et al., 1984). However, women have been rarely included in sufficient numbers to examine their social support in relation to coronary disease (Brezinka & Kittel, 1996; Bunker et al., 2003). Women are known to differ from men in their provision of social support, use of social support, and in their social skills (Shumaker & Hill, 1991). Thus, it is important to study the effects of social support on women’s cardiovascular disease. Moreover, the precise association of social support with the underlying pathology of coronary arteries has not been well studied. Quantitative coronary arteriography (QCA) provides computer-assisted evaluation of repeated coronary angiography visualizing the change overtime of the entire coronary tree. It was developed to study the effects of pharmacological treatment on the progression of coronary atherosclerosis (Jukema et al., 1995; Watts et al., 1992), and is a sensitive indicator for coronary atherosclerosis progression, but few studies have related it to psychosocial factors. In cross-sectional analyses, we have found that lack of social support was an independent risk factor for the severity and extent of coronary atherosclerosis as assessed by coronary angiography in middle-aged women. This association was independent of important medical and lifestyle risk factors such as smoking, diet and sedentary lifestyle (Orth-Gomer 1998). Using the same measures of social support, which had been previously shown to predict incident coronary disease in men (Orth-Gome´r et al., 1993), poor social support was also found to be an independent risk factor for recurrent clinical events in women (Horsten, Mittleman, Wamala, Schenck-Gustafsson, & Orth-Gome´r, 2000). In the present 3-year followup study, we examined the relationship between social support and the progression of coronary disease in middle-aged women. Social support was assessed using two instruments: a modified version of the Interview Schedule for Social Interaction (ISSI) and a condensed version of the Interpersonal

Support Evaluation List (ISEL). Three aspects of social support were studied: emotional support, social integration, and interpersonal social relations. QCA was used to evaluate the progression of coronary atherosclerosis.

Methods Study population The Stockholm female coronary risk study design, recruitment, and study population have been previously described (Al-Khalili, Wamala, Orth-Gomer, & Schenck-Gustafsson, 2000; Orth-Gome´r et al., 1997). In brief, between February 1991 and February 1994, 292 women, aged 30–65, residing in the greater Stockholm area and consecutively admitted to coronary care units with AMI or unstable angina (UAP) were enrolled in the study. Three to 6 months after hospitalization, we assessed a detailed medical history, lifestyle and social demographic data, anthropometric measures, and a full lipid and routine laboratory profile. Of the 292 women enrolled in the Female coronary risk study, 131 underwent baseline QCA within 3 months of their initial evaluations (Orth-Gome´r et al., 1998). Among them, 12 refused the second evaluation and 119 women underwent the second QCA evaluation an average of 3.25 years (range 2–5 years) later. Of these, 10 had poor quality angiograms either at baseline (n ¼ 2) or at followup (n ¼ 8) and three women’s follow-up angiograms were not available for quantitative evaluation. Among the remaining patients, one had missing information on social support, resulting in 102 patients with valid and comparable QCA measurement and social support information. Assessment of social support Social support was assessed using two instruments: a modified version of the ISSI (Henderson, Duncan-Jones, Byrne, & Scott, 1980; Unde´n & Orth-Gome´r, 1989) and a condensed version of the ISEL (Cohen, Mermelstein, Kamarck, & Hoberman, 1985). A paper and pencil version of the ISSI measure has been developed for use in population studies, and examined for reliability and validity in our research laboratory (Unde´n & Orth-Gome´r, 1989). The instrument yields two scales: one describes availability of deep emotional relationships or attachment—emotional support, and the other the availability of more peripheral contacts that provides social integration. The attachment scale describes the availability of close emotional support, mainly from family and close friends. It consists of 6 items with a total score ranging from 0 to 6. One example item from the scale is: ‘‘Do you have someone that can share your inmost thoughts with and

ARTICLE IN PRESS H.-X. Wang et al. / Social Science & Medicine 60 (2005) 599–607

601

confide in?’’ The social integration scale describes the structural measurement, incorporating both the quantitative and qualitative characteristics of the social network (Orth-Gome´r, 1979). It is comprised of six items with a total score ranging from 6 to 36. One example item from the scale is: ‘‘How many people, of those you know, do you meet or talk with during an ordinary week (do not count persons that you meet by chance and that you hardly will meet again, e.g. customers in a store)?’’ The condensed version of ISEL consists of 13 items that address interpersonal social relations: belongingness (5 items), appraisal support (5 items), and tangible support (3 items). Examples of the items are: ‘‘There are several different people with whom I enjoy spending time’’ (belonging). ‘‘When I need suggestions for how to deal with a personal problem I know there is someone that I can turn to’’ (appraisal support). ‘‘If I have to be away for a couple of weeks, there is someone who can take care of my home’’ (tangible support). Items were scored from 1 to 4. A high score indicates high-level social support. The total scale is comprised of 13 items with scores ranging from 13 to 52.

