J Behav Med DOI 10.1007/s10865-014-9559-4
Spouse health behavior outcomes from a randomized controlled trial of a spouse-assisted lifestyle change intervention to improve patient low-density lipoprotein cholesterol Heather A. King • Amy S. Jeffreys • Megan A. McVay Cynthia J. Coffman • Corrine I. Voils
•
Received: April 28, 2013 / Accepted: February 10, 2014 Ó Springer Science+Business Media New York (outside the USA) 2014
Abstract This study evaluated spouse health behavior outcomes from a randomized controlled trial of a spouseassisted lifestyle intervention to reduce patient low-density lipoprotein cholesterol and improve patient health behaviors. Participants were 251 spouses of patients from the Durham Veterans Affairs Medical Center randomized to intervention or usual care. The intervention comprised 9 monthly telephone calls to patients and spouses. Outcomes were assessed at baseline, 6 and 11 months. At 11 months, there were no differences in spouse outcomes between intervention and usual care groups for moderate intensity physical activity (i.e., frequency, duration) or dietary intake (i.e., total calories, total fat, percentage of calories from total fat, saturated fat, percentage of calories from saturated fat, cholesterol, fiber). To improve spouse outcomes, couple interventions may need to include spouse behavior change goals and reciprocal support between patients and spouses and consider the need for improvement in spouse outcomes. Keywords Spouse Couple Dietary intake Physical activity Intervention Social support H. A. King (&) A. S. Jeffreys M. A. McVay C. J. Coffman C. I. Voils Center for Health Services Research in Primary Care, Durham Veterans Affairs Medical Center, HSR&D (152), 508 Fulton St., Durham, NC 27705, USA e-mail: [email protected] M. A. McVay C. I. Voils Department of Medicine, Duke University Medical Center, Durham, NC, USA C. J. Coffman Department of Biostatistics and Bioinformatics, Duke University Medical Center, Durham, NC, USA
Introduction Couple interventions (i.e., both patient and partner are active participants) for chronic illness have significant small effects on some patient outcomes and in some cases are more efficacious than both usual care and psychosocial interventions involving only the patient (Martire et al., 2010). Unclear is whether such interventions potentially benefit only patients or also their spouses. Observational studies indicate significant concordance between patients and their spouses for behavioral risk factors such as smoking, dietary intake, and physical activity (Di Castelnuovo et al., 2009; Macken et al., 2000). Therefore, if spouse health behavior outcomes were also positively affected by couple interventions, then such interventions may be particularly efficacious and efficient in terms of time and cost. In a recent review and meta-analysis of couple interventions for chronic illness conducted by Martire et al. (2010), only half of the reviewed studies examined effects on spouse outcomes. Those that did had various methodological limitations and focused primarily on psychosocial outcomes. Only one reviewed study assessed spouse health behaviors (Wing et al., 1991): this study on obese Type 2 diabetes patients and their overweight spouses found greater spouse weight loss and changes in eating behavior strategies among spouses in a couple-focused behavioral weight loss program compared to a patient-only program. In addition, a recent trial evaluated the effectiveness of a couple intervention, specifically a spouse-assisted lifestyle intervention, to reduce patient cholesterol levels and improve patient health behaviors such as dietary intake and physical activity (CouPLES; Voils et al., 2009). In a partnerassisted lifestyle intervention, the role of the partner or spouse is to support patient behavior change and intervention adherence (Baucom et al., 1998). Among patients, no
123
J Behav Med
significant differences between intervention and usual care groups were found for low-density lipoprotein cholesterol or physical activity at the final assessment (11 months). However, patient caloric intake, total and saturated fat intake, and percentage of calories from total fat were lower in the intervention than in the usual care control group (Voils et al., 2013). In this paper, we expand on these previous analyses by examining a distinct research question: whether health behaviors of physical activity and dietary intake improved in spouses as a function of their involvement in this spouse-assisted intervention.
National Heart, Lung, and Blood Institute, 2001) and information on providing spousal support; baseline measurements were collected; and then couples were randomized to usual care or the intervention, stratified by patient race (White vs. non-White) and patient coronary heart disease risk level (low/moderate vs. high). Patients in the usual care group received clinical management of lipid disorders by their providers. Neither patients nor spouses received further contact from study staff except for outcome assessment visits. The trial was approved by ethics committees and conducted at the Durham Veterans Affairs Medical Center. Patient recruitment and enrollment
Method Participants Detailed methods and the CONSORT flow diagram outlining enrollment and intervention allocation have been reported elsewhere (Voils et al., 2009, 2011, 2013). Ultimately, 255 couples were randomized (127 to the intervention, 128 to usual care) following patient baseline assessments. In four cases, patients attended baseline appointments alone, and spouses never returned for assessments; therefore, data from 251 spouses (124 intervention, 127 usual care) were available for analyses. Assessments were completed for 248 spouses at baseline (122 intervention, 126 usual care), 164 at the 6-month follow-up (88 intervention, 76 usual care), and 176 (90 intervention, 86 usual care) at the 11-month follow-up. Spouse baseline characteristics are provided in Table 1.
