EMPIRICAL RESEARCH

T H E I M P A C T O F S O C I A L S U P P O R T IN P U L M O N A R Y R E H A B I L I T A T I O N OF PATIENTS WITH CHRONIC OBSTRUCTIVE PULMONARY DISEASE 1

Sharon Grodner, Ph.D., Lela M. Prewitt, Beth A. Jaworski, B.A., Roseann Myers, R.N., Robert M. Kaplan, Ph.D., and Andrew L. Ries, M.D., M.P.H. University of California, San Diego

ABSTRACT

havioral research and practice ~1). Results of studies examining the possible relationship between social support and onset of illness are inconsistent. However, there is substantial evidence suggesting that social support is protective in the presence of chronic disease (2). Although there are various ways to conceptualize social support, most fail into two main categories----quantitative and qualitative (3). Social support is typically defined as the number of individuals in a person's network who are perceived as caring and dependable. Several authors have emphasized the importance of considering the social support network size and satisfaction as separate constructs (4-6). Previous studies substantiate the impact of social support on patients with chronic illness in modifying social stressors (7,8) and encouraging health promoting behaviors that may affect survival and other health outcomes (9,10). Studies on chronic illness and social support center around the stress buffering model of support, in which chronic illness is characterized as a chronic stress situation and support is thought to modify the deleterious effects of illness (11). Interactions between social support and morbidity and mortality have been studied in several chronic disease populations, including patients with cancer (l l), heart disease (12), diabetes (13), end stage renal disease (14), and arthritis (15,16). Positive relationships have been observed between increased social support and reduction of the stresses of chronic illness, adherence and compliance to medical treatment, and lower morbidity and mortality (17,18). However, there have been no studies examining social support in adult pulmonary disease populations. The societal impact of chronic obstructive pulmonary disease (COPD) is significant. COPD is a major cause of morbidity and mortality. In the United States, it is currently the fourth leading cause of death and is present in approximately 10% of the older adult population. Chronic illnesses like COPD produce major life stresses that require physical and psychological adjustments. They often disrupt social relationships, increase isolation, and place enormous burdens on social networks and health care providers. For these individuals, social support may be a valuable resource for coping. In this study, we examined the relationship of social support to morbidity and mortality in a group of patients with chronic obstructive pulmonary disease participating in a clinical trial of pulmonary rehabilitation. We hypothesized that social support would be related to measures of health outcome.

Social support has been shown to be an important mediator of health status and survival in chronic illness, but little information is available in patients with lung diseases. We used the Social Support Questionnaire (SSQ) to examine the relationships of number of persons (SSQ-N) and satisfaction (SSQ-S) with other measures of health status, treatment changes, and survival in 110 patients with chronic obstructive pulmonary disease (COPD) participating in a randomized, controlled clinical trial of pulmonary rehabilitation (PR). Included in the analyses were measures of lung function (FEVI.o), exercise tolerance (maximum and endurance), symptoms ratings, age, self-efficacy, depression, and gender. At baseline, SSQ-N and SSQ-S were correlated positively with self-efficacy and negatively with depression and self-reported shortness of breath (SOB). SSQ-N was also correlated with disease severity and maximum exercise tolerance (FEV1.o and VO 2 max). Using the Cox Proportional Hazard Model, SSQ-S was significantly related to improved survival up to six years. However, in multivariate analysis, after adjusting for FEVI. o and SOB which were better predictors of survivaL SSQ-S was marginally significant. SSQ-S and survival were computed separately for males and females across treatment groups. SSQ-S was significantly related to mortality for women but not for men. We conclude that social support is related to measures of physical and psychological function in patients with COPD and may influence improvement and survival after pulmonary rehabilitation.

(Ann Behav Med 1996, 18(3):139-145)

INTRODUCTION The impact of social support on psychological and physicai health has gained increasing, attention in medical and he1Preparation of this manuscript was supported in part by Grants HL 34732 and HL 02215 from the National Heart, Lung and Blood Institute, Grant HD/HL 30912 from the National Center for Rehabilitation Research, National Institute of Child Health and Human Development, and Grant RR 00827 from the Division of Research Resources for the UCSD Clinical Research Center.

Reprint Address: A.L. Ries, M.D., M.P.H., Division of Pulmonary and Critical Care Medicine, UCSD Medical Center, 200 West Arbor Drive, San Diego, CA 92103-8377. 9 1996 by The Society of Behavioral Medicine.

