The Health Action Process Approach as a Motivational Model of Dietary Self-Management for People With Multiple Sclerosis

Abstract

The main objective of this study was to evaluate the health action process approach (HAPA) as a motivational model for dietary self-management for people with multiple sclerosis (MS). Quantitative descriptive research design using path analysis was used. Participants were 209 individuals with MS recruited from the National MS Society and a neurology clinic at a university teaching hospital in the Midwest. Outcome was measured by the Healthy Dietary Stages of Change Instrument, along with 10 predictor measures. The HAPA dietary self-management model fit the data relatively well (goodness-of-fit index = .95, normed fit index = .90, comparative fit index = .95, and root mean square error approximation = 0.07) explaining 15% of the variance in dietary self-management behavior. Recovery self-efficacy and action and coping planning directly contributed to the prediction of dietary health behaviors. Action self-efficacy, outcome expectancy, risk perception, and social support influenced intention, and the relationship between intention and dietary health behaviors is mediated by action and coping planning. Action self-efficacy, maintenance self-efficacy, and recovery self-efficacy directly or indirectly affected dietary health behaviors. Empirical support was found for the HAPA model of dietary self-management for people with MS. The HAPA model can be used to design behavioral health promotion interventions for people with disabilities in vocational rehabilitation.

Keywords

multiple sclerosis health promotion self-management dietary behavior

There are about 400,000 individuals with multiple sclerosis (MS) in the United States and 200 new patients per week, most of whom were diagnosed between ages 20 and 50 years. MS is a progressive disease of the central nervous system (National Multiple Sclerosis Society [NMSS], 2011). MS degrades mobility (e.g., fatigue, difficulty in walking and balancing) and impacts on mental health (e.g., depression and cognitive impairment; Fleming, 2002). Decreased mobility, fatigue, and depression very likely cause inactivity and decline in work capability. Unemployment has been a common outcome after the onset of MS. The unemployment rate has been reported as significantly higher than for people without disabilities (Khan, Ng, & Turner-Stokes, 2009). A sedentary lifestyle puts people with MS at risk of becoming overweight or obese, which is associated with secondary health conditions such as cardiovascular diseases, diabetes mellitus, osteoarthritis, and depression (Centers for Disease Control and Prevention [CDC], 2011; Physical Activity Guidelines Advisory Committee, 2008; Stuifbergen & Roberts, 1997). These secondary conditions negatively impact work performance and are inversely correlated with obtaining and retaining employment (Ipsen, Seekins, & Ravesloot, 2010).

Regular exercise has been found to reduce the negative impacts of secondary conditions and improve or maintain daily functional performance (e.g., work; Stuifbergen, Becker, Blozis, Timmerman, & Kullberg, 2003). In addition to regular exercise and physical activity, diet and nutrition can be complementary health promotion strategies for people with MS (Bombardier, Wadhwani, & LaRotonda, 2005). For people with significant ambulatory impairments, dietary management is very important for controlling secondary conditions such as being overweight and obese (U.S. Department of Agriculture [USDA] and U.S. Department of Health and Human Services, 2010). Dietary self-management is needed and helpful for people with MS who need to monitor their symptoms and secondary conditions (e.g., gaining weight, hypertension, and high blood sugar); however, the importance of dietary management for people with MS has not been recognized or studied extensively (Habek, Hojsak, & Brinar, 2010; Schwarz & Leweling, 2005).

Slawta et al. (2003) recruited 123 individuals with MS living in the community and surveyed their blood lipids, lipoprotein-cholesterol, and glucose; body composition; and eating habits, dietary composition, and so on. They found that the average body mass index, waist circumference, and total body fat of participants exceeded the recommended levels by the World Health Organization, namely, 50% overweight, 25% obese, 68% over a body fat percentage of 35%, 52% waist circumference over 80 cm, and 16% waist circumference over 96.5 cm. Intake of fat was over the suggested level by the USDA. Slightly less than half met the recommended intake of vegetables and fruits.

A large-scale survey from the Veterans Health Administration by Khurana and associates (2009) investigated the prevalence of being overweight and obese among 4,703 veterans (3,763 were men and 576 were women) with MS. They found 42.8% of men with MS were overweight and 21.2% were obese, whereas 28% of women were overweight and 25% were obese. Interestingly, depression had a strong trend of increasing the odds of being overweight or obese. Hewson, Phillips, Simpson, Drury, and Crawford (1984) investigated 142 people with MS about their weight, height, and 7-day dietary intake, and found that 40% of females and 44% of males were overweight. On the other hand, some people with MS are underweight and malnourished due to their decreased mobility (e.g., manipulating food and fluid in mouth, dysphasia, and relying on convenience food), fatigue (e.g., quickly tire when chewing), impaired cognition (e.g., poor insight to care regime and a weight problem), and side effects of drugs (e.g., loss of appetite because of “Baclofen” causing nausea and sedation; Payne, 2001). These conditions can cause difficulties in shopping for food, choosing food, preparing meals, cooking, and eating. Payne (2001) suggested that every 6 to 12 months, people with MS should be assessed for nutrition status and be given updated and appropriate dietary guidance. Given such findings, it would appear that people with MS could benefit from dietary health management to control their weight and body fat.