Briefly, couples received a recruitment letter prior to a patient primary care visit, followed by a screening telephone call. To be eligible, patients had to be married and have elevated low-density lipoprotein cholesterol. Patient and spouse exclusion criteria assessed during the screening call included inconsistent access to telephone; impaired cognition (Callahan et al., 2002) or hearing; health problem that precluded participation; and resident of nursing home or receiving long-term care. Couples passing the telephone screening were scheduled for baseline assessments, where written informed consent was obtained and self-report measures were completed. Patients and spouses initially received $10 each for each outcome visit; later, they received $20 to address increases in transportation expenses. Couples intervention
Procedure Setting and design In this two-group, randomized controlled trial, all couples received print educational materials containing guidelineconcordant recommendations (NCEP ATP III guidelines;
The intervention was informed by social cognitive theory. It was delivered via phone by a nurse interventionist using a custom software package to maximize intervention fidelity. Delivering an intervention by phone is less expensive and can reach more patients compared to faceto-face intervention delivery. Intervention telephone calls
Table 1 Baseline characteristics of enrolled spouses Demographic variable
All (N = 251)
Intervention (N = 124)
Usual care (N = 127)
Age, M (SD)
58.8 (12.0)
59.4 (12.3)
58.2 (11.6)
White (%) Female (%)
64.1 94.0
66.4 91.1
61.9 96.9
High school grad or less (%)
26.6
25.4
27.8
Full-time employment (%)
34.3
27.1
41.3
Weekly frequency of moderate intensity physical activity, median (IQR)
7 (2, 12)
6 (2, 11)
7 (3, 12)
Weekly duration of moderate intensity physical activity (hours), median (IQR)
3.75 (1.0, 9.5)
3.75 (1.0, 7.5)
4 (1.5, 10.0)
3 missing for age, race, education, and employment status (2 intervention, 1 usual care). 7 missing for weekly frequency and duration of moderate intensity physical activity (3 intervention, 4 usual care), Participants with missing data are excluded from percentage calculations M mean, SD standard deviation, IQR interquartile range
123
J Behav Med
were delivered monthly first to patients and then to spouses. In month 1, patients and spouses received information about hypercholesterolemia and an overview of self-management principles. Spouses also received an orientation on strategies to support patient goal achievement and were encouraged to provide the patients with support and praise, offer help, reward desired health behaviors rather than criticizing poor health behaviors, focus on patients’ goals instead of what spouses thought patients should be doing, and help patients establish non-food related rewards for progress toward their goals. In months 2–5 and 7–10, calls with patients focused on goal-setting and problem-solving. In all calls, patients created behavioral goals and plans, most often in relation to diet or physical activity. Following patient calls, spouses were informed of patients’ successes, goals, and action plans for the upcoming month and were asked to record them. Spouses then collaborated with the nurse interventionist to generate a plan to support patient goal achievement. For diet or physical activity goals, spouses were asked if they planned to make the same changes that the patient planned to make, and recommended support behaviors were tailored accordingly. For example, spouses might prepare foods differently or join the patient on walks if they intended to make the same changes or provide verbal reinforcement if they did not intend to make the same changes. Spouses indicated they planned to make the same behavior changes as patients in 96.7 % of calls in which the patient set dietary change goals and 65.4 % of calls in which the patient set physical activity goals. Assessments and measures In addition to baseline assessments, couples attended outcome assessments at 6 and 11 months, which were conducted by blinded research personnel. Because the intervention was designed to improve patient low-density lipoprotein cholesterol, spouse low-density lipoprotein cholesterol was not assessed. However, spouses completed the same self-report measures as patients, each assessing their own behaviors. For physical activity, the Community for Healthy Activities Model Program for Seniors (CHAMPS) questionnaire (Stewart et al., 2001) was administered to assess frequency (times per week) and duration (hours per week) of physical activity over the past 4 weeks. The type of physical activity we analyzed from the CHAMPS was moderate intensity physical activity (e.g., heavy housework, walking briskly) because it is stressed in US physical activity guidelines and is important for health maintenance (U.S. Department of Health and Human Services, 2008). For dietary intake, we examined total fat, saturated fat, dietary cholesterol, and fiber from the Brief Food Frequency
Questionnaire (FFQ) (Block, 2000) because they are targeted in the NCEP ATP III guidelines for cholesterol management (National Heart, Lung, and Blood Institute, 2001). We also examined total caloric intake (energy) and percentage of calories from total fat and saturated fat as the NCEP ATP III guidelines suggest weight loss if needed (National Heart, Lung, and Blood Institute, 2001). Because it was lengthy, the FFQ was completed at home and returned by mail. Data analysis Analyses were conducted on an intent-to-treat basis; spouses were analyzed in the arm to which the couple was randomized, using all data up to the 11-month follow-up or last available measurement prior to exclusion or dropout. Statistical analyses were performed using SAS for Windows (Version 9.2: SAS Institute, Cary, NC). Because frequency and duration of physical activity outcomes are count type data, we fit generalized linear mixed models using a negative-binomial distribution with a log link function (Diggle et al., 2002). Continuous dietary outcomes were log-transformed (except for percentage of calories from total fat and saturated fat) and analyzed with linear mixed models (Verbeke & Molenbergh, 2000). The primary predictors in all of these models included a common intercept and indicator variables for the 6- and 11-month time points and treatment group by time interaction. Our primary inference was on the treatment by 11-month follow-up time point indicator variable. The estimation procedure applied in the linear mixed modeling yields unbiased estimates of parameters when missing outcomes are related to either observed covariates or response variables (Little & Rubin, 2002).
Results At 11 months, there were no overall time effects for spouse self-reported moderate intensity physical activity (i.e., frequency, p = 0.56; duration, p = 0.41) or two of the dietary intake outcomes (i.e., percentage of calories from total fat, p = 0.88; percentage of calories from saturated fat, p = 0.40). For the remaining dietary intake outcomes, however, total calories (p = 0.002), total fat (p = 0.007), saturated fat (p = 0.004), cholesterol (p \ 0.0001), and fiber (p \ 0.001) were lower at 11 months compare to baseline. For spouses, the estimated rate of frequency and duration of moderate physical activity did not differ between intervention and usual care arms at 11 months (Table 2), and no differences were found in spouse dietary intake outcomes between intervention and usual care groups (Table 3). The estimated baseline rate of 6.5 times a week of moderate intensity physical activity with an estimated
123
J Behav Med Table 2 Model estimates of spouse self-reported physical activity by group and time point Measurement and study time point
Intervention
Usual care
Incidence rate ratio (95 % CI)
p value
Frequency of moderate intensity physical activity (times) per week Baseline
6.5
6.5
6 months
6.5
6.6
1.0 (0.7, 1.3)
0.89
11 months
6.3
6.1
1.0 (0.8, 1.4)
0.84
Duration of moderate intensity physical activity (hours) per week Baseline
4.6
6 months
4.1
4.6 4.1
1.0 (0.7, 1.4)
0.97
11 months
4.2
4.2
1.0 (0.7, 1.4)
0.92
Spouses missing frequency and duration of moderate intensity physical activity (measured by the Community for Healthy Activities Model Program for Seniors (CHAMPS) questionnaire): baseline = 7 spouses (3 intervention, 4 usual care (UC)), 6 months = 91 spouses (39 intervention, 52 UC), 11 months = 75 spouses (34 intervention, 41 UC) CI confidence interval
duration of 4.6 hours per week was similar across all times points in both arms for spouses. Unadjusted raw means and standard deviations of spouse dietary intake outcomes at each time point are shown in Table 3.