139

140

ANNALS OF BEHAVIORAL

MEDICINE

METHODS Subjects The subjects included in this analysis were 110 of 119 patients with COPD who participated in a randomized clinical trial of pulmonary rehabilitation. Methods and preliminary results of this clinical trial have been reported previously (19). The nine patients excluded did not complete the social support questionnaire. All subjects met the following entry criteria: (a) clinical diagnosis of COPD confirmed by history, physical examination, spirometry, and chest roentgenogram; (b) no other significant lung disease; (c) clinically stable on an acceptable medical regimen; and (d) no unstable cardiac or other disease that would limit participation in the rehabilitation program. Assessment Each subject underwent comprehensive pulmonary function and exercise tests and completed psychosocial questionnaires at baseline, after the interventions, and at follow-up assessments for up to six years. Pulmonary function was evaluated using comprehensive tests following standard methods (20,21). For this analysis, we utilized FEVt0, an important measure of disease severity in COPD and one that was identified in a previous factor analysis as representative of airflow obstruction in these subjects (22). Exercise performance was evaluated with tests of both maximal exercise tolerance and endurance. Maximal exercise tolerance was measured with a symptom-limited exercise test to the highest tolerable level on a treadmill. In this incremental exercise test, the treadmill speed was increased at one-minute intervals by 0.5 miles per hour up to 3.0 miles per hour with further work increments made by increasing elevation by 2% each minute. During the exercise test, expired gases were analyzed to measure oxygen uptake (VO2) and other related variables (23). Maximum VO2 (VO2 max) is a standard measure of maximal exercise tolerance. An electrocardiogram was used to measure heart rate and monitor for arrhythmias or cardiac ischemia. Perceived symptoms of breathlessness and muscle fatigue were rated at the end of the exercise test using a scale adapted from Borg (24). Exercise endurance was measured on a separate day. This test was performed at a constant work level chosen from the initial maximum exercise test to estimate each subject's symptom-limited capacity for steady-state walking (25). On average, the target exercise level represented 95% of initial maximal work load. Patients were instructed to walk up to 20 minutes at this level and, if possible, for an additional 10 minutes at a higher level. Measurements were made of the total time at the target levels on the treadmill (maximum = 30 minutes); perceived symptom ratings of breathlessness and muscle fatigue were obtained at the end of the exercise test using the modified Borg scale. Psychosocial questionnaires completed at baseline and used in this analysis included the following:

Social Support: The Social Support Questionnaire (SSQ) (26) measures the number of persons in the social support network (SSQ-N) and self-perception of satisfaction with the available support (SSQ-S). The original SSQ questionaire has 27 self-administered items. In the first section, patients are requested to list individuals on whom they depend in a variety of situations. In the second section, the subjects rate their satisfaction with their support network on a six-point scale. For

G r o d n e r et al. this study, we used a modified twelve-item version of the SSQ. The short form of the SSQ has been shown to correlate well with the original version (27). The SSQ yields two scores: SSQ-N, reflecting the mean number of persons listed across the items; and SSQ-S, reflecting the mean satisfaction rating across all items. When compared with 23 other social support scales, the SSQ has been found to have a strong record of reliability and validity (28,29). In studies using over 200 subjects, alpha reliability has been reported at 0.94 and above.

Self-Efficacy: The self-efficacy questionnaire for this study was adapted from a previous version used by by Kaplan, Atkins, and Reinsch (30). The subscale used in this analysis emphasizes walking and includes the following statements: walk one block (approximately 5 minutes), walk two blocks (10 minutes), walk three blocks (15 minutes) . . . walk three miles (90 minutes). The subject rates the degree of confidence or strength of expectation to perform that activity on a 100-point probability scale, ranging in 10-point intervals from 0 (complete uncertainty) to 100 (complete certainty). The self-efficacy score reflects the highest walking level for which the patient expressed 100% confidence. Well Being: The Quality of Well Being Scale (QWB) is a comprehensive measure of health-related quality-of-life that includes several components. First, it obtains observable levels of functioning at a point in time. The levels of functioning are obtained from three separate scales: mobility, physical activity, and social activity. Second, symptoms or problems are selected from a standard, comprehensive list, and the most undesirable symptom is noted. Then, the level of function and symptom reports are weighted by preference or the desirability of the state on a scale ranging from 0 (dead) to 1.0 (optimum function). The weights are standard ones, obtained from independent samples of judges who rate the desirability of the observable health status and symptom problem combination. This system has been used extensively in medical and health services research applications (31). In addition, specific validity and reliability studies using this measure for patients with COPD have been published (32). These studies demonstrate that the QWB scale is sensitive to relatively minor changes in health status and that it is correlated with a variety of physical and functional measures of health. Depression: The Centers for Epidemiologic Studies Depression Scale (CES-D) is a general measure of depressive symptoms that has been used extensively in epidemiologic studies (33). The scale includes 20 items and measures dimensions of depressed mood, feelings of guilt and worthlessness, appetite loss, sleep disturbance, and fatigue. These items are assumed to be representative of the major components of depressive symptomatology. The CES-D has demonstrated high internal consistency. The alpha coefficients ranged from 0.84 to 0.90. Since the scale was devised to measure current depressive symptoms, test-retest reliability decreased as time increased between repeat administrations of the scale. Nevertheless, correlations were in the range of 0.45 to 0.70, which are values similar to those reported for other depression scales (34). The CES-D was designed for assessing non-psychiatric populations (35). Although several studies have demonstrated that the CES-D is associated with clinical diagnosis of depression, it has been found to be better as a screening instrument than as a diagnostic tool. Scores on the CES-D scale of less than 16 are highly associated with clinical judgments of non-