Unfortunately, research studying dietary behavior or dietary interventions specific for people with MS is limited. Health education interventions for improving fruit and vegetable consumption and low-fat intake behaviors that are based on social cognitive theory, the transtheoretical model, and theory of planned behavior have been undertaken for people with diabetes and college students. These theories have been useful in identifying potentially important and modifiable healthy dietary determinants for people without disabilities. However, there is a lack of research using these theories individually or collectively to study health-promoting behaviors of people with disabilities, such as MS. There is a need for an integrated health promotion model studying the dietary management behavior of people with MS specifically. In addition, knowing how to independently self-manage their diet is one way to empower people with disabilities to pursue better health and quality of life. In the current study, we are interested in validating Schwarzer’s health action process approach (HAPA) as a dietary self-management model for people with MS. The model, if validated, can provide invaluable information for developing health promotion interventions for people with MS and further empower them to self-manage their health so that they can increase their work capabilities and pursue a better quality of life.

The HAPA as a Health Promotion Model

The HAPA integrates the social cognitive theory (Bandura, 1997), the planned behavior theory (Ajzen & Fishbein, 1980), and the stages of change (SOC) theory (Prochaska, DiClement, & Norcross, 1992) to predict engagement in health behavior. This model appears to have great potential as a motivational model for dietary self-management for people with MS.

The HAPA model mainly builds on social cognitive theory, which is composed of knowledge of health risks and benefits, perceived self-efficacy, outcome expectations, goals, and perceived facilitators and obstacles (Bandura, 1997). Specifically, the HAPA model indicates that the adoption, initiation, and maintenance of health behaviors can be conceptualized as a process composed of motivation and volition phases. Health self-efficacy, outcome expectancies, and risk perception all play into motivation. Conversely, volition includes planning, action, and maintenance behaviors, with perceived self-efficacy and other cognitions playing a crucial role (Bandura, 1997). In the HAPA model, both action self-efficacy and outcome expectancies are considered primary variables for motivating change. However, if a person has no experience with a behavior, outcome expectancies would have a stronger direct influence on intention than self-efficacy. The effect of risk perception on intention is considered the weakest among the three cognitions (Bandura, 1997; Schwarzer, 2008). Risk perception can lead a person to contemplate making health behavior change in the motivation phase but not at the volition phase. Similarly, health outcome expectancies (i.e., balancing the pros and cons and the consequences of making behavioral changes) are important in the motivation phase, but the effects of health outcome expectancies on intention will be significantly reduced once the decision is made. The volition phase can be described along three levels: cognitive, behavioral, and situational. The cognitive level refers to self-regulatory processes that mediate between the intentions and the actions. This volitional process contains action plans and action control and is strongly influenced by perceived self-efficacy and perceived situational barriers. Figure 1 provides a graphical depiction of the HAPA health promotion model.

Figure 1.

The HAPA model of health promotion

Schwarzer (2008) validated HAPA as a health promotion model in a series of studies. Results indicated that HAPA constructs are predictive of exercise, drinking, and dietary behavior. Renner et al. (2008) applied the HAPA model on a diet low in fat and high in vitamins of 697 South Korean men and women recruited from universities, homes for the elderly, clerical institutions, and police departments. They included three predictors of intention measured at baseline (i.e., action self-efficacy, outcome expectancy, and objective health risk), and 6 months later, the other three predictors of behavior were measured (i.e., intention, coping self-efficacy, and planning). For all 697 participants, the HAPA model fit the data well: χ² = 276, df = 188, χ²/df = 1.47, comparative fit index (CFI) = .97, and root mean square error approximation (RMSEA) = 0.03. Furthermore, they examined whether there are sex differences in the structure of social cognitive variables. They found structural differences in the prediction patterns between men (33% of nutrition behavior variance explained by the model, 25% planning, 9% intention) and women (40% of nutrition behavior variance explained by the model, 38% planning, 28% intention). Coping self-efficacy and planning had strong predictive coefficients in each men and women’s model.