Discussion Based on evidence that partners have concordance in health behaviors, we anticipated that improvements seen in patient health behaviors during a lifestyle intervention might translate into spouse behavior change, particularly given that spouses had an active role in the intervention. However, this hypothesis was not supported. One possible explanation has to do with the goal and nature of the intervention. Baucom et al. (1998) distinguished between disorder-specific interventions, which focus on issues concerning the target condition and couple relationship functioning as it impacts the disorder and management, and partner-assisted interventions, in which the role of the spouse/partner is to encourage and reinforce patient behavior change and adherence to intervention protocol. This intervention was designed as a partnerassisted intervention. Because we did not want to create undue burden on spouses, we allowed for differing amounts and nature of spousal support. For example, we asked spouses if they intended to make the same changes because we recognized that spouses may not have the same ability or desire to change their health behaviors. Spouses indicated they planned to make the same behavior changes as patients in most calls in which the patient set dietary change goals and two-thirds of calls in which the patient set physical activity goals (Voils et al., 2013). These findings suggest that, if the goal is to improve spouse outcomes in addition to patient outcomes, interventions may need to focus on spouse health status, health
123
behaviors, and behavior change goals. This was done in a recent weight loss intervention; partners were encouraged to attend all group sessions, provided with a 10 % weight loss goal, and advised to make the same dietary and physical activity changes as the primary study participant (Gorin et al., 2013). The intervention resulted in significant weight loss and reduced fat intake in partners. Interventions may also be more effective if they included provision of reciprocal support to patients and spouses. Rather than teaching just one member of the dyad to provide support, social support strategies could be conveyed to both members to engender common understanding, as was done in a previous couple intervention in which spouse change was observed (Wing et al., 1991). Interventions may also be improved by requesting greater involvement of spouses. However, the potential benefits of this must be balanced by possible negative effects on recruitment, retention, and intervention adherence (Voils et al., 2011). Another possible explanation is that spouses may have been healthy and not required behavior change. Although partners tend to show concordance in health behaviors and coronary risk factors (Di Castelnuovo et al., 2009; Macken et al., 2000), we did not require spouse elevated low-density lipoprotein cholesterol for study eligibility. Spouses, however, did generally plan to make the same behavior changes as patients (Voils et al., 2013), which could reflect a desire to provide support and/or the need to improve their own health behaviors. As with any study, it is important to note the limitations. The physical activity and dietary intake outcomes were assessed via self-report and may be biased, leading to imprecise estimates. For example, frequency and duration of moderate intensity physical activity were high, which may reflect a self-report bias or study selection bias such that those spouses who enrolled in the study were more physically active with limited room for improvement. The
J Behav Med Table 3 Spouse dietary energy and nutrient intake by group and time point; observed means (SD) along with model estimated differences and associated 95 % confidence intervals and p values Observed mean (SD)a Macronutrient and study time point
Model estimatesb,c
Intervention
Usual care
Est. diff.
95 % CI LL, UL
p value
Baseline
1,209 (503)
1,227 (612)
6 months 11 months
971 (440) 1,008 (436)
1,011 (492) 1,058 (473)
-0.05 -0.03
-0.18, 0.08 -0.16, 0.10
0.46 0.65
Baseline
51.3 (23.2)
52.5 (27.0)
6 months
38.5 (20.6)
11 months
41.2 (20.3)
42.2 (21.9)
-0.12
-0.27, 0.03
0.12
45.2 (21.8)
-0.08
-0.22, 0.06
0.27
Baseline
38.6 (7.5)
38.6 (9.4)
6 months
35.3 (7.1)
11 months
36.5 (7.9)
38.1 (9.2)
-2.51
-4.80, -0.22
0.03
38.7 (7.8)
-1.47
-3.54, 0.60
0.16
Baseline
16.8 (8.2)
16.7 (9.2)
6 months
12.4 (6.8)
11 months
13.3 (6.6)
13.1 (6.9)
-0.11
-0.26, 0.05
0.17
14.2 (6.9)
-0.08
-0.22, 0.07
0.31
Baseline 6 months
12.7 (3.0) 11.3 (2.7)
12.1 (3.0) 11.8 (2.9)
-0.64
-1.40, 0.11
0.09
11 months
11.8 (2.8)
12.1 (2.6)
-0.41
-1.14, 0.32
0.26
Baseline
162.8 (83.7)
159.9 (116.4)
6 months
126.7 (79.2)
123.0 (80.0)
-0.06
-0.23, 0.10
0.46
11 months
125.5 (80.3)
117.1 (64.8)
-0.01
-0.18, 0.15
0.87
Baseline
12.7 (6.2)
13.4 (7.9)
6 months
10.7 (5.8)
11.8 (7.9)
0.00
-0.17, 0.18
0.96
11 months
11.3 (5.7)
11.7 (8.5)
0.03
-0.12, 0.19
0.69
Energy, kcal/db
Total fat, g/db
Total fat, %
c
Saturated fat, g/db
Saturated fat, %c
Cholesterol, mg/db
b
Fiber, g/d
SD standard deviation, Est. Diff. model estimated difference between intervention and usual care, CI confidence interval, LL lower limit, UL upper limit a
Unadjusted raw data with no imputations for missing data. Sample size (number missing) for baseline, 6 months, and 11 months are n = 95 (29), n = 71 (53), n = 78 (46), respectively, for the intervention group and n = 86 (41), n = 61 (66), n = 76 (51), respectively, for the usual care group. Brief Food Frequency Questionnaire (FFQ) data were missing for all time points for n = 28 (10 intervention group; 18 usual care group—included in above summary of individual time point sample size and missing) because the FFQ was completed at home and returned by mail b
Est. diff., 95 % CI, and p value for test of difference between intervention and usual care at 6 and 11 months from linear mixed models of logtransformed FFQ nutrient outcomes
c Est. diff., 95 % CI, and p value for test of difference between intervention and usual care at 6 and 11 months from linear mixed models of FFQ percentage of calories from nutrient outcomes
reported rates of physical activity may not be surprisingly high, however, when considering the activities the CHAMPS questionnaire measures (e.g., heavy housework, heavy gardening). Moreover, we used the brief version of the Food Frequency Questionnaire, which contains fewer food items than the full version. Despite the fact that the estimated means may be inflated due to these issues, the
relative differences in the self-report measures between groups and across time are informative. In addition, we had a significant amount of missing data for the Brief Food Frequency Questionnaire; no spouse data was available for any time point for some patients. We note that, in the main outcomes paper (Voils et al., 2013), no differences were found in characteristics and outcomes between patients
123
J Behav Med
who completed the Brief Food Frequency Questionnaire and those who did not. The results and generalizability are also limited by the high rates of refusal and focus on relatively healthy older male patients who received care from a single Veterans Affairs Medical Center and their female spouses. Our study has several strengths, most significant of which is measurement and examination of spouse outcomes, particularly health behaviors, which is not a common practice in the literature on couple interventions for chronic illness (Martire et al., 2010). These findings suggest several directions for future research. For example, studies might examine whether spouse outcomes are moderated by level of patient improvement. Studies might also examine the role of psychological intimacy and cooperation within couples. In qualitative interviews with a subset of patients and spouses who received this intervention, spouses of patients with same or better low-density lipoprotein cholesterol talked more about operating together, and spouses of patients with worse low-density lipoprotein cholesterol talked more about acting individually (Sperber et al., 2013). More research is needed to directly compare types of couple interventions (e.g., disorder-specific, partner-assisted) across various chronic diseases and their effects on patient and spouse health behaviors, physical health, and clinical outcomes. In sum, although couple interventions for chronic illness have significant small effects on some patient outcomes and in some cases are more efficacious than usual care and patient-only psychosocial interventions, more attention to spouse outcomes is needed. In addition, considering the type of couple intervention and the role of the spouse are crucial to design and evaluation and may help to maximize effects. Acknowledgments The first author was supported by a post-doctoral fellowship from the Department of Veterans Affairs (DVA), Office of Academic Affiliations, Health Services Research and Development Service (HSR&D; TPP-020). The original trial was supported by a grant from the DVA HSR&D (IIR 05-273, funding dates: 9/1/06-8/31/10; the ClinicalTrials.gov registration number is NCT00321789) and a Career Development Award from DVA HSR&D (MRP 04-216) to the last author. The third author was supported by a post-doctoral fellowship from the Agency for Healthcare Quality and Research (AHRQ; 5T32HS000079-15). The views expressed in this article are those of the authors and do not necessarily represent the position or policy of the DVA or the United States government. Conflict of interest
No author declares any conflict of interest.