S o c i a l S u p p o r t in C O P D

VOLUME

TABLE 1 Results of Selected Baseline Measures in 110 Patients with COPD

18, N U M B E R

3, 1 9 9 6

141

TABLE 2 Correlation of SSQ-N and SSQ-S with Baseline Data

Variable

Mean -+ SD

Variable

SSQ-N

SSQ-S

Age Gender FEVL0, L VO2 max, ml/kg/min Exercise endurance, min Perceived breathlessness Perceived fatigue SSQ-N SSQ-S Self-efficacy--Walking QWB CES-D SOBQ

62.9 - 7.1 (range = 42 to 77) 78M/32F 1.39 -+ 0.64 16.7 _+ 6.0 11.7 _+ 7.9 4.7 _+ 2.2 4.2 _+ 2.1 2.3 -+ 1.7 5.3 - 1.1 3.9 -+ 3.2 0.659 -2- 0.081 14.5 -+ 8.9 34.5 -+ 19.0

Age FEVI.0 VO2 max Exercise endurance Perceived breathlessness Perceived fatigue SSQ-S Self-efficacy--Walking QWB CES-D SOBQ

.08 .24* .21" .07 .08 .11 .25** .22* - .03 -.31"* -.23*

.04 .17 .16 .09 - . 10 - . 10 -.24* .01 -.26"* -.18

FEV~.0 = forced expiratory volume in one second; VO2 max = maximum oxygen uptake during treadmill exercise test; SSQ-N = number of persons in the social support network; SSQ-S = satisfaction with the social support network; QWB = Quality of Well Being Scale; CES-D = Centers for Epidemiology Depression Scale; SOBQ = UCSD Shortness of Breath Questionnaire. depression; however, scores of 17 or higher are only moderately related to clinical diagnoses of depression.

Shortness of Breath: The Shortness of Breath Questionnaire (SOBQ), version A, was developed and has been used widely in the Pulmonary Rehabilitation Program at the University of California, San Diego (36). Subjects are asked to indicate how frequently they experience shortness of breath on a six-point scale (0 = never, 1 = sometimes, 2 = half of the time, 3 = most of the time, 4 = all of the time, and NA = not applicable or unable to perform) during 21 different activities of daily living associated with varying levels of activity. Additionally, three questions inquiring overall about limitations due to shortness of breath, fear of self-harm from overexertion, and fear of shortness of breath are included for a total of 24 items. The score used in this analysis is the total shortness of breath reported over all 24 questions. Missing responses or responses marked "not applicable" are scored as 0. Clinical Trial Interventions After baseline assessment, subjects in this study were randomly assigned to an eight-week intervention of either comprehensive pulmonary rehabilitation (N = 54) or an education control program (N = 56). The rehabilitation program involved twelve sessions in eight weeks. Each four-hour session included two classes or group sessions plus supervised exercise training. Patients were enrolled in groups of three to five at approximately one-month intervals. During the eight-week core program, each patient was given individualized instruction, including education, respiratory and physical care techniques, exercise training, and psychosocial support. The education control group involved four biweekly meetings, at which time patients received information about pulmonary disease but did not receive the behavioral components or the individualized instruction of the rehabilitation program. In particular, those in the education control group did not participate in supervised exercise training. All patients had moderate to severe COPD, marked exercise limitation and perceived symptoms, and considerable dis-

N = 110; *p < .05; **p < .01. FEV1.0 = forced expiratory volume in one second; VO2 max = maximum oxygen uptake during treadmill exercise test; SSQ-N = number of persons in the social support network; SSQ-S = satisfaction with the social support network; QWB = Quality of Well Being Scale; CES-D = Centers for Epidemiology Depression Scale; SOBQ = UCSD Shortness of Breath Questionnaire.

ability from their disease. There were no significant differences between the experimental and control groups at baseline.