Schwarzer and Renner (2000) recruited 580 residents of Berlin. They measured residents’ risk perception, outcome expectancies, action self-efficacy, and intention. Six months later, they followed up residents and measured their coping self-efficacy and dietary fat intake as well as high-fiber dietary intake. Overall, the model fit the data well: χ²(197) = 348.93, p < .001, goodness-of-fit index (GFI) = .98, and RMSEA = 0.038. Interestingly, the better predictor of low fat was intention, whereas coping self-efficacy was the better predictor for high fiber. Forty-eight percent of variance in the low-fat diet and 33% in the high-fiber diet were accounted for by intention and coping self-efficacy. Furthermore, they tested modeling by gender and grouped samples into slim and heavy groups. Each model fit well and without significant differences.

The HAPA as a Motivational Model for Dietary Self-Management

The primary purpose of the present study was to formulate and test a HAPA motivation model for dietary self- management for people with MS. The HAPA model posits that the adoption, initiation, and maintenance of health behaviors (e.g., dietary self-management) can be conceptualized as a process comprising two stages: the motivation phase and the volition phase (Schwarzer, 1992, 2008). Action self-efficacy, outcome expectancies, and risk perception contribute to intention formation in the motivation phase, and action and coping planning, maintenance self-efficacy, and recovery self-efficacy are used to bridge the gap between intention and action in the volition phase. Severity of disability (Lynch & Chiu, 2009; Ravesloot et al., 2011), social support (Anderson, Winett, & Wojcik, 2007; Ford, Ahluwalia, & Galuska, 2000; Ghaderi, 2003; Limbert, 2010; Steptoe, Perkins-Porras, Rink, Hilton, & Cappuccio, 2004), and perceived barriers (Shaikh, Yaroch, Nebeling, Yeh, & Resnicow, 2008) also influence engagement and maintenance of healthy dietary self-management behavior.

This research study had three goals: (a) to translate expected theoretical relationships among HAPA constructs into a causal model, (b) to examine the strength of structural relationships among constructs that influence engagement in healthy dietary behavior, and (c) to evaluate the general compatibility (i.e., goodness of fit) of the model with the data. Figure 2 provides a graphical depiction of our direct translation of the HAPA model as a motivational model for healthy dietary self-management for people with MS.

Figure 2.

The HAPA dietary self-management model for people with MS

The model made the following a priori specification based on the hypothetical relationships among different constructs in the HAPA model:

Action self-efficacy, outcome expectancy, risk perception, severity of MS symptoms, and social support are associated with intention to engage in healthy dietary self-management. In the HAPA model, action self-efficacy and outcome expectancies are seen as major predictors of intentions. The effect of risk perception on intention is considered a less strong predictor (Schwarzer, 2008). The HAPA model does not discuss the effect of severity of disability on healthy dietary intentions. However, researchers called for the attention on the disability severity associated with healthy dietary intention (Habek et al., 2010; Ravesloot et al, 2011; Schwarz & Leweling, 2005). In the HAPA model, action self-efficacy is posited to be directly associated with maintenance self-efficacy; maintenance self-efficacy is related to recovery self-efficacy, and in turn, recovery self-efficacy would be directly associated with dietary self-management (Schwarzer, 2008).

Both maintenance self-efficacy and intention to engage in dietary self-management are directly associated with action and coping planning, and in turn, action and coping planning is directly associated with self-management (Schwarzer, 2008).

Severity of MS is directly associated with perceived barriers to dietary self-management, and in turn, barriers to dietary self-management are directly associated with dietary self-management level (Habek et al., 2010; Ravesloot et al., 2011; Schwarz & Leweling, 2005).

Method

Participants

We recruited 209 individuals with a self-reported diagnosis of MS from the NMSS and the neurology clinic of a university teaching hospital in the Midwest. Participants enrolled by clicking on the online anonymous survey link from a research announcement on the NMSS website. To protect anonymity, the survey did not ask where participants lived or where they heard the research announcement. Inclusion criteria specified that the participants were diagnosed with MS, 18 to 65 years old, and living in the community. Participant ages ranged from 19 to 67 years, with a mean of 47.39 years (SD = 10.12). Most were women (n = 181, 88.3%), married (n = 149, 71.3%), and White (n = 182, 87.1%). Participants were relatively well educated with 34.4% having some college education, 34.9% having graduated from college, and 21.1% having completed graduate school education. About 38.3% of the participants were retired due to MS, 33.5% were employed full-time, and 12.4% were unemployed. Sixty-three (30.1%) participants were professionals, 19 were managers (9.1%), and 32 (15.3%) were in clerical or sales. In addition, 171 (81.8%) of the participants reported being treated for secondary health problems (e.g., obesity, high blood pressure, and diabetes).