References Baucom, D. H., Shoham, V., Mueser, K. T., Daiuto, A. D., & Stickle, T. R. (1998). Empirically supported couple and family interventions for marital distress and adult mental health problems. Journal of Consulting and Clinical Psychology, 66 (1), 53–88. Retrieved from http://www.apa.org/pubs/journals/ccp/index.aspx
123
Block, G. (2000). Block brief 2000 food frequency questionnaire (FFQ). Callahan, C. M., Unverzagt, F. W., Hui, S. L., Perkins, A. J., & Hendrie, H. C. (2002). Six-item screener to identify cognitive impairment among subjects for clinical research. Medical Care, 40, 771–781. doi:10.1097/01.MLR.0000024610.33213.C8 Di Castelnuovo, A., Quacquaruccio, G., Donati, M. B., de Gaetano, G., & Iacoviello, L. (2009). Spousal concordance for major coronary risk factors: A systematic review and meta-analysis. American Journal of Epidemiology, 169, 1–8. doi:10.1093/aje/kwn234 Diggle, P. J., Heagerty, P., Liang, K., & Zeger, S. L. (2002). Analysis of longitudinal data. New York, NY: Oxford University Press. Gorin, A. A., Raynor, H. A., Fava, J., Maguire, K., Robichaud, E., Trautvetter, J., et al. (2013). Randomized controlled trial of a comprehensive home environment-focused weight-loss program for adults. Health Psychology, 32, 128–137. doi:10.1037/a0026959 Little, R. J. A., & Rubin, D. B. (2002). Statistical analysis with missing data. Hoboken, NJ: Wiley. Macken, L. C., Yates, B., & Blancher, S. (2000). Concordance of risk factors in female spouses of male patients with coronary heart disease. Journal of Cardiopulmonary Rehabilitation, 20 (6), 361–368. Retrieved from http://journals.lww.com/jcrjournal/ pages/default.aspx Martire, L. M., Schulz, R., Helgeson, V. S., Small, B. J., & Saghafi, E. M. (2010). Review and meta-analysis of couple-oriented interventions for chronic illness. Annals of Behavioral Medicine, 40, 325–342. doi:10.1007/s12160-010-9216-2 National Heart, Lung, and Blood Institute. (2001). Third report of the National Cholesterol Education Program expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III): Executive summary (NIH Publication No. 01-3670). Retrieved from http://www.nhlbi.nih. gov/guidelines/cholesterol/atp3xsum.pdf Sperber, N. R., Sandelowski, M., & Voils, C. I. (2013). Spousal support in a behavior change intervention for cholesterol management. Patient Education and Counseling, 92 (1), 121–126. http://dx.doi.org/10.1016/j.pec.2013.02.015 Stewart, A. L., Mills, K. M., King, A. C., Haskell, W. L., Gillis, D., & Ritter, P. L. (2001). CHAMPS physical activity questionnaire for older adults: Outcomes for interventions. Medicine and Science in Sports and Exercise, 33, 1126–1141. U.S. Department of Health and Human Services. (2008). 2008 physical activity guidelines for Americans. Retrieved from http:// www.health.gov/paguidelines/pdf/paguide.pdf Verbeke, G., & Molenbergh, G. (2000). Linear mixed models for longitudinal data. New York, NY: Springer. Voils, C. I., Coffman, C. J., Yancy, W. S., Jr., Weinberger, M., Jeffreys, A. S., Datta, S., Bosworth, H. B. (2013). A randomized controlled trial to evaluate the effectiveness of CouPLES: A spouse-assisted lifestyle change intervention to improve lowdensity lipoprotein cholesterol. Preventive Medicine, 56 (1), 46–52. http://dx.doi.org/10.1016/j.ypmed.2012.11.001 Voils, C. I., Yancy, W. S, Jr, Kovac, S., Coffman, C. J., Weinberger, M., Oddone, E. Z., et al. (2009). Study protocol: Couples partnering for lipid enhancing strategies (CouPLES)—a randomized, controlled trial. Trials, 10, 10. doi:10.1186/1745-6215-10-10 Voils, C. I., Yancy, W. S, Jr, Weinberger, M., Bolton, J., Coffman, C. J., Jeffreys, A., et al. (2011). The trials and tribulations of enrolling couples in a randomized, controlled trial: A self-management program for hyperlipidemia as a model. Patient Education and Counseling, 84, 33–40. doi:10.1016/j.pec.2010.06.005 Wing, R. R., Marcus, M. D., Epstein, L. H., & Jawad, A. (1991). A ‘‘family-based’’ approach to the treatment of obese type II diabetic patients. Journal of Consulting and Clinical Psychology, 59 (1), 156–162. Retrieved from http://www.apa.org/pubs/ journals/ccp/index.aspx
Spouse health behavior outcomes from a randomized controlled trial of a spouse-assisted lifestyle change intervention to improve patient low-density lipoprotein cholesterol Heather A. King • Amy S. Jeffreys • Megan A. McVay Cynthia J. Coffman • Corrine I. Voils
•
Received: April 28, 2013 / Accepted: February 10, 2014 Ó Springer Science+Business Media New York (outside the USA) 2014