Statistical Analysis Descriptive statistics (mean, standard deviation, and range) were calculated for baseline data on all study subjects as well as for each experimental treatment group. Comparisons between groups were evaluated by t-tests. Correlations were calculated to describe the relationship of baseline social support measures with other baseline variables for all subjects as well as change in measurements after interventions for each experimental group. The relationship of social support and other measured variables on survival was evaluated using the Cox Proportional Hazard technique. For this analysis, group assignment (rehabilitation and education control) was coded and entered as a categorical variable. A median split on SSQ-S was then performed to compare the survival of subjects with low versus high satisfaction of social support. Univariate survivor functions were also computed separately across treatment groups for males and females. RESULTS Results of baseline measures are summarized in Table 1. These patients had moderate to severe COPD, marked exercise limitation and symptoms, and considerable disability from their disease. Correlations between SSQ-N and SSQ-S and other baseline variables are presented in Table 2. Significant positive correlations were found at baseline between scores of both SSQ-N and SSQ-S and self-efficacy for walking (SEW). At baseline, scores on the SSQ-N and SSQ-S were found to correlate negatively with the CES-D depression scale. The measures of disease severity (FEV1.0), maximum exercise tolerance (VO2 max), and self-reported shortness of breath (SOBQ) were correlated significantly with SSQ-N but not with SSQ-S. After the treatment interventions, correlations between baseline measures of social support (SSQ-N and SSQ-S) and changes in outcome variables at two, six, and twelve months were used to assess the influence of social support on response

142

ANNALS

OF BEHAVIORAL

Grodner

MEDICINE

TABLE 3 Comparison of Baseline Measures in Survivors and Non-Survivors Variable

Survivors Mean (SD)

Non-Survivors Mean (SD)

N Age Gender FEV,.0, L FEVL0/FVC (%) VO 2 max, ml/kg/min Exercise endurance, min Perceived breathlessness Perceived fatigue SSQ-N SSQ-S Self-efficacy--Walking QWB CES-D SOBQ

68 62.5 (7.8) 49M/19F 1.61 (0.67) 47.7 (12.4) 19.0 (5.7) 12.5 (7.9) 4.7 (2.2) 4.2 (2.1) 2.4 (1.9) 5.4 (0.9) 4.5 (3.3) 0.664 (0.074) 14.8 (9.6) 29.8 (16.3)

42 63.6 (5.9) 29M/13F 1.03 (0.38)t 39.2 (10.9)t 13.1 (4.6)1" 10.4 (7.7) 4.7 (2.2) 4.1 (2.1) 2.1 (1.4) 5.0 (1.3) 3.0 (3.0)* 0.652 (0.092) 14.1 (7.8) 42.2 (2.7)**

*p < .05; **p < .01; t p < .001. FEVL0 = forced expiratory volume in one second; VO2 max = maximum oxygen uptake during treadmill exercise test; SSQ-N = number of persons in the social support network; SSQ-S = satisfaction with the social support network; QWB = Quality of Well Being Scale; CES-D = Centers for Epidemiology Depression Scale; SOBQ = UCSD Shortness of Breath Questionnaire.

to the treatments. In this study, patients in the pulmonary rehabilitation group (N = 54) demonstrated highly significant improvement in exercise performance and symptoms. For these patients, SSQ-N was significantly correlated with improvement in perceived breathlessness at two months (r = 0.37, N = 49), six months (r = 0.30, N = 49), and twelve months (r = 0.32, N = 40). None of the other correlations between measures of social support and changes after intervention were significant for either the experimental (rehabilitation) group or control (education) group. Over the course of six years of follow-up, there were 68 survivors and 42 non-survivors in this study. A comparison of selected baseline measures for survivors and non-survivors is presented in Table 3. Non-survivors had significantly lower lung function (FEV~.0), maximum exercise tolerance (VO2 max), self-efficacy for walking, and shortness of breath ratings (SOBQ). Using the Cox Proportional Hazards Model, SSQ-S (and not SSQ-N) was related to improved survival for up to six years [p < 0.05; hazard ratio = 0.79; (95% C.I. 0.63-0.99) reflecting a 21% change in the likelihood of survival for one unit change in SSQ-S]; however, when controlling for FEV~.0 and SOBQ, SSQ-S was not significant [p < 0.16; hazard ratio = 0.84 (95% C.I. 0.66-1.07]. The Kaplan-Meier survival curves of the subjects with high and low SSQ-S (median split) are presented in Figure 1 for all subjects and in Figures 2 and 3 for males and females only, respectively. For all subjects (Figure 1), these curves were not significantly different (p < 0.10), although there was a trend of reduced long-term survival for the subjects with low SSQ-S. Using the Proportional Hazards Model, the hazard ratio for subjects with low SSQ-S was 1.61 (95% C.I. 0.88-2.97) compared to subjects with high SSQ-S (p < 0.12). However, when the analysis was performed adjusting for FEV~.0, the hazard ratio was reduced to 1.33 (95% C.I. 0.72-2.44; p = 0.36).