Instrument

Healthy Dietary Stages of Change Instrument (HDSC)

The HDSC was adopted from the Physical Activity Stages of Change Instrument developed by Nigg et al. (2005) to operationalize the concept of readiness to engage in physical activity. The current study replaced physical activity with healthy diet. In the present study, the HDSC was used as an outcome measure to assess the degree of engagement in a healthy diet. The HDSC is composed of four items (e.g., “Do you currently eat healthy foods?”). Items are rated on a dichotomous “yes” or “no” format. A scoring algorithm was provided by Nigg et al. to convert the scores in the four items to represent the degree of engagement in physical activity along a 5-point continuum: 1 (precontemplation [PC]), 2 (contemplation [C]), 3 (preparation [P]), 4 (action [A]), and 5 (maintenance [M]). Individuals with scores of 4 and 5 are considered actively engaging in a healthy diet for the purpose of this study. The current study regarded the stages of healthy diet as how much a person engaged in a healthy diet. The higher the stages, the more regularly healthy eating was taking place. The present study used this as the proxy of healthy diet in the path analysis.

Minimal Record of Disability (MRD)

The MRD was developed by the International Federation of Multiple Sclerosis Societies (1984) to operationalize the severity of MS by evaluating MS symptoms and performance of activities of daily living (ADL). It is composed of 23 items and two subscales: (a) the Incapacity Status Scale (ISS), 16 items (e.g., “stair climbing,” “speech and hearing,” and “mood and though”), focused on functional disability in ADL, and (b) the Environment Status Scale (ESS), 7 items (e.g., “work status,” “personal residence or home,” and “transportation”), measures social impairment resulting from the illness. ISS items are rated on a 5-point Likert-type scale from 0 (no disability) to 4 (most disability). ESS items are rated on a 6-point Likert-type scale ranging from 0 (no disability) to 5 (totally lost). The ISS, with its focus on ADL, was used to operationally define MS severity in the current study. Moderate intraclass correlation coefficients (ICC) were found between the MRD and the Kurtzke Functional System (KFS), ranging from .26 with the KFS sensory scale to .69 with the KFS pyramidal function scale (Solari et al., 1993). Also, high concordance (ICC = .84) was found between the MRD and the Expanded Disability Status Scale. Cronbach’s alpha coefficients for the ISS in the present study were computed to be .90.

Action Self-Efficacy Scale–Healthy Dietary (ASES-HD)

The ASES was developed by Renner and Schwarzer (2005) to operationalize health action self-efficacy, including nutrition, physical exercise, alcohol reduction, and smoking cessation self-efficacy subscales. The two-item ASES-HD subscale was used in this study (e.g., “I can manage to stick to healthy foods, even when I have to try several times until it works.”). ASES items are rated using a 4-point Likert-type rating scale from 1 (very uncertain) to 4 (very certain). The internal consistency reliability coefficient (Cronbach’s alpha) for the HD scale was reported to be .79 by Renner and Schwarzer and .92 for the present study.

Outcome Expectancy Scale–Healthy Dietary (OES-HD)

The OES was developed by Renner and Schwarzer (2005) to operationalize the concept of health outcome expectancies. It is composed of 12 items and four subscales (i.e., nutrition, physical exercise, alcohol reduction, and smoking cessation outcome expectancy subscales). The OES-HD scale was used in this study (e.g., “If I eat healthy foods that will be good for my blood pressure and cholesterol level.”). OES items are rated using a 4-point Likert-type rating scale from 1 (not at all true) to 4 (exactly true). The internal consistency reliability coefficient (Cronbach’s alpha) for the HD scale was reported to be .81 by Renner and Schwarzer and .73 for the present study.

Health/Safety Risk Perceptions and Expected Benefits Scale

The scale includes two subscales, Health/Safety Risk Perceptions subscale (HRPS) and Health/Safety Expected Benefits subscale (HEBS), developed by Weber, Blais, and Betz (2002). The HRPS was operationalized to measure the likelihood that people would engage in risky and harmful health behaviors, along with the perception of the magnitude of the risks related to these risky health activities. The HRPS is composed of six items (e.g., “Regularly eating high cholesterol foods.”). HRPS items are rated on a 5-point Likert-type scale from 1 (not at all risky) to 5 (extremely risky). The total score can range from 6 to 30. The internal consistency reliability coefficient (Cronbach’s alpha) for the HRPS was reported to be .81 by Weber et al. and .65 for the present study. The HEBS was operationalized to measure the perception of the benefits of engaging in positive health activities. It is composed of six items (e.g., “Eating two servings of fruits and three servings of vegetables per day.”). HEBS items are rated on a 5-point Likert-type scale from 1 (no benefits at all) to 5 (great benefits). The total score ranges from 6 to 30. The internal consistency reliability coefficient (Cronbach’s alpha) for HEBS was reported to be .82 by Weber et al. and .49 for the present study. The current study used the average score of HRPS and HEBS to measure risk perception in the path analysis.