proximal right coronary artery (RCA), (9) mid RCA, and (10) distal RCA. Mean segment diameter was calculated as the mean of all diameters measured along a given segment. We calculated the change over time by subtracting the first from the second measurement. The procedure was blinded to patient’s identity and the order of examination. Special procedures were undertaken to replicate the biologic and imaging conditions of the original angiogram during the followup. The catheterization laboratory was calibrated initially and then twice yearly. The baseline angiogram was reviewed, and a copy of the baseline arteriography procedure log was used as the template for the followup angiogram. Catheters of the same size and type were used in the followup and baseline assessments. The filming sequence used in the baseline angiogram was exactly replicated. For each view, the image intensifiers were restored to their positions from the baseline angiogram. To ensure that the coronary arteries were seen in the same place on the X-ray monitor, patients were positioned as in the baseline angiogram. Intra-coronary nitrates were given in identical doses in the first and second examinations.

Quantitative coronary angiography

Covariates

Selective coronary arteriography was performed at the Department of Thoracic Radiology at Karolinska Hospital using the Judkins’ technique and a standard clinical angiographic procedure. Before and after intracoronary injection of nitroglycerine, identical angiographic projections were obtained. We followed a standard clinical angiograph procedure using 7 French non-nylon catheters. After engaging the coronary segment under study with the injection catheter, the angiographic view was optimized with short test injections. During a breath hold, filming was started, before contrast injection, to show the catheter. Next, we injected dye to render the segments of interest opaque for at least 3 cardiac cycles. Imaging conditions (angiographic view angles, catheter size, and field size) were recorded in an arteriography procedure log. At the end of the procedure, the distal 10 cm of each catheter was saved and labeled for later identification and measurement. All angiograms were recorded as cine films with a frame rate 25 frame/s. The Angiographic Image Processing Laboratory of the Division of Cardiology, University of Texas performed computer-assisted quantitative evaluations of the angiographic films (14). For each angiogram, absolute luminal diameter (in mm) was measured in 10 pre-defined coronary segments: (1) left main coronary artery, (2) proximal left anterior descending artery (LAD), (3) mid LAD, (4) first diagonal branch of the LAD, (5) proximal left circumflex artery (LCX), (6) mid LCX, (7) first obtuse marginal branch of the LCX, (8)

Age at examination was obtained from the census register. Educational level was classified into mandatory and high school or college/university. Smoking status was categorized as ever versus never (never or former smokers). Alcohol consumption was reported as the habitual weekly intake of beer, wine, strong spirits and calculated as average daily consumption of absolute alcohol in grams. Physical exercise was assessed according to the World Health Organization questionnaire and categorized as sedentary lifestyle and active lifestyle. Body mass index (BMI) was calculated as weight (kg)/ height (m2). Menopausal status was classified as premenopausal, postmenopausal with- and without- hormone-replacement therapy. Family history of CHD was defined as having at least one close family member diagnosed with AMI. The clinical diagnosis at the index event was categorized into AMI or UAP. Severity of angina pectoris symptoms was graded as: I—no angina, II— mild angina, III—moderately severe angina, and IV— very severe angina, by the Canadian Cardiovascular Society Scale for Angina Pectoris (Kong, 1992). Patients were categorized as severe angina (III+IV) and mild or no angina (I+II). A detailed medical history including hyperlipidemia, hypertension, diabetes, and left ventricular dysfunction, as well as medication was obtained during the clinical examination and verified by scrutiny of hospital charts. Medications were classified as: aspirin, ACE inhibitors, beta-blockers, calcium, antagonists, insulin and lipid-lowering medication.

ARTICLE IN PRESS 602

H.-X. Wang et al. / Social Science & Medicine 60 (2005) 599–607

All blood samples were drawn in the morning 6–10 am after an overnight fast. Plasma samples were frozen and sent in batches to the analyzing laboratory (CALAB). Total cholesterol was determined by enzymatic methods. High-density lipoproteins (HDL) determination was based on the isolation of LDL and VLDL from serum by precipitation. The cholesterol content of the supernatant was measured enzymatically using an automated Multichannel analyzer (Riepponen, Marniemi, & Rautaoja, 1987). Serum fibrinogen (g/l) was determined using a polymerization rate method (Vermylen, de Vreker, & Verstraete, 1963). Statistical analyses Baseline characteristics were compared across levels of social support for the three measures. The Mann– Whitney test was used for continuous variables and Fisher’s exact test for categorical variables. We conducted mixed model analysis of variance to test the association between social support and changes in mean luminal diameter. In all analyses, we used the difference in the mean luminal diameters between the baseline and followup measurements as the primary measure of progression- or regression- of coronary atherosclerosis. Segments were treated as experimental units, clustered within individuals, in order to control for correlations of segments within subject. The scale scores on emotional support were skewed, 68% of women had the maximum score of 6. Thus, it was dichotomized as low (lower and middle tertiles, scoreo6) and high (upper tertile, score=6) levels of social attachment. The scale scores for social integration were normally distributed, and a comparison was made among women with low scores (lower tertile, scoreo17), with intermediate scores (middle tertile, 16oscoreo23), and with high scores (upper tertile, score>22). In the analysis for social integration, we found that women in the middle and in the upper tertile had similar magnitude of progression; we regrouped them into women with low (lower tertile, scoreo17) and women with high (middle and upper tertiles, score>16) social integration. Similarly, according to the scores on interpersonal social relations, we regrouped the women into two groups, as low (lower and middle tertiles, scoreo47) and high (upper tertile, score> 46) interpersonal social relations. For each of the three social support measures (emotional support, social integration and interpersonal social relations), we conducted crude, age adjusted, and multivariate-adjusted analyses controlling for potential confounders including age, BMI, menopausal status, smoking history, and diagnosis of index event of AMI. In a second set of models, we further adjusted for educational level, marital status, sedentary lifestyle, alcohol consumption, fibrinogen, total/HDL cholesterol