1.0

et a l .

--

0.90.8"~ 0 . 7 ._> w

0.6 .~" 0.5 o o..

0.40.30.2All subjects (n=l 10)

0.4 0.0

I 0

i

I

i

I

I

i

1

2

3

4

5

6

Years

Survival of all 110 subjects for high versus F I G U R E 1: low social support satisfaction.

There was a difference between males and females in the relationship between SSQ-S and survival. For males (Figure 2), there was no significant difference in survival between the low and high SSQ-S groups. For females, however (Figure 3), survival for subjects with high SSQ-S was significantly better than for those with low SSQ-S (p < 0.05).

1.0 0.9 0.8 .>

%

0.7

Low

0.8

I

o.5 ..Q

0.4

0.3 0.2 0.1

0.0

Males (n=78)

i 0

i 1

I

I

I

I

I

2

3

4

5

6

Years

Survival of 78 males for high versus low F I G U R E 2: social support satisfaction.

Social S u p p o r t in C O P D 1.0

I

0.9

[__

.>_

o

High

I

0.7 0.6

~

L

1

0.8

V O L U M E 18, N U M B E R 3, 1996

I L_

L_ L 1

Low

0.5 0.4

0.

0.3

0.2 Females (n=32)

0.1 0.0 0

I

I

I

I

I

I

1

2

3

4

5

6

Years

Survival of 32 females for high versus low F I G U R E 3: social support satisfaction. DISCUSSION The results of this study suggest that measures of social support may be related to measures of health status and survival in patients with moderate to severe COPD. The data were analyzed several different ways. Using the Cox Proportional Hazards Model, social support satisfaction but not social support network size was found to be a significant prospective predictor of mortality. The Kaplan-Meier analysis, which requires splitting social support into low and high categories, was used on the same data. In this analysis, the groups were divided by median split. Even this crude analysis indicated a trend (p = .10) toward improved long-term survival among those with higher social support satisfaction at baseline. Further analysis demonstrated that social support was significant for women but not for men. We remain uncertain about the exact explanation for these results. It is not clear whether the effect is related to physiological or psychological factors. For example, FEV1.0 is an important indicator of disease severity and the most significant predictor of overall mortality for patients with COPD (19). When using multivariate analysis to adjust for disease severity, the predicted relationship between social support and mortality diminishes. Correlation analysis demonstrates that the number of people in the patient's social support network was correlated with FEVI.0 as well as exercise tolerance and shortness of breath. These findings suggest that more functional patients experience greater social support. In other words, increased severity of illness may interfere with an individual's ability to elicit, make use of, and be satisfied with what support is offered, thereby deepening social isolation. Measures of social support were obtained only at baseline, Consequently, it is not possible to determine from our longitudinal study whether deterioration in health status preceded disruption in social support. At baseline, there were significant correlations between measures of social network size and sat-