Berlin Social-Support Scales (BSSS)

The BSSS was developed by Schwarzer and Schulz (2000) to assess cognitive and behavioral aspects of social support; quantity, type, and function of social support in general and in stressful circumstances; as well as dyadic support interactions in stressful situations. The BSSS is composed of 52 items and six subscales (perceived available support [PAS], need for support [NS], support seeking [SS], received support, provided support, and protective buffering). Only the first three subscales were used in this study: (a) PAS, 8 items (e.g., “There are people who offer me help when I need it”); (b) NS, 4 items (e.g., “It is important for me always to have someone who listens to me”); and (c) SS, 5 items (e.g., “Whenever I need help, I ask for it”). BSSS items are rated using a 4-point Likert-type scale from 1 (strongly disagree) to 4 (strongly agree). The internal consistency reliability coefficients (Cronbach’s alpha) for the PAS, NS, and SS were reported to be .83, .63, and .81, respectively (Schwarzer & Schulz, 2000), and .94, .74, and .81 for the present study, respectively.

Health Behavior Intention Scale (HBIS)

The HBIS was developed by Renner and Schwarzer (2005) to operationalize the concept of intention to engage in heath behaviors. It is composed of 10 items (e.g., “I intend to eat as healthy as possible.”). HBIS items are rated on a 7-point Likert-type scale from 1 (do not intend at all) to 7 (strongly intend). The total score for the intention variable can range from 10 to 70. The internal consistency reliability coefficient (Cronbach’s alpha) for HBIS was reported to be .65 by Renner and Schwarzer and was .82 for the present study.

Maintenance Self-Efficacy Scale— Healthy Diet (MSES-HD) Healthy Diet (MSES-HD)

The MSES-HD was developed by Luszczynska and Sutton (2006) to assess people’s perception of their maintenance self-efficacy for dietary health. It is composed of four items (e.g., “I am confident that I am able to stick to a healthy diet, even if my blood pressure and cholesterol level do not improve immediately”). Items are rated on a 4-point Likert-type scale of 1 (not at all true), 2 (barely true), 3 (mostly true), and 4 (exactly true). The total score can range from 4 to 16. The coefficient alpha for this scale was reported to be .81. The internal consistency reliability coefficient (Cronbach’s alpha) for MSES-HD was computed to be .90 for the present study.

Action Planning and Coping Planning Scale–Healthy Diet (APCPS-HD)

The APCPS-HD was developed by Sniehotta, Schwarzer, Scholz, and Schuz (2005) to measure the metacognition of action planning and coping planning for a healthy diet. It is composed of five items and two subscales: (a) action planning subscale, two items (e.g., “I already have concrete plans how to change my nutrition habits.”), and (b) coping planning, three items (e.g., “I already have concrete plans how to deal with relapses.”). Items are rated on a 4-point Likert-type scale from 1 (not at all true) to 4 (exactly true). The total score for the action planning subscale can range from 2 to 8, and the total score for the coping planning subscale can range from 3 to 12. Sniehotta et al. reported that the internal consistency reliability coefficients (Cronbach’s alpha) for action planning and coping planning were .92 and .90, respectively, and the reliability coefficients were .96 and .93 for the present study.

Barriers to Health Promoting Activities for Disabled Persons Scale (BHADP)

The BHADP was developed by Becker, Stuifbergen, and Sands (1991) to measure perceptions of barriers to health promotion activities. It is composed of 18 items and three subscales: (a) intrapersonal barriers (e.g., “too tired”), (b) interpersonal barriers (e.g., “other responsibilities”), and (c) environmental barriers (e.g., “lack of transportation”). Items are rated using a 4-point Likert-type scale from 1 (never) to 4 (routinely). The total score ranges from 18 to 72. A high score means greater perceived barriers. The internal consistency reliability coefficient (Cronbach’s alpha) for BHADP was reported to be .82 by Becker et al., and it was .85 for the present study.

Recovery Self-Efficacy Scale–Healthy Diet (RSES-HD)

The RSES-HD was developed by Luszczynska and Sutton (2006) to operationalize the concept of recovery self-efficacy for dietary health. It is composed of three items (e.g., “I am sure I can resume healthy nutrition habits again regularly, even if I have already paused for several weeks.”). Items are rated on a 4-point Likert-type scale from 1 (not at all true) to 4 (exactly true). The total score ranges from 3 to 12. The internal consistency reliability coefficient (Cronbach’s alpha) for the RSES-E was reported by Luszczynska and Sutton to be .85, and it was .94 for the present study.