Abstract This study evaluated spouse health behavior outcomes from a randomized controlled trial of a spouseassisted lifestyle intervention to reduce patient low-density lipoprotein cholesterol and improve patient health behaviors. Participants were 251 spouses of patients from the Durham Veterans Affairs Medical Center randomized to intervention or usual care. The intervention comprised 9 monthly telephone calls to patients and spouses. Outcomes were assessed at baseline, 6 and 11 months. At 11 months, there were no differences in spouse outcomes between intervention and usual care groups for moderate intensity physical activity (i.e., frequency, duration) or dietary intake (i.e., total calories, total fat, percentage of calories from total fat, saturated fat, percentage of calories from saturated fat, cholesterol, fiber). To improve spouse outcomes, couple interventions may need to include spouse behavior change goals and reciprocal support between patients and spouses and consider the need for improvement in spouse outcomes. Keywords Spouse Couple Dietary intake Physical activity Intervention Social support H. A. King (&) A. S. Jeffreys M. A. McVay C. J. Coffman C. I. Voils Center for Health Services Research in Primary Care, Durham Veterans Affairs Medical Center, HSR&D (152), 508 Fulton St., Durham, NC 27705, USA e-mail: [email protected] M. A. McVay C. I. Voils Department of Medicine, Duke University Medical Center, Durham, NC, USA C. J. Coffman Department of Biostatistics and Bioinformatics, Duke University Medical Center, Durham, NC, USA
Introduction Couple interventions (i.e., both patient and partner are active participants) for chronic illness have significant small effects on some patient outcomes and in some cases are more efficacious than both usual care and psychosocial interventions involving only the patient (Martire et al., 2010). Unclear is whether such interventions potentially benefit only patients or also their spouses. Observational studies indicate significant concordance between patients and their spouses for behavioral risk factors such as smoking, dietary intake, and physical activity (Di Castelnuovo et al., 2009; Macken et al., 2000). Therefore, if spouse health behavior outcomes were also positively affected by couple interventions, then such interventions may be particularly efficacious and efficient in terms of time and cost. In a recent review and meta-analysis of couple interventions for chronic illness conducted by Martire et al. (2010), only half of the reviewed studies examined effects on spouse outcomes. Those that did had various methodological limitations and focused primarily on psychosocial outcomes. Only one reviewed study assessed spouse health behaviors (Wing et al., 1991): this study on obese Type 2 diabetes patients and their overweight spouses found greater spouse weight loss and changes in eating behavior strategies among spouses in a couple-focused behavioral weight loss program compared to a patient-only program. In addition, a recent trial evaluated the effectiveness of a couple intervention, specifically a spouse-assisted lifestyle intervention, to reduce patient cholesterol levels and improve patient health behaviors such as dietary intake and physical activity (CouPLES; Voils et al., 2009). In a partnerassisted lifestyle intervention, the role of the partner or spouse is to support patient behavior change and intervention adherence (Baucom et al., 1998). Among patients, no
123
J Behav Med
significant differences between intervention and usual care groups were found for low-density lipoprotein cholesterol or physical activity at the final assessment (11 months). However, patient caloric intake, total and saturated fat intake, and percentage of calories from total fat were lower in the intervention than in the usual care control group (Voils et al., 2013). In this paper, we expand on these previous analyses by examining a distinct research question: whether health behaviors of physical activity and dietary intake improved in spouses as a function of their involvement in this spouse-assisted intervention.
National Heart, Lung, and Blood Institute, 2001) and information on providing spousal support; baseline measurements were collected; and then couples were randomized to usual care or the intervention, stratified by patient race (White vs. non-White) and patient coronary heart disease risk level (low/moderate vs. high). Patients in the usual care group received clinical management of lipid disorders by their providers. Neither patients nor spouses received further contact from study staff except for outcome assessment visits. The trial was approved by ethics committees and conducted at the Durham Veterans Affairs Medical Center. Patient recruitment and enrollment
Method Participants Detailed methods and the CONSORT flow diagram outlining enrollment and intervention allocation have been reported elsewhere (Voils et al., 2009, 2011, 2013). Ultimately, 255 couples were randomized (127 to the intervention, 128 to usual care) following patient baseline assessments. In four cases, patients attended baseline appointments alone, and spouses never returned for assessments; therefore, data from 251 spouses (124 intervention, 127 usual care) were available for analyses. Assessments were completed for 248 spouses at baseline (122 intervention, 126 usual care), 164 at the 6-month follow-up (88 intervention, 76 usual care), and 176 (90 intervention, 86 usual care) at the 11-month follow-up. Spouse baseline characteristics are provided in Table 1.