143

isfaction with social support with less severe depression and self-efficacy for walking. However, the multivariate model did not indicate that these psychological variables mediated the relationship between social support and survival. It did not appear that measures of social support were related to outcome from the pulmonary education or rehabilitation intervention programs in this clinical trial. However, the significant correlation between SSQ-N and perceived breathlessness found after rehabilitation is noteworthy because changes in breathlessness are the most important and consistent effects of pulmonary rehabilitation. One explanation is that social support is provided by the program structure, the staff, and the other patients. Thus, while participating in the rehabilitation program, patients may have been less dependent upon their own support networks. Since social support data was only available at baseline, it is difficult to determine psychological aspects of the relationship between support and survival. For example, the stressbuffering model emphasizes the role social support plays in absorbing stress caused by the deleterious effects of illness or other life circumstances, and in promoting positive adaptive health behaviors (37-39). However, a comprehensive assessment of the stress-buffering model would require measures of life stress that were not available in this data set. In order to fully address these issues, future studies need to include measures of life stress. One of the most important findings in this study was that women with low social support satisfaction (SSQ-S) were significantly more likely to die within six years than women with high satisfaction in their social relationships. In contrast, no differences in survival were found for males with low versus high social support satisfaction. The literature on gender differences in social support and survival is inconsistent. Some studies indicate that the relationship of social support to mortality is stronger for men than for women. The Tecumseh, Michigan Community Health Study measured social support networks including marital status, visits with friends and family, and organizational involvement. It was found that greater social support was associated with increased mortality in elderly women (40-42). On the other hand, several studies have shown that social support predicts increased longevity for women rather than for men. For example, the Framingham Heart Study demonstrated that women, but not men, with unsupportive spouses were more likely to die of heart disease (43). On the other hand, the Alameda County Study found that social support was associated with survival for both males and females (44). The gender differences in the nature and quality of social support for males and females has just recently been evaluated. Our results are consistent with other studies showing that women typically have larger networks than men (42). However, in our study, life expectancy was associated with quality and not quantity of relationships for women. This finding may reflect that although women's networks are larger and provide more opportunity for support, they may also increase the demands on resources and negate potential benefits. Bolger and Eckkenrode (45) reported that social visits with friends and neighbors would serve as buffers against stress only when the contacts were "discretionary" and not "obligatory." Women are more likely than men to be support providers as well as support recipients, and women are more likely than men to provide support that places emotional burdens on themselves. Until recently, COPD was primarily a disease of men. However, with the increased

144

ANNALS OF BEHAVIORAL

MEDICINE

number of women smoking cigarettes in the 1950s, 1960s, and 1970s, the proportion of women with COPD is approaching that of men. Consequently, studies of gender differences in adaptation to COPD warrants consideration. The generalizability of our findings is hindered by a variety of factors. It should be noted that there were only 42 deaths out of the 110 subjects. Therefore, statistical power to detect a 16% survival benefit of social support is limited. Further, we used the social support questionaire to assess network size and satisfaction with social relationships. These constructs, used widely in previous studies, are said to be predictive of health outcome. However, there are a wide variety of measures of social support and different approaches that capture other aspects of this construct (46--52). The possible effects of specific types of support (advice, education, tangible aid, and emotional support), particular providers (spouse, family, friends, or physician), or positive or negative aspects were not assessed; it may be of benefit to study them separately (53-58). We encourage the use of other measures in future studies in order to enhance the generalizability of the results. We are limited in our ability to test specific models of the relationship between social support and health outcomes, since social support measures were obtained at baseline only. Additionally, measures of psychological stress were not collected as part of this protocol. It is apparent that the connection between health and social support is complicated. Previous studies of social support and chronic illness have not included COPD patients. The current study demonstrates the potential benefits of social support on morbidity and mortality in patients with this syndrome. It would be helpful to understand how naturally occurring social support networks influence outcomes for patients with COPD. Identifying the factors that promote and maintain support will aid in designing optimal interventions for these individuals. Additional research is needed to further clarify the influence of social support on illness severity, progression, and survival in patients with chronic lung disease.

REFERENCES (1) Veiel HOE Baumann V (eds): The Meaning and Measurement of Social Support. New York: Hemisphere, 1992. (2) Wallston BS, Alagna SW, DeVeUisB, DeVellis RF: Social support and physical health. Health Psychology. 1983, 2:367-391. (3) Wellman B: From social support to social network. In Sarason IG, Sarason BR (eds), Social Support: Theory Research and Applications. Boston, MA: Martinus Nijhoff, 1985, 205-222. (4) Kaplan RM, Hartwell SL: Differential effects of social support and social network on physiological and social outcomes in men and women with Type II diabetes mellitus. Health Psychology. 1987, 6:387-398. (5) Schwartz LS, Springer J, Flaherty JA, Kiani R: The role of recent life events and social support in the control of diabetes mellitus. General Hospital Psychiatry. 1986, 8:212-216. (6) Gomer KO, Johnson JV: Social network interaction and mortality. A six-year follow-up study of a random sample of the Swedish population. Journal of Chronic Diseases. 1987, 40:949-957. (7) Blazer DG: Social support and mortality in an elderly community population. American Journal of Epidemiology. 1982, 115:684694. (8) Ruberman W, Weinblatt E, Goldberg JD: Education, psychosocial stress, and sudden cardiac death. Journal of Chronic Diseases. 1983, 36:151-160.