Procedure

NMSS was contacted to elicit their support for this research project. Participants were recruited from several state chapters of the NMSS, through a survey link on the “Researchers Need You” section of the NMSS website. This survey link could be accessed by interested participants from every state. Additional participants were recruited from a neurology clinic at a university teaching hospital in the Midwest. Participants who volunteered to participate in the project were given a link to complete the research packet developed by the first author on the surveymonkey.com website. Participants also received a US$10 gift card as a token of appreciation for participating in the study.

Data Analysis

Path analysis was used to test the hypothesized model. All model estimations were conducted with AMOS 4.0 using maximum-likelihood estimation (Arbuckle & Wothke, 1999). In addition to the chi-square goodness-of-fit statistic, we also used several additional fit indices recommended by Weston, Gore, Chan, and Catalano (2008) to assess model fit. These indices are (a) χ²/df in the range of 2; (b) the CFI, with values greater than .90 indicating reasonably good fit; (c) the GFI, with values greater than .90 indicating reasonably good fit; (d) the Bentler and Bonett’s normed fit index (NFI), with values greater than .90 indicating reasonably good fit; and (e) the RMSEA, with values less than 0.08 indicating adequate fit.

Results

Data Screening

All variables were screened using SPSS 19.0 statistical software for accuracy of data entry, missing values, multivariate outliers, normality, linearity, and homoscedasticity. The current study deleted all cases with any missing values and outliers. We deleted 50 cases due to missing values (18.9%).To test the HAPA dietary self-management model using path analysis, we first assessed normality by visually inspecting histograms of all variables and by examining the kurtosis and skewness indices. All variables were below the critical values of skewness and kurtosis values. Using the Mahalanobis distance squared formula (Gao, Mokhtarian, & Johnston, 2008), we deleted six multivariate outliers from a complete sample of 215, resulting in a final sample of 209 participants.

To have sufficient statistical power to detect relationships and to achieve an acceptable level of precision for parameter estimates in path analysis, Kline (2005) indicated that samples with fewer than 100 participants are small, those with 100 to 200 participants are medium, and those with more than 200 participants are large. Quintana and Maxwell (1999) suggested that sample size should be calculated using the criterion of 5 to 10 participants per estimated parameter suggested by Bentler and Chou (1987). With 24 estimated parameters in the path model (9 participants per estimated parameter and sample size a little more than 200), the sample size is deemed adequate for path analysis.

Descriptive Statistics

Means, standard deviations, and intercorrelations among the variables in the HAPA dietary self-management model are shown in Table 1.

Table 1.

Means, Standard Deviations, and Correlations of Variables in the HAPA Model

Variable		M (SD)	1.	2.	3.	4.	5.	6.	7.	8.	9.	10.	11.
1.	Healthy diet	4.42 (1.10)	—
2.	Severity of disability	0.93 (0.63)	−.14*	—
3.	Action self-efficacy	2.94 (0.88)	.30**	−.08	—
4.	Outcome expectancy	3.06 (0.07)	.12	−.08	.21**	—
5.	Risk perception	4.46 (0.67)	.15*	.14*	.22**	.18**	—
6.	Social support	3.04 (0.54)	.18*	−.11	.24**	.12	.15*	—
7.	Intention	4.89 (1.31)	.25**	.03	.43*	.30**	.38**	.29**	—
8.	Maintenance self-efficacy	2.90 (0.79)	.36**	−.03	.46**	.34**	.33**	.13	.53**	—
9.	Action and coping planning	2.72 (0.95)	.32**	.01	.30**	.29**	.30**	.23**	.63**	.56**	—
10.	Barriers to health promotion	1.80 (0.45)	−.17*	.49**	−.27**	−.09	−.05	−.29**	−.17*	−.23**	−.16*	—
11.	Recovery self-efficacy	3.06 (0.78)	.37**	−.13*	.53**	.24**	.31**	.17*	.49**	.66**	.52**	−.33**	—

p < .05. **p < .01.