ratio, angina severity, use of medications, as well as medical history. These factors were introduced into the model one at a time, respectively.

Results Women who underwent arteriography (n ¼ 131) did not differ from the remaining women (n ¼ 161) on a number of clinical and psychosocial variables including standard risk factors and severity of disease symptoms. As compared with the 102 women who had valid measurements of QCA, the 29 women who had either no valid measurement of QCA or information of social support had similar distribution of the variables studied. At baseline, women patients with low levels of social integration were more likely to have a diagnosis of AMI, hypertension and a sedentary lifestyle than women with high levels of social integration. Women with low levels of emotional support were also more likely to have a diagnosis of AMI than those with high levels. Women who had higher levels of interpersonal social relations were more likely to have severe anginal symptoms (Table 1). Among 102 women patients who had information on social support, there were 890 repeatedly examined coronary segments available for analyses of emotional support and 900 segments available for analyses of social integration and interpersonal social relations in relation to the progression. At baseline, women with high levels of support had greater mean luminal diameters across all three measures than those with lower levels (Table 2). During the 3-year followup period, there was greater coronary artery luminal narrowing in women with low levels of social support than among women with high levels across all three measures. Adjusting for age and other standard risk factors including history of smoking, BMI, menopausal status, and diagnosis of index event at baseline, the progression over the 3-year period remained virtually the same for all three measurements of social support (Table 2). In another set of models, we considered other potential confounders of the social support—atherosclerosis progression relationship in addition to those in our basic multivariate models. These variables were added to our basic model one at a time and included marital status, educational level, sedentary lifestyle, alcohol consumption, fibrinogen, total/HDL cholesterol ratio, as well as a medical history of hyperlipidemia, hypertension, diabetes, family history of CHD and left ventricular dysfunction, angina severity, and use of medications. No substantial variations were observed when controlling for these clinical and other confounders. For example, after controlling for alcohol consumption, women with low level of emotional support showed a progression of 0.14 mm in coronary artery

ARTICLE IN PRESS H.-X. Wang et al. / Social Science & Medicine 60 (2005) 599–607

603

Table 1 Clinical characteristics of 102 women enrolled in the female coronary risk study who underwent serial quantitative coronary angiography measured by segments according to emotional support, social integration, and interpersonal social relations Emotional supporta

Social integration

Interpersonal social relations

Low n ¼ 32

High n ¼ 69

Low n ¼ 33

High n ¼ 69

Low n ¼ 66

High n ¼ 36

Age (years)

55.777.9

55.677.2

56.376.9

55.477.6

55.477.6

56.277.0

Diagnosis at index event of AMI UAP History of severe anginal symptoms (III+IV) History of left ventricular dysfunction History of hyperlipidemia History of hypertension History of diabetes mellitus Family history of CHD

62.5 37.5 51.6 18.8 48.4 54.8 18.8 56.3

33.3 66.7 64.1 13.9 33.3 43.1 11.6 52.2

60.6 39.4 50.0 15.6 48.4 46.9 12.2 57.6

34.8 65.2 63.6 15.2 32.8 46.2 14.5 50.7

48.5 51.5 48.3 12.9 35.5 46.8 13.6 51.5

33.3 66.7 77.8 19.4 42.4 45.7 13.9 55.6

Menopausal status Premenopausal Postmenopausal with HRT Postmenopausal without HRT Cigarette smoking (ever smoked) Alcohol intake (g/day) Sedentary lifestyle Body mass index (kg/m2) Serum levels of total/HDL cholesterol Serum levels of fibrinogen (g/l)