G r o d n e r et al. (9) Wilcox B: Social support, life stress, and psychological adjustment: A test of the buffering hypothesis. American Journal of Community Psychology. 1981, 9:371-386. (10) Wiklund I, Oden A, Sanne H, et al: Prognostic importance of somatic and psychosocial variables after a first myocardial infarction. American Journal of Epidemiology. 1988, 128.'786-795. (11) Taylor SE, Falke RL, Shoptaw SJ, Lichtman RR: Social support, support groups, and the cancer patient. Journal of Consulting and Clinical Psychology. 1986, 54:608-615. (12) Kaplan GA, Salonen JT, Cohen RD, et al: Social connections and mortality from all causes and from cardiovascular disease: Prospective evidence from eastern Finland. American Journal of Epidemiology. 1988, 128:370-380. (13) Marteau TM, Bloch S, Baum JD: Family life and diabetic control. Journal of Child Psychology and Psychiatry. 1987, 28:823-833. (14) Siegal BR, Calsyn ILl, Cuddihee RM: The relationship of social support to psychological adjustment in the end-stage renal disease patients. Journal of Chronic Diseases. 1987, 40:337-344. (15) Fitzpatrick R, Newman S, Lamb R, Shipley M: Social relationships and psychological well-being in rheumatoid arthritis. Social Science and Medicine. 1988, 27:399-403. (16) DeVellis RF, Sauter DeVellis BM, Sauter SVH, Cohen JL: Predictors of pain and functioning in arthritis. Health Education and Research. 1986, 1:61-67. (17) House JS, Landis KR, Umberson D: Social relationships and health. Science. 1988, 241:540-545. (18) Neale AV, Tilley BC, Vernon SW: Marital status, delay in seeking treatment, and survival from breast cancer. Social Science and Medicine. 1986, 23:305-312. (19) Ries AL, Kaplan RM, Limberg TL, Prewitt LM: Effects of pulmonary rehabilitation on physiologic and psychosocial outcomes in patients with chronic obstructive pulmonary disease. Annals of Internal Medicine. 1995, 122:823-832. (20) American Thoracic Society: ATS statement: Snowbird workshop on standardization of spirometry. American Review of Respiratory Disease. 1979, 119:831-839. (21) Clausen JL, Zarins LP: Pulmonary Function Testing Guidelines and Controversies: Equipment, Methods, and Normal Values. New York: Academic Press, 1982. (22) Ries AL, Kaplan RM, Blumberg E: Use of factor analysis to consolidate multiple outcome measures in chronic obstructive pulmonary disease. Journal of Clinical Epidemiology. 1991, 44:497503. (23) Wasserman K, Hansen JE, Sue DY, Whipp BJ: Principles of exercise testing and interpretation. Philadelphia, PA: Lea & Febiger, 1987. (24) Borg GAV: Psychophysical bases of perceived exertion. Medicine and Science in Sports Exercise. 1982, 14:377-381. (25) Punzal PA, Ries AI, Kaplan RM, Prewitt LM: Maximum intensity exercise training in patients with chronic obstructive pulmonary disease. Chest. 1991, 100:618-623. (26) Sarason IG, Levin HM, Bashan RB, Sarason BR: Assessing social support: The Social Support Questionnaire. Journal of Personality and Social Psychology. 1983, 44:127-139. (27) Sarason IG, Sarason BR, Shearin EN, Pierce GR: A brief measure of social support: Practical and theoretical implications. Journal of Social and Personal Relationships. 1986, 4:497-510. (28) Heitzmann CA, Kaplan RM: Assessment of methods for measuring social support. Health Psychology. 1988, 7:75-109. (29) Sarason IG, Levine HM, Basham RB, Sarason BR: Assessing social support: The Social Support Questionnaire. Journal of Personality and Social Psychology. 1983, 44:127-139. (30) Kaplan RM, Atkins CJ, Reinsch S: Self-efficacy expectations mediate exercise compliance in patients with COPD. Health Psychology. 1984, 3:223-242. (31) Kaplan RM, Anderson JP: A general health model policy: Update and applications. Health Service Research. 1988, 23:203-235.