The average ISS score of 0.93 (severity of disability; SD = 0.63) indicated that participants’ functional disability in terms of ADL is relatively modest. Participants rated themselves as less than certain about their ability to overcome barriers to engage in a healthy diet (action self-efficacy; M = 2.94, SD = 0.88), rated themselves as moderately high in terms of outcome expectancies (M = 3.06, SD = 0. 07), agreed strongly with the risk of unhealthy eating behavior and the benefits of healthy eating behaviors (M = 4.46, SD = 0.67), and somewhat agreed with their accessibility, need, and search of social support (M = 3.04, SD = 0.54). Individuals with MS in the present study showed relatively strong intention to engage in a healthy diet (M = 4.89, SD = 1.31). They rated their maintenance self-efficacy (M = 2.90, SD = 0.79) as moderately high. They committed to have action and coping planning for engaging in a healthy diet (M = 2.72, SD = 0.95) and rated themselves as moderately low in terms of barriers to engage in a healthy diet (M = 1.80, SD = 0.45). They rated their recovery self-efficacy (M = 3.06, SD = 0.78) as moderately high. Finally, in regard to the stage status of engaging in a healthy diet, the stage distributions of the participants in the present study were 4.8% in precontemplation (PC), 2.9% in contemplation (C), 10.5% in preparation (P), 9.6% in action (A), and 72.2% in maintenance (M), comparing to the distribution of PC = 5% (±10), C = 10% (±10), P = 40% (±10), A = 10% (±10), and M = 35% (±10) in the general population reported by Nigg et al. (2005).

Path Analysis

A review of the original HAPA model (Figure 1) suggests that the diagram presented by Schwarzer (2008) may not fully capture all the structural relationships among the HAPA constructs and dietary self-management. For example, Bandura’s (1997) self-efficacy postulates that outcome expectancy correlates with self-efficacy, although self-efficacy has a stronger influence on intention and behavior. Positive and negative self-efficacy interacting with positive and negative outcome expectancy generates various behavioral performances (Bandura, 1997). Studies of social cognitive predictors of physical activity among rehabilitation clients found that self-efficacy has a small correlation with risk perception, but self-efficacy has a medium correlation with outcome expectancy, while outcome expectancy has a small to moderate correlation with risk perception (Schwarzer, Luszczynska, Ziegelmann, Scholz, & Lippke, 2008).

Ravesloot et al. (2011) in their review of four health belief theories (Health Belief, Theory of Planned Behavior, Social Cognitive, and Transtheoretical) within the disability context indicate that severity of disability may be associated with outcome expectancies by increasing sensitivity to the possibility of future disabling conditions and to the risks of behavior change. Disability can also limit the individual’s problem-solving and coping options, reducing self-efficacy. Because some MS patients believe nutrition management may alleviate their symptoms, severity of MS would be assumed to influence on self-efficacy and intention to eat healthily. Some studies of vegetable and fruit intake have reported barriers to hinder people to access healthy foods, whereas others have indicated social support facilitating people to eat healthily (Shaikh et al., 2008). These relationships are not depicted in Schwarzer’s HAPA model (Figure 1) but are further modeled in our HAPA dietary self-management model (Figure 2). On the basis of empirical guidance and conceptual and theoretical reasons, we correlated action self-efficacy, outcome expectancy, risk perception, and severity in the respecified model. In addition, we added the paths between severity and intention, severity and perceived barriers, social support and intention, social support and perceived barriers, and perceived barriers and the stages of healthy dietary behavior.

The modified hypothesized HAPA model of dietary self-management was evaluated using path analysis. This priori model test revealed a significant chi-square, χ²(34, N = 209) = 110.36, p < .001; χ²/df = 3.25; GFI = .92, NFI = .67, CFI = .88; and RMSEA = 0.10, indicating a less than adequate fit of the model to the data. Because the model was not adequate (underspecified), a reconceptualization of the model is warranted. One advantage of using modeling techniques is that hypothesized models can be modified according to theory and empirical guidance from the statistical output to improve the fit of the model (Byrne, 2001). After further examining the modification index and empirical considerations, they indicated significant improvement in model fit, which can be achieved by deleting paths (i.e., severity of MS→intention, barriers→stages of dietary) and adding paths (i.e., action self-efficacy→recovery self-efficacy, barriers→recovery self-efficacy).

Testing the respecified model revealed a significant chi-square statistic, χ²(33, N = 209) = 67.01, p < .001. However, several alternative fit indexes are acceptable: χ²/df = 2.03; GFI = .95, NFI = .90, and CFI = .95; and RMSEA = 0.07. The results showed a relatively good fit of the respecified model to the data. The parameter estimates for all the added paths in the respecified model were statistically significant. Figure 3 shows a schematic depiction of the respecified model with standardized path coefficients.

Figure 3.

The respecified HAPA dietary self-management model for people with MS

Table 2 shows the squared multiple correlation coefficients (R²) for the endogenous variables in the respecified model. The R² value represents the proportion of variance that is explained by the predictors of the variable in question. In this model, R² coefficients ranged from .15 to .52, with variables predicting intention to dietary self-management accounting for 30%; variables predicting maintenance self-efficacy accounting for 33%; variables predicting action and coping planning accounting for 45%; variables predicting perceived barriers accounting for 28%; variables predicting recovery self-efficacy accounting for 52% of the variance; and variables predicting healthy eating self-management accounting for 15%.