21.9 3.1 75.0 75.0 5.579.3 25.0 27.074.1 4.671.4 3.770.7

23.2 13.0 63.8 75.4 5.377.1 21.7 27.874.5 4.971.7 3.670.9

21.2 3.0 75.8 72.7 4.678.0 39.4 27.173.8 4.971.6 3.970.9

23.2 13.0 63.8 76.8 5.777.7 15.9 27.774.6 4.571.6 3.570.8

22.7 6.1 71.2 74.2 5.778.2 27.3 27.174.1 4.871.5 3.670.8

22.2 16.7 61.1 77.8 4.577.0 16.7 28.274.7 4.471.6 3.671.0

Medication use Aspirin ACE inhibitors Beta blockers Calcium antagonists Insulin Lipid-lowering drugs

75.0 12.5 78.1 25.0 16.3 16.1

73.1 10.6 73.1 34.9 6.1 10.6

71.9 12.5 75.0 32.5 9.7 6.5

75.0 10.5 75.0 34.3 9.0 14.9

75.0 9.5 79.7 31.3 8.1 9.7

72.2 13.9 66.7 31.4 11.1 16.7

 po0:01: a Information on emotional support missing for one woman. Figures represent percent (%) for categorical variables and mean 7 standard deviation for continuous variables. P values from Fisher’s exact tests for categorical variables and Mann–Whitney tests for continuous variables. AMI=acute myocardial infarction; UAP=unstable angina pectoris; Severe anginal symptoms=severe and very severe angina; CHD=coronary heart disease; HRT=hormone-replacement therapy; HDL=high-density lipoproteins

luminal narrowing whereas women with high levels of emotional support progressed 0.06 mm (p ¼ 0:02). Similarly, women with low levels of social integration and interpersonal social relationships showed a progression of 0.16 and 0.12 mm in coronary artery luminal narrowing whereas women with high levels of social integration and interpersonal social relations progressed 0.06 mm (po0:007) and 0.05 mm (po0:03), respectively.

Discussion In this 3-year longitudinal study of 102 middle-aged women with coronary heart disease, low levels of

emotional support and social integration, as well as poor interpersonal social relations were associated with significantly increased progression of coronary atherosclerosis as measured by QCA. These associations were independent of conventional risk factors including age, BMI, menopausal status, smoking history, and CHD diagnosis. These associations of low social support with atherosclerosis progression were not explained by educational level, marital status, sedentary lifestyle, alcohol consumption, fibrinogen, total/HDL cholesterol ratio, systolic blood pressure, left ventricular dysfunction, angina severity, or family history of CHD, diabetes, and medications for these ailments.

ARTICLE IN PRESS 604

H.-X. Wang et al. / Social Science & Medicine 60 (2005) 599–607

Table 2 Change of coronary artery luminal diameter (in mm) and 95% confidence intervals, over a 3-year period, among 102 women in the Female coronary risk study, according to level of emotional support, social integration, and interpersonal social relations

Emotional support Low High Social integration Low High Interpersonal social relations Low High a

No. of segments assessed

Mean luminal diameter at baseline

287 603

2.9570.59 3.1970.60

291 609

2.9670.52 3.2070.63

586 314

3.0670.59 3.2270.62

p

Age adjusted mean luminal diameter changes

0.07

p

Multivariate adjusted mean luminal diameter changesa

0.0003 0.16 (0.11, 0.22) 0.05 (0.01, 0.08)

0.07

0.003 0.15 (0.1, 0.22) 0.05 (0.01, 0.1)

0.007 0.15 (0.1, 0.2) 0.06 (0.03, 0.1)

0.20

0.04 0.14 (0.08, 0.2) 0.07 (0.02, 0.12)

0.001 0.13 (0.09, 0.17) 0.03 (0.02, 0.08)

p

0.003 0.13 (0.08, 0.18) 0.04 (0.02, 0.1)

Multivariate models adjusted for age, smoking history, body mass index, menopausal status, and diagnosis of index event.

To assess social support, we used two standardized measures previously found to predict coronary disease in men and heart rate variability in healthy women (Horsten, Ericsson, Perski, Wamala, & Orth-Gome´r, 1999; Orth-Gome´r et al., 1993). We studied the subscale of ISSI, in terms of emotional support and social integration, separately. Emotional support focuses more on close emotional support, and social integration covers both the quantitative and qualitative aspects of social networks. The similar findings from emotional support and social integration suggest that both emotional support and social integration are equally important prognostic factors for the progression of coronary heart disease. ISEL concentrates on aspects of interpersonal social relations including belongingness, appraisal support, and tangible support. Similar effect of interpersonal social relations on disease progression confirmed the findings from emotional support and social integration. Comparison with previous studies To our knowledge, only one prior study has investigated social support in relation to progression of coronary atherosclerosis. In that study, 162 patients, both men and women aged 75 years and younger with CAD documented by standardized angiography at baseline had a second angiogram after 2 years. Emotional support was assessed by the self-report questionnaires. The extent of coronary atherosclerosis at the beginning was defined as the number of stenosis with X50% luminal narrowing, and the progression of the disease was defined as a score of +1, +2 or +3 when comparing the followup with the baseline angiogram. Compared with patients with higher levels of emotional support, those with low emotional support were eight times more likely to have disease progression (Angerer et al., 2000).