Social Support in COPD (32) Kaplan RM, Atkins CJ, Timms R: Validity of a quality of wellbeing scale as an outcome measure in chronic obstructive pulmonary disease. Journal of Chronic Diseases. 1984, 37.'85-95. (33) Weissman MM, Sholomskas D, Pottenger M, Prusoff BA, Locke BZ: Assessing depressive symptoms in five psychiatric populations: A validation study. American Journal of Epidemiology. 1977, 6:203-214. (34) Radloff LS: The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychology Measurement. 1977, 1:385. (35) Gottlieb IH, Cine DB: Self-report assessment of depression and anxiety. In Kendill PC, Watson D (eds), Anxiety and Depression: Distinctive and Overlapping Features. San Diego, CA: Academic Press, 1989, 131-169. (36) Archibald CJ, Guidotti TL: Degree of objectivity measured impairment and perceived shortness of breath with activities of daily living in patients with chronic obstructive pulmonary disease. Canadian Journal of Clinical Rehabilitation~ 1987, 1:45-54. (37) Wallston BS, Wallston KA: Social psychological models of health behavior. In Banm A, Taylor S, Singer JE (eds),. Handbook of Psychology and Health (Vol. IV). Hillsdale, NJ: Lawrence Erlbaum Associates, 1983. (38) Cohen S, Wills TA: Stress, social support, and the buffering hypothesis. Psychological Bulletin. 1985, 98:310-357. (39) Cohen S: Psychosocial models of the role of social support in the etiology of physical disease. Health Psychology. 1988, 7.'269-297. (40) Berkman LE Syme SL: Social networks, host resistance, and mortality: A nine-year follow-up study of Alameda county residents. American Journal of Epidemiology. 1979, 109:186-204. (41) House JS, Robbins C, Mejzneer HL: The association of social relationships with mortality: Prospective evidence from the Tecumseh Community Health Study. American Journal of Epidemiology. 1982, 116:123-t40. (42) Shumaker SA, Hill DR: Gender difference in social support and physical health. Health Psychology. 1991, 10:i02-i i I. (43) Haynes SG, Feinleib M: Women, work, and coronary heart disease: Prospective findings from the Framingham heart study. American Journal of Public Health. 1980, 70:133-141. (44) Berkman LF: Social networks, support, and health: Taking the next step forward. American Journal of Epidemiology. 1986, 123:559562.

V O L U M E 18, N U M B E R 3, 1996

145

(45) Bolger N, Eckenrode J: Social relationships, personality, and anxiety during a major stressful event. Journal of Personality and Social Psychology. 1991, 61:440-449. (46) Cassileth BR, Walsh WP, Lusk EJ: Psychosocial correlates of cancer survival: A subsequent report 3 to 8 years after cancer diagnosis. Journal of Clinical Oncology. 1988, 6:1753-1759. (47) Thoits PA: Life stress, social support, and psychological vulnerability: Epidemiological considerations. Journal of Community Psychology. 1982, 10:341-362. (48) Wortman CB: Social support and the cancer patient. Cancer. 1984, 53:2339-2360. (49) Lieberman M: The effects of social support on responses to stress. In Goldberg L, Breznitz S (eds), Handbook of Stress. New York: The Free Press, 1982, 764-782. (50) Lazarus RS, Folkman S: Stress Appraisal and Coping. New York: Springer, 1984. (51) Dakoff GA, Taylor SE: Victims' perceptions of social support: What is helpful from whom? Journal of Personality and Social Psychology. 1990, 58:80-89. (52) Sarason BR, Pierce GR, Sarason IG: Social support and interactional processes: A triadic hypothesis. Journal of Social and Personal Relationships. 1990, 7:495-506. (53) Block AR: An investigation of the response of the spouse to the chronic pain behavior. Psychosomatic Medicine. 1981, 43:415422. (54) Flor H, Turk D, Rudy T: Relationship of pain impact and significant other reinforcement of pain behaviors: The mediating role of gender, marital status, and marital satisfaction. Pain. 1989, 38:4550. (55) Kaplan RM, Chadwick MW, Schimmel LE: Social learning intervention to promote metabolic control in Type I diabetes meUitus: Pilot experiment results. Diabetes Care. 1985, 8:152-155. (56) Wortman CB, Dunkel-Schetter C: Conceptual and methodological issues in the study of social support. In Baum A, Singer JE (eds), Handbook of Psychology and Health. Hillsdale, NJ: Lawrence Erlbaum Associates, 1987, 63-108. (57) DiMatteo MR, Hays R: Social support and serious illness. In Gottlieb BH (ed), Social Networks and Social Support. Beverly Hills, CA: Sage, 1981, 117-148. (58) Stephens MAR, Kinney JM, Norris VK, Ritchie SW: Social networks as assets and liabilities in recovery from stroke by geriatric patients. Psychology andAging. 1987, 2:125-229.