Briefly, couples received a recruitment letter prior to a patient primary care visit, followed by a screening telephone call. To be eligible, patients had to be married and have elevated low-density lipoprotein cholesterol. Patient and spouse exclusion criteria assessed during the screening call included inconsistent access to telephone; impaired cognition (Callahan et al., 2002) or hearing; health problem that precluded participation; and resident of nursing home or receiving long-term care. Couples passing the telephone screening were scheduled for baseline assessments, where written informed consent was obtained and self-report measures were completed. Patients and spouses initially received $10 each for each outcome visit; later, they received $20 to address increases in transportation expenses. Couples intervention
Procedure Setting and design In this two-group, randomized controlled trial, all couples received print educational materials containing guidelineconcordant recommendations (NCEP ATP III guidelines;
The intervention was informed by social cognitive theory. It was delivered via phone by a nurse interventionist using a custom software package to maximize intervention fidelity. Delivering an intervention by phone is less expensive and can reach more patients compared to faceto-face intervention delivery. Intervention telephone calls
Table 1 Baseline characteristics of enrolled spouses Demographic variable
All (N = 251)
Intervention (N = 124)
Usual care (N = 127)
Age, M (SD)
58.8 (12.0)
59.4 (12.3)
58.2 (11.6)
White (%) Female (%)
64.1 94.0
66.4 91.1
61.9 96.9
High school grad or less (%)
26.6
25.4
27.8
Full-time employment (%)
34.3
27.1
41.3
Weekly frequency of moderate intensity physical activity, median (IQR)
7 (2, 12)
6 (2, 11)
7 (3, 12)
Weekly duration of moderate intensity physical activity (hours), median (IQR)
3.75 (1.0, 9.5)
3.75 (1.0, 7.5)
4 (1.5, 10.0)
3 missing for age, race, education, and employment status (2 intervention, 1 usual care). 7 missing for weekly frequency and duration of moderate intensity physical activity (3 intervention, 4 usual care), Participants with missing data are excluded from percentage calculations M mean, SD standard deviation, IQR interquartile range
123
J Behav Med
were delivered monthly first to patients and then to spouses. In month 1, patients and spouses received information about hypercholesterolemia and an overview of self-management principles. Spouses also received an orientation on strategies to support patient goal achievement and were encouraged to provide the patients with support and praise, offer help, reward desired health behaviors rather than criticizing poor health behaviors, focus on patients’ goals instead of what spouses thought patients should be doing, and help patients establish non-food related rewards for progress toward their goals. In months 2–5 and 7–10, calls with patients focused on goal-setting and problem-solving. In all calls, patients created behavioral goals and plans, most often in relation to diet or physical activity. Following patient calls, spouses were informed of patients’ successes, goals, and action plans for the upcoming month and were asked to record them. Spouses then collaborated with the nurse interventionist to generate a plan to support patient goal achievement. For diet or physical activity goals, spouses were asked if they planned to make the same changes that the patient planned to make, and recommended support behaviors were tailored accordingly. For example, spouses might prepare foods differently or join the patient on walks if they intended to make the same changes or provide verbal reinforcement if they did not intend to make the same changes. Spouses indicated they planned to make the same behavior changes as patients in 96.7 % of calls in which the patient set dietary change goals and 65.4 % of calls in which the patient set physical activity goals. Assessments and measures In addition to baseline assessments, couples attended outcome assessments at 6 and 11 months, which were conducted by blinded research personnel. Because the intervention was designed to improve patient low-density lipoprotein cholesterol, spouse low-density lipoprotein cholesterol was not assessed. However, spouses completed the same self-report measures as patients, each assessing their own behaviors. For physical activity, the Community for Healthy Activities Model Program for Seniors (CHAMPS) questionnaire (Stewart et al., 2001) was administered to assess frequency (times per week) and duration (hours per week) of physical activity over the past 4 weeks. The type of physical activity we analyzed from the CHAMPS was moderate intensity physical activity (e.g., heavy housework, walking briskly) because it is stressed in US physical activity guidelines and is important for health maintenance (U.S. Department of Health and Human Services, 2008). For dietary intake, we examined total fat, saturated fat, dietary cholesterol, and fiber from the Brief Food Frequency
Questionnaire (FFQ) (Block, 2000) because they are targeted in the NCEP ATP III guidelines for cholesterol management (National Heart, Lung, and Blood Institute, 2001). We also examined total caloric intake (energy) and percentage of calories from total fat and saturated fat as the NCEP ATP III guidelines suggest weight loss if needed (National Heart, Lung, and Blood Institute, 2001). Because it was lengthy, the FFQ was completed at home and returned by mail. Data analysis Analyses were conducted on an intent-to-treat basis; spouses were analyzed in the arm to which the couple was randomized, using all data up to the 11-month follow-up or last available measurement prior to exclusion or dropout. Statistical analyses were performed using SAS for Windows (Version 9.2: SAS Institute, Cary, NC). Because frequency and duration of physical activity outcomes are count type data, we fit generalized linear mixed models using a negative-binomial distribution with a log link function (Diggle et al., 2002). Continuous dietary outcomes were log-transformed (except for percentage of calories from total fat and saturated fat) and analyzed with linear mixed models (Verbeke & Molenbergh, 2000). The primary predictors in all of these models included a common intercept and indicator variables for the 6- and 11-month time points and treatment group by time interaction. Our primary inference was on the treatment by 11-month follow-up time point indicator variable. The estimation procedure applied in the linear mixed modeling yields unbiased estimates of parameters when missing outcomes are related to either observed covariates or response variables (Little & Rubin, 2002).
Results At 11 months, there were no overall time effects for spouse self-reported moderate intensity physical activity (i.e., frequency, p = 0.56; duration, p = 0.41) or two of the dietary intake outcomes (i.e., percentage of calories from total fat, p = 0.88; percentage of calories from saturated fat, p = 0.40). For the remaining dietary intake outcomes, however, total calories (p = 0.002), total fat (p = 0.007), saturated fat (p = 0.004), cholesterol (p \ 0.0001), and fiber (p \ 0.001) were lower at 11 months compare to baseline. For spouses, the estimated rate of frequency and duration of moderate physical activity did not differ between intervention and usual care arms at 11 months (Table 2), and no differences were found in spouse dietary intake outcomes between intervention and usual care groups (Table 3). The estimated baseline rate of 6.5 times a week of moderate intensity physical activity with an estimated
123
J Behav Med Table 2 Model estimates of spouse self-reported physical activity by group and time point Measurement and study time point
Intervention
Usual care
Incidence rate ratio (95 % CI)
p value
Frequency of moderate intensity physical activity (times) per week Baseline
6.5
6.5
6 months
6.5
6.6
1.0 (0.7, 1.3)
0.89
11 months
6.3
6.1
1.0 (0.8, 1.4)
0.84
Duration of moderate intensity physical activity (hours) per week Baseline
4.6
6 months
4.1
4.6 4.1
1.0 (0.7, 1.4)
0.97
11 months
4.2
4.2
1.0 (0.7, 1.4)
0.92
Spouses missing frequency and duration of moderate intensity physical activity (measured by the Community for Healthy Activities Model Program for Seniors (CHAMPS) questionnaire): baseline = 7 spouses (3 intervention, 4 usual care (UC)), 6 months = 91 spouses (39 intervention, 52 UC), 11 months = 75 spouses (34 intervention, 41 UC) CI confidence interval
duration of 4.6 hours per week was similar across all times points in both arms for spouses. Unadjusted raw means and standard deviations of spouse dietary intake outcomes at each time point are shown in Table 3.