Our findings are in line with previous studies of social support and cardiovascular disease incidence and prognosis (Angerer et al., 2000; Berkman, Leo-Summers, & Horwitz, 1992; Craig, Lynch, & Quartner, 2000; Eng et al., 2002; Hemingway & Marmot, 1999; Krumholz et al., 1998; Orth-Gome´r et al., 1993). In previous studies that used the same measurements of social support (ISSI), both poor emotional support and social integration were found to be independent predictors of 5-year subsequent acute CHD events (Hedblad et al., 1992) and 6-year incidence of AMI or CHD death (Orth-Gome´r et al., 1993) in middle-aged Swedish men. Lack of social support was also found to be associated with a doubling of the risk for coronary and all cause mortality in middle-aged men and women in Gothenburg (Welin et al., 2000). Studies using other measurements of social support have also shown consistent results. In a study comparing social network function and structure, Seeman and Syme reported that the feeling of being loved, emotional support, and instrument support were important indicators for coronary atherosclerosis (Seeman, Kaplan, Knudsen, Cohen, & Guralnik, 1987). In a Multicenter study of patients aged 25–75 years from community and academic hospitals in the United States and Canada, living alone was found to be an independent prognostic risk factor for a subsequent major recurrent cardiac event during an average of 2.1 years, in patients aged 25–75 years (Case, Moss, & Case, 1992). In a 6-year followup study of patients undergoing cardiac catheterization, patients without a confidant had twice the risk for an acute event as compared to patients who had a confidant (Williams et al., 1992). In a study of 292 patients aged 65 and older, who were hospitalized with clinical heart failure, Krumholz et al. (1998) reported that lack of emotional support was a strong predictor of fatal and non-fatal cardiovascular events in subsequent 1 year in women but not in men (Krumholz et al., 1998). In a 15-year

ARTICLE IN PRESS H.-X. Wang et al. / Social Science & Medicine 60 (2005) 599–607

followup study of 2603 Americans, Vogt and colleagues found that poor social network scope, poor contact frequency, and poor network size were related to increased risks of CHD incidence (Vogt, Mullooly, Ernst, Pope, & Hollis, 1992). In a study of 196 men and women, 65 years or older, who were hospitalized for acute myocardial infarction, lack of emotional support was found to be independently associated with 6-month mortality (Berkman et al., 1992). Although results of individual intervention studies have been mixed, 2 meta-analyses of psychosocial interventions following MI have reported a reduction in cardiac morbidity and all-cause mortality (Dusseldorp, van Elderen, Maes, Meulman, & Kraaij, 1999; Linden, Stossel, & Maurice, 1996). In contrast, in a recent multi-centered randomized clinical trail from the ENRICHD study, involving in 2481 MI patients, psychosocial intervention did not have a benefit on mortality and recurrent infarction. The negative findings may be caused by duration and timing of the intervention (Berkman et al., 2003). The observation that women with high levels of social support were more likely to have severe anginal symptoms at baseline may be due to various causes. On one hand, women with frequent symptoms may get more attention from family and close friends, and therefore received more support. On the other hand, women with severe anginal symptoms may be more actively seeking help concerning the disease and consulting more people from either the existing social network or newly established relations and thereby getting more support. Alternatively, in women, network strain and the burden of support providing and care giving may override the beneficial effect of social support (Shumaker & Hill, 1991). Strengths and limitations A strength of the present study is that we used two well-characterized standard instruments, ISSI and ISEL, to assess social support, simultaneously. For the outcome assessment, we used QCA, a reliable method, to evaluate the progression of coronary artherosclerosis. A limitation of this study is that a relatively high proportion (22%) of women patients did not have valid/comparable measurements for QCA. It is unlikely that selection bias would have led to our results because many of the baseline characteristics did not vary between the participants and non-participants. Second, although we have controlled many confounders in our multivariate analyses, as in any observational study, uncontrolled and residual confounding cannot be ruled out. Third, our study population consisted of only middle-aged women hospitalized for AMI or UAP. Thus, our results could not be generalized to the general population.

605

Possible mechanisms Among the three proposed mechanisms in the literature linking social support with coronary heart disease, our findings support the direct pathophysiological mechanism involving psycho-neuro-endocrine pathway. It has been suggested that lack of social support is etiologically related to coronary artery lesion development through sympathetic-adrenomedullary influences on platelet function, heart rate, and blood pressure, and pituitary–adrenal cortical factors involved in smooth muscle cell proliferation during progression of the lesion after injury has taken place. Findings from both human and animal studies have demonstrated an association between social isolation and hypercortisolemia, and between degree of social support and resting heart rates (Knox & Uvnas-Moberg, 1998; Sapolsky, Alberts, & Altmann, 1997; Unde´n, Orth-Gomer, & Elofsson, 1991). A buffering mechanism was demonstrated by epidemiological studies showing that people with better availability of social support had moderate cardiovascular reactivity to stressful situations (Barefoot et al., 2000; Gerin, Milner, Chawla, & Pickering, 1995; Uchino & Garvey, 1997) and that social isolation may worsen immune function (Glaser, Kiecolt-Glaser, Malarkey, & Sheridan, 1998; Kiecolt-Glaser et al., 1987). Although social support has often been found to enhance health-promoting behaviors and poor social support may lead to adverse lifestyle, such as poor diet, smoking, and sedentary lifestyle (Cohen, 1988), behavioral pathways are not supported by the present study. We found little difference in lifestyle factors such as smoking, alcohol consumption and BMI across different levels of social support. Only sedentary lifestyle was related to levels of social support. Higher social support might help subjects to better access medical care, especially for those with more severe disease, therefore hindering the disease progress. This is less likely to be the case because access to medical care is equally available and free for all residents in Sweden. Moreover, at baseline, women who had higher levels of interpersonal social relations were more likely to have severe angina symptoms, further supported that there is no differential behavior in access to health-care system among those with high or low levels of social support. Conclusions Our results suggest that lack of emotional support, social isolation, and lack of interpersonal social relations may be important risk factors for accelerated progression of coronary heart disease among middle-aged women. These data suggest that novel interventions aimed at enhancing social integration and support may improve the prognosis of women with established coronary artery disease.

ARTICLE IN PRESS 606

H.-X. Wang et al. / Social Science & Medicine 60 (2005) 599–607

Acknowledgements This work was supported by Grant HL 45785 from the US National Institutes of Health, the Swedish Bank Tercentenary Fund, the Swedish Medical Research Council, and the Swedish Council for Work Life Research and the Swedish Labour Market Insurance Company. The authors are indebted to May Blom, RN and Birgitta Lindvall, RN who coordinated the QCA followup of women from the FemCorRisk study.

References Al-Khalili, F., Wamala, S., Orth-Gomer, K., & SchenckGustafsson, K. (2000). Prognostic value of exercise testing in women after acute coronary syndromes (The Stockholm female coronary risk study). American Journal of Cardiology, 86(2), 211–213. Angerer, P., Siebert, U., Kothny, W., Muhlbauer, D., Mudra, H., & von Schacky, C. (2000). Impact of social support, cynical hostility and anger expression on progression of coronary atherosclerosis. Journal of the American College of Cardiology, 36(6), 1781–1788. Barefoot, J., Brummett, B., Helms, M., Mark, D., Siegler, I., & Williams, R. (2000). Depressive symptoms and survival of patients with coronary artery disease. Psychosomatic Medicine, 62(6), 790–795. Berkman, L., Blumenthal, J., Burg, M., Carney, R., Catellier, D., Cowan, M., Czajkowski, S., DeBusk, R., Hosking, J., Jaffe, A., Kaufmann, P., Mitchell, P., Norman, J., Powell, L., Raczynski, J., & Schneiderman, N. (2003). Effects of treating depression and low perceived social support on clinical events after myocardial infarction: The enhancing recovery in coronary heart disease patients (ENRICHD) randomized trial. Journal of the American Medical Association, 289(23), 3106–3116. Berkman, L., Leo-Summers, L., & Horwitz, R. (1992). Emotional support and survival after myocardial infarction. A prospective, population-based study of the elderly. Annals of Internal Medicine, 117(12), 1003–1009. Brezinka, V., & Kittel, F. (1996). Psychosocial factors of coronary heart disease in women: A review. Social Science & Medicine, 42(10), 1351–1365. Bunker, S., Colquhoun, D., Esler, M., Hickie, I., Hunt, D., Jelinek, V., Oldenburg, B., Peach, H., Ruth, D., Tennant, C., & Tonkin, A. (2003). ‘‘Stress’’ and coronary heart disease: Psychosocial risk factors. Medical Journal of Australia, 178(6), 272–276. Case, B. R., Moss, A. J., & Case, N. (1992). Living alone after myocardial infarction. Journal of the American Medical Association, 267, 520–524. Cohen, S. (1988). Psychosocial models of the role of social support in the etiology of physical disease. Health Psychology, 7(3), 269–297. Cohen, S., Mermelstein, R., Kamarck, T., & Hoberman, H. (1985). Measuring the function components of social support. Dordrecht: Martinus Nijholt. Craig, F., Lynch, J., & Quartner, J. (2000). The perception of available social support is related to reduced cardiovascular