Discussion Based on evidence that partners have concordance in health behaviors, we anticipated that improvements seen in patient health behaviors during a lifestyle intervention might translate into spouse behavior change, particularly given that spouses had an active role in the intervention. However, this hypothesis was not supported. One possible explanation has to do with the goal and nature of the intervention. Baucom et al. (1998) distinguished between disorder-specific interventions, which focus on issues concerning the target condition and couple relationship functioning as it impacts the disorder and management, and partner-assisted interventions, in which the role of the spouse/partner is to encourage and reinforce patient behavior change and adherence to intervention protocol. This intervention was designed as a partnerassisted intervention. Because we did not want to create undue burden on spouses, we allowed for differing amounts and nature of spousal support. For example, we asked spouses if they intended to make the same changes because we recognized that spouses may not have the same ability or desire to change their health behaviors. Spouses indicated they planned to make the same behavior changes as patients in most calls in which the patient set dietary change goals and two-thirds of calls in which the patient set physical activity goals (Voils et al., 2013). These findings suggest that, if the goal is to improve spouse outcomes in addition to patient outcomes, interventions may need to focus on spouse health status, health
123
behaviors, and behavior change goals. This was done in a recent weight loss intervention; partners were encouraged to attend all group sessions, provided with a 10 % weight loss goal, and advised to make the same dietary and physical activity changes as the primary study participant (Gorin et al., 2013). The intervention resulted in significant weight loss and reduced fat intake in partners. Interventions may also be more effective if they included provision of reciprocal support to patients and spouses. Rather than teaching just one member of the dyad to provide support, social support strategies could be conveyed to both members to engender common understanding, as was done in a previous couple intervention in which spouse change was observed (Wing et al., 1991). Interventions may also be improved by requesting greater involvement of spouses. However, the potential benefits of this must be balanced by possible negative effects on recruitment, retention, and intervention adherence (Voils et al., 2011). Another possible explanation is that spouses may have been healthy and not required behavior change. Although partners tend to show concordance in health behaviors and coronary risk factors (Di Castelnuovo et al., 2009; Macken et al., 2000), we did not require spouse elevated low-density lipoprotein cholesterol for study eligibility. Spouses, however, did generally plan to make the same behavior changes as patients (Voils et al., 2013), which could reflect a desire to provide support and/or the need to improve their own health behaviors. As with any study, it is important to note the limitations. The physical activity and dietary intake outcomes were assessed via self-report and may be biased, leading to imprecise estimates. For example, frequency and duration of moderate intensity physical activity were high, which may reflect a self-report bias or study selection bias such that those spouses who enrolled in the study were more physically active with limited room for improvement. The
J Behav Med Table 3 Spouse dietary energy and nutrient intake by group and time point; observed means (SD) along with model estimated differences and associated 95 % confidence intervals and p values Observed mean (SD)a Macronutrient and study time point
Model estimatesb,c
Intervention
Usual care
Est. diff.
95 % CI LL, UL
p value
Baseline
1,209 (503)
1,227 (612)
6 months 11 months
971 (440) 1,008 (436)
1,011 (492) 1,058 (473)
-0.05 -0.03
-0.18, 0.08 -0.16, 0.10
0.46 0.65
Baseline
51.3 (23.2)
52.5 (27.0)
6 months
38.5 (20.6)
11 months
41.2 (20.3)
42.2 (21.9)
-0.12
-0.27, 0.03
0.12
45.2 (21.8)
-0.08
-0.22, 0.06
0.27
Baseline
38.6 (7.5)
38.6 (9.4)
6 months
35.3 (7.1)
11 months
36.5 (7.9)
38.1 (9.2)
-2.51
-4.80, -0.22
0.03
38.7 (7.8)
-1.47
-3.54, 0.60
0.16
Baseline
16.8 (8.2)
16.7 (9.2)
6 months
12.4 (6.8)
11 months
13.3 (6.6)
13.1 (6.9)
-0.11
-0.26, 0.05
0.17
14.2 (6.9)
-0.08
-0.22, 0.07
0.31
Baseline 6 months
12.7 (3.0) 11.3 (2.7)
12.1 (3.0) 11.8 (2.9)
-0.64
-1.40, 0.11
0.09
11 months
11.8 (2.8)
12.1 (2.6)
-0.41
-1.14, 0.32
0.26
Baseline
162.8 (83.7)
159.9 (116.4)
6 months
126.7 (79.2)
123.0 (80.0)
-0.06
-0.23, 0.10
0.46
11 months
125.5 (80.3)
117.1 (64.8)
-0.01
-0.18, 0.15
0.87
Baseline
12.7 (6.2)
13.4 (7.9)
6 months
10.7 (5.8)
11.8 (7.9)
0.00
-0.17, 0.18
0.96
11 months
11.3 (5.7)
11.7 (8.5)
0.03
-0.12, 0.19
0.69
Energy, kcal/db
Total fat, g/db
Total fat, %
c
Saturated fat, g/db
Saturated fat, %c
Cholesterol, mg/db
b
Fiber, g/d
SD standard deviation, Est. Diff. model estimated difference between intervention and usual care, CI confidence interval, LL lower limit, UL upper limit a
Unadjusted raw data with no imputations for missing data. Sample size (number missing) for baseline, 6 months, and 11 months are n = 95 (29), n = 71 (53), n = 78 (46), respectively, for the intervention group and n = 86 (41), n = 61 (66), n = 76 (51), respectively, for the usual care group. Brief Food Frequency Questionnaire (FFQ) data were missing for all time points for n = 28 (10 intervention group; 18 usual care group—included in above summary of individual time point sample size and missing) because the FFQ was completed at home and returned by mail b
Est. diff., 95 % CI, and p value for test of difference between intervention and usual care at 6 and 11 months from linear mixed models of logtransformed FFQ nutrient outcomes
c Est. diff., 95 % CI, and p value for test of difference between intervention and usual care at 6 and 11 months from linear mixed models of FFQ percentage of calories from nutrient outcomes
reported rates of physical activity may not be surprisingly high, however, when considering the activities the CHAMPS questionnaire measures (e.g., heavy housework, heavy gardening). Moreover, we used the brief version of the Food Frequency Questionnaire, which contains fewer food items than the full version. Despite the fact that the estimated means may be inflated due to these issues, the
relative differences in the self-report measures between groups and across time are informative. In addition, we had a significant amount of missing data for the Brief Food Frequency Questionnaire; no spouse data was available for any time point for some patients. We note that, in the main outcomes paper (Voils et al., 2013), no differences were found in characteristics and outcomes between patients
123
J Behav Med
who completed the Brief Food Frequency Questionnaire and those who did not. The results and generalizability are also limited by the high rates of refusal and focus on relatively healthy older male patients who received care from a single Veterans Affairs Medical Center and their female spouses. Our study has several strengths, most significant of which is measurement and examination of spouse outcomes, particularly health behaviors, which is not a common practice in the literature on couple interventions for chronic illness (Martire et al., 2010). These findings suggest several directions for future research. For example, studies might examine whether spouse outcomes are moderated by level of patient improvement. Studies might also examine the role of psychological intimacy and cooperation within couples. In qualitative interviews with a subset of patients and spouses who received this intervention, spouses of patients with same or better low-density lipoprotein cholesterol talked more about operating together, and spouses of patients with worse low-density lipoprotein cholesterol talked more about acting individually (Sperber et al., 2013). More research is needed to directly compare types of couple interventions (e.g., disorder-specific, partner-assisted) across various chronic diseases and their effects on patient and spouse health behaviors, physical health, and clinical outcomes. In sum, although couple interventions for chronic illness have significant small effects on some patient outcomes and in some cases are more efficacious than usual care and patient-only psychosocial interventions, more attention to spouse outcomes is needed. In addition, considering the type of couple intervention and the role of the spouse are crucial to design and evaluation and may help to maximize effects. Acknowledgments The first author was supported by a post-doctoral fellowship from the Department of Veterans Affairs (DVA), Office of Academic Affiliations, Health Services Research and Development Service (HSR&D; TPP-020). The original trial was supported by a grant from the DVA HSR&D (IIR 05-273, funding dates: 9/1/06-8/31/10; the ClinicalTrials.gov registration number is NCT00321789) and a Career Development Award from DVA HSR&D (MRP 04-216) to the last author. The third author was supported by a post-doctoral fellowship from the Agency for Healthcare Quality and Research (AHRQ; 5T32HS000079-15). The views expressed in this article are those of the authors and do not necessarily represent the position or policy of the DVA or the United States government. Conflict of interest
No author declares any conflict of interest.