reactivity in Phase II cardiac rehabilitation patients. Integrative Physiological and Behavioral Science, 35(4), 272–283. Dusseldorp, E., van Elderen, T., Maes, S., Meulman, J., & Kraaij, V. (1999). A meta-analysis of psychoeducational programs for coronary heart disease patients. Health Psychology, 18(5), 506–519. Eng, P., Rimm, E., Fitzmaurice, G., & Kawachi, I. (2002). Social ties and change in social ties in relation to subsequent total and cause-specific mortality and coronary heart disease incidence in men. American Journal of Epidemiology, 155(8), 700–709. Eriksen, W. (1994). The role of social support in the pathogenesis of coronary heart disease. A literature review. Family Practice, 11(2), 201–209. Frasure-Smith, N., Lesperance, F., Gravel, G., Masson, A., Juneau, M., Talajic, M., & Bourassa, M. (2000). Social support, depression, and mortality during the first year after myocardial infarction. Circulation, 101(16), 1919–1924. Fratiglioni, L., Wang, H., Ericsson, K., Maytan, M., & Winblad, B. (2000). Influence of social network on occurrence of dementia: A community-based longitudinal study. Lancet, 355(9212), 1315–1319. Gerin, W., Milner, D., Chawla, S., & Pickering, T. (1995). Social support as a moderator of cardiovascular reactivity in women: A test of the direct effects and buffering hypotheses. Psychosomatic Medicine, 57(1), 16–22. Glaser, R., Kiecolt-Glaser, J., Malarkey, W., & Sheridan, J. (1998). The influence of psychological stress on the immune response to vaccines. Annals of the New York Academy of Sciences, 840, 649–655. Hedblad, B., Ostergren, P., Hanson, B., Janzon, L., Johansson, B., & Juul-Moller, S. (1992). Influence of social support on cardiac event rate in men with ischaemic type ST segment depression during ambulatory 24-h long-term ECG recording. The prospective population study Men born in 1914, Malmo, Sweden. European Heart Journal, 13(4), 433–439. Hemingway, H., & Marmot, M. (1999). Evidence based cardiology: Psychosocial factors in the aetiology and prognosis of coronary heart disease. Systematic review of prospective cohort studies. British Medical Journal, 318(7196), 1460–1467. Henderson, S., Duncan-Jones, P., Byrne, D., & Scott, R. (1980). Measuring social relationships. The Interview Schedule for Social Interaction. Psychological Medicine, 10(4), 723–734. Horsten, M., Ericsson, M., Perski, A., Wamala, S. P., & OrthGome´r, K. (1999). Psychosocial factors and heart rate variability in healthy women. Psychosomatic Medicine, 61(1), 49–52. Horsten, M., Mittleman, M. A., Wamala, S. P., SchenckGustafsson, K., & Orth-Gome´r, K. (2000). Depressive symptoms and lack of social integration in relation to prognosis of CHD in middle-aged women: The Stockholm female coronary risk study. European Heart Journal, 21(13), 1072–1080. House, J. S., Landis, K. R., & Umberson, D. (1988). Social relationships and Health. Science, 241, 540–545. Johnson, J., Stewart, W., Hall, E., Fredlund, P., & Heorell, T. (1996). Long-term psychosocial work environment and cardiovascular mortality among Swedish men. American Journal of Public Health, 86(3), 301–303.

ARTICLE IN PRESS H.-X. Wang et al. / Social Science & Medicine 60 (2005) 599–607 Jukema, J., Bruschke, A., van Boven, A., Reiber, J., Bal, E., Zwinderman, A., Jansen, H., Boerma, G., van Rappard, F., Lie, K., & Al, E. (1995). Effects of lipid lowering by pravastatin on progression and regression of coronary artery disease in symptomatic men with normal to moderately elevated serum cholesterol levels. The Regression Growth Evaluation Statin Study (REGRESS). Circulation, 91(10), 2528–2540. Kawachi, I., Colditz, G., Ascherio, A., Rimm, E., Giovannucci, E., Stampfer, M., & Willett, W. (1996). A prospective study of social networks in relation to total mortality and cardiovascular disease in men in the USA. Journal of Epidemiology and Community Health, 50(3), 245–251. Kiecolt-Glaser, J., Fisher, L., Ogrocki, P., Stout, J., Speicher, C., & Glaser, R. (1987). Marital quality, marital disruption, and immune function. Psychosomatic Medicine, 49(1), 13–34. Knox, S., & Uvnas-Moberg, K. (1998). Social isolation and cardiovascular disease: An atherosclerotic pathway? Psychoneuroendocrinology, 23(8), 877–890. Krumholz, H., Butler, J., Miller, J., Vaccarino, V., Williams, C., Mendes de Leon, C., Seeman, T., Kasl, S., & Berkman, L. (1998). Prognostic importance of emotional support for elderly patients hospitalized with heart failure. Circulation, 97(10), 958–964. Linden, W., Stossel, C., & Maurice, J. (1996). Psychosocial interventions for patients with coronary artery disease: A meta-analysis. Archives of Internal Medicine, 156(7), 745–752. Orth-Gome´r, K. (1979). Ischemic heart disease and psychological stress in Stockholm and New York. Journal of Psychosomatic Research, 23, 165–173. Orth-Gome´r, K., Horsten, M., Schenck-Gustafsson, K., Kirkeeide, R., Svane, B., Wamala, S. P., Mittleman, M. A., & Ryde´n, L. (1998). Social relations and extent and severity of coronary artery disease. European Heart Journal, 19, 1648–1656. Orth-Gome´r, K., Mittleman, M., Schenck-Gustafsson, K., Wamala, S., Eriksson, M., Belkic, K., Kirkeeide, R., Svane, B., & Ryde´n, L. (1997). Lipoprotein (a) as a determinant of coronary heart disease in young women. Circulation, 95, 329–334. Orth-Gome´r, K., Rosengren, A., & Wilhelmsen, L. (1993). Lack of social support and incidence of coronary heart disease in middle-aged Swedish men. Psychosomatic Medicine, 55, 37–43. Orth-Gome´r, K., Unde´n, A.-L., & Edwards, M.-E. (1988). Social isolation and mortality in ischemic heart disease. A 10-year followup study of 150 middle-aged men. Acta Medica Scandinavica, 224, 115–205. Pirie, P., Rooney, B., Pechacek, T., Lando, H., & Schmid, L. (1997). Incorporating social support into a community-wide smoking-cessation contest. Addict Behavior, 22(1), 131–137. Riepponen, P., Marniemi, J., & Rautaoja, T. (1987). Immunoturbidimetric determination of Apolipoprotein A-1 and B in serum. Scandinavian Journal of Clinical & Laboratory Investigation, 47, 739.

607

Rosengren, A., Orth-Gomer, K., & Wilhelmsen, L. (1998). Socioeconomic differences in health indices, social networks and mortality among Swedish men. A study of men born in 1933. Scandinavian Journal of the Society of Medicine, 26(4), 272–280. Rozanski, A., Blumenthal, J., & Kaplan, J. (1999). Impact of psychological factors on the pathogenesis of cardiovascular disease and implications for therapy. Circulation, 99(16), 2192–2217. Ruberman, W., Weinblatt, E., & Goldberg, J. D. (1984). Psychosocial influences on mortality after myocardial infarction. New England Journal of Medicine, 311, 552–559. Sapolsky, R., Alberts, S., & Altmann, J. (1997). Hypercortisolism associated with social subordinance or social isolation among wild baboons. Archives of General Psychiatry, 54(12), 1137–1143. Seeman, T., Kaplan, G., Knudsen, L., Cohen, R., & Guralnik, J. (1987). Social network ties and mortality among the elderly in the Alameda county study. American Journal of Epidemiology, 126(4), 714–723. Shumaker, S. A., & Hill, D. R. (1991). Gender differences in social support and physical health. Health Psychology, 10, 102–111. Uchino, B., & Garvey, T. (1997). The availability of social support reduces cardiovascular reactivity to acute psychological stress. Journal of Behavioral Medicine, 20(1), 15–27. Unde´n, A., Orth-Gomer, K., & Elofsson, S. (1991). Cardiovascular effects of social support in the work place: Twentyfour-hour ecg monitoring of men and women. Psychosomatic Medicine, 53(1), 50–60. Unde´n, A.-L., & Orth-Gome´r, K. (1989). Development of a sicial support instrument for use in population surveys. Social Sciences & Medicine, 29(12), 1387–1392. Vermylen, C., de Vreker, R. A., & Verstraete, M. (1963). A rapid enzymatic method for assay fibrinogen: Fibrin polymerization time (FPT test). Clinica Chimica Acta, 8, 418–424. Vogt, T., Mullooly, J., Ernst, D., Pope, C., & Hollis, J. (1992). Social networks as predictors of ischemic heart disease, cancer, stroke and hypertension: Incidence, survival and mortality. Journal of Clinical Epidemiology, 45(6), 659–666. Watts, G., Lewis, B., Brunt, J., Lewis, E., Coltart, D., Smith, L., Mann, J., & Swan, A. (1992). Effects on coronary artery disease of lipid-lowering diet, or diet plus cholestyramine, in the St Thomas’ Atherosclerosis Regression Study (STARS). Lancet, 339(8793), 563–569. Welin, C., Lappas, G., & Wilhelmsen, L. (2000). Independent importance of psychosocial factors for prognosis after myocardial infarction. Journal of Internal Medicine, 247(6), 629–639. Williams, R. B., Barefoot, J. C., Califf, R. M., Haney, T. L., Saunders, W. B., Pryor, D. B., Hlatky, M. A., Siegler, I. C., & Mark, D. B. (1992). Prognostic importance of social and economic resources among medically treated patients with angiographically documented coronary artery disease. Journal of the American Medical Association, 267, 520–524.