References Baucom, D. H., Shoham, V., Mueser, K. T., Daiuto, A. D., & Stickle, T. R. (1998). Empirically supported couple and family interventions for marital distress and adult mental health problems. Journal of Consulting and Clinical Psychology, 66 (1), 53–88. Retrieved from http://www.apa.org/pubs/journals/ccp/index.aspx
123
Block, G. (2000). Block brief 2000 food frequency questionnaire (FFQ). Callahan, C. M., Unverzagt, F. W., Hui, S. L., Perkins, A. J., & Hendrie, H. C. (2002). Six-item screener to identify cognitive impairment among subjects for clinical research. Medical Care, 40, 771–781. doi:10.1097/01.MLR.0000024610.33213.C8 Di Castelnuovo, A., Quacquaruccio, G., Donati, M. B., de Gaetano, G., & Iacoviello, L. (2009). Spousal concordance for major coronary risk factors: A systematic review and meta-analysis. American Journal of Epidemiology, 169, 1–8. doi:10.1093/aje/kwn234 Diggle, P. J., Heagerty, P., Liang, K., & Zeger, S. L. (2002). Analysis of longitudinal data. New York, NY: Oxford University Press. Gorin, A. A., Raynor, H. A., Fava, J., Maguire, K., Robichaud, E., Trautvetter, J., et al. (2013). Randomized controlled trial of a comprehensive home environment-focused weight-loss program for adults. Health Psychology, 32, 128–137. doi:10.1037/a0026959 Little, R. J. A., & Rubin, D. B. (2002). Statistical analysis with missing data. Hoboken, NJ: Wiley. Macken, L. C., Yates, B., & Blancher, S. (2000). Concordance of risk factors in female spouses of male patients with coronary heart disease. Journal of Cardiopulmonary Rehabilitation, 20 (6), 361–368. Retrieved from http://journals.lww.com/jcrjournal/ pages/default.aspx Martire, L. M., Schulz, R., Helgeson, V. S., Small, B. J., & Saghafi, E. M. (2010). Review and meta-analysis of couple-oriented interventions for chronic illness. Annals of Behavioral Medicine, 40, 325–342. doi:10.1007/s12160-010-9216-2 National Heart, Lung, and Blood Institute. (2001). Third report of the National Cholesterol Education Program expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III): Executive summary (NIH Publication No. 01-3670). Retrieved from http://www.nhlbi.nih. gov/guidelines/cholesterol/atp3xsum.pdf Sperber, N. R., Sandelowski, M., & Voils, C. I. (2013). Spousal support in a behavior change intervention for cholesterol management. Patient Education and Counseling, 92 (1), 121–126. http://dx.doi.org/10.1016/j.pec.2013.02.015 Stewart, A. L., Mills, K. M., King, A. C., Haskell, W. L., Gillis, D., & Ritter, P. L. (2001). CHAMPS physical activity questionnaire for older adults: Outcomes for interventions. Medicine and Science in Sports and Exercise, 33, 1126–1141. U.S. Department of Health and Human Services. (2008). 2008 physical activity guidelines for Americans. Retrieved from http:// www.health.gov/paguidelines/pdf/paguide.pdf Verbeke, G., & Molenbergh, G. (2000). Linear mixed models for longitudinal data. New York, NY: Springer. Voils, C. I., Coffman, C. J., Yancy, W. S., Jr., Weinberger, M., Jeffreys, A. S., Datta, S., Bosworth, H. B. (2013). A randomized controlled trial to evaluate the effectiveness of CouPLES: A spouse-assisted lifestyle change intervention to improve lowdensity lipoprotein cholesterol. Preventive Medicine, 56 (1), 46–52. http://dx.doi.org/10.1016/j.ypmed.2012.11.001 Voils, C. I., Yancy, W. S, Jr, Kovac, S., Coffman, C. J., Weinberger, M., Oddone, E. Z., et al. (2009). Study protocol: Couples partnering for lipid enhancing strategies (CouPLES)—a randomized, controlled trial. Trials, 10, 10. doi:10.1186/1745-6215-10-10 Voils, C. I., Yancy, W. S, Jr, Weinberger, M., Bolton, J., Coffman, C. J., Jeffreys, A., et al. (2011). The trials and tribulations of enrolling couples in a randomized, controlled trial: A self-management program for hyperlipidemia as a model. Patient Education and Counseling, 84, 33–40. doi:10.1016/j.pec.2010.06.005 Wing, R. R., Marcus, M. D., Epstein, L. H., & Jawad, A. (1991). A ‘‘family-based’’ approach to the treatment of obese type II diabetic patients. Journal of Consulting and Clinical Psychology, 59 (1), 156–162. Retrieved from http://www.apa.org/pubs/ journals/ccp/index.aspx