Table 2.

Squared Multiple Correlations for the Endogenous Variables in the Respecified Model of Healthy Dietary Behavior

Variable	R ²
Healthy dietary behavior	.15
Intention	.30
Maintenance self-efficacy	.33
Action and coping planning	.45
Perceived barriers	.28
Recovery self-efficacy	.52

All paths’ coefficients were significant in the post hoc respecified model, and some correlations between disability severity and social cognitive predictors for the healthy dietary behavior were significant. Action self-efficacy correlates with outcome expectancy and risk perception moderately (r = .21 and .22, respectively, p < .01). Outcome expectancy correlates with risk perception mildly to moderately (r = .18, p < .01). Interestingly, severity only significantly correlates with risk perception (r = .14, p < .05), neither with action self-efficacy nor outcome expectancy. The paths between action self-efficacy and intention (β = .31), outcome expectancy and intention (β = .18), and risk perception and intention (β = .27) are significant. Action self-efficacy also has a significant direct effect on maintenance self-efficacy (β = .29) and recovery self-efficacy (β = .26). It is noticeable that intention has a significant direct effect on maintenance self-efficacy (β = .40), bridging the motivation stage and the volition stage. Social support also has a significant direct effect on intention (β = .16) and inversely correlates with the perceived barriers significantly (β = −.25). Severity of MS has a strong direct effect on perceived barriers (β = .46). It is worth notice that barriers have a negative path coefficient on recovery self-efficacy (β = −.14). Intention has a strong direct effect on action and coping planning (β = .45). Maintenance self-efficacy also has a strong direct effect on action and coping planning (β = .32) and recovery self-efficacy (β = .41). Action and coping planning has a mildly direct effect on recovery self-efficacy (β = .19). Both recovery self-efficacy (β = .27) and action and coping planning (β = .17) have direct effects on the stages of change in dietary self-management mildly to moderately.

Discussion

This study proposed a motivational model for dietary self-management for people with MS. The strength of the model was the utilization of HAPA as a framework to conceptualize what motivates people with chronic illness and disability to engage in health promotion behaviors such as healthy dietary behavior. The HAPA model suggests a phased approach that develops client action self-efficacy, intention, and action coping and planning to achieve healthy diet outcomes. Dietary health promotion strategies might begin with the development of client action self-efficacy. The literature describes several ways to enhance dietary action self-efficacy, including observing successful examples, applying personal successful experiences, receiving encouragement and expectations from friends and family, and setting achievable goals (Bandura, 1997). Action self-efficacy contributes directly to client intention and maintenance self-efficacy, which directly influences on action and coping planning. Client intentions can also be influenced with clear descriptions of the risks of poor diet and benefits of good diet for individuals with MS. Disability is often viewed as a significant barrier to engaging in healthy eating behaviors; however, social support can reduce perceived barriers. Models should include some evaluation of social supports to help clients identify people in their immediate network that might assist them in addressing dietary concerns. Social supports are important to move clients toward intention and help them overcome barriers that negatively influence their recovery self-efficacy. Also, when participants develop healthy eating behaviors, it is important that they further develop their recovery self-efficacy by planning and rehearsing how they will respond to a relapse in unhealthy eating behaviors.

The original model did not fit the data well and included a couple of nonsignificant paths (i.e., the paths between severity and intention, and perceived barriers and stages of healthy diet). The less-than-adequate model fit suggested that the HAPA-based model may be underspecified. An examination of the modification indexes indicated adding several new paths to the model could be supported by health behavior change theories. The respecified model fit the data relatively well and provided a more adequate depiction of the predictive sequence of factors in the motivational and volition phases.

Specifically, action self-efficacy, outcome expectancies, and risk perceptions are moderately correlated with each other, with the relationship between action self-efficacy and risk perceptions stronger than the relationship between outcome expectancy and risk perception, and slightly stronger than action self-efficacy and outcome expectancy. Severity of MS only had a mild to modest correlation with risk perception. Contrary to the expectations of the model predicting the relationships between severity and social cognitive predictors and intention, severity of MS did not significantly correlate with action self-efficacy and outcome expectancy, and severity did not have a direct effect on intention to eat healthily. It might be because MS influences an individual’s motor functions, and the effects of a healthy diet are not as noticeable as the effects of exercise to improve mobility. The current findings showed that although MS resulted in some impairment in ADLs, the participants still had strong intention to engage in healthy eating behaviors. Therefore, regardless of severity of disability, vocational rehabilitation (VR) counselors can focus on dietary self-management interventions to help improve their clients’ health, control their MS symptoms, and increase the possibility of successful VR case closure, as the severity of MS may have little adverse effect on a clients’ intention to engage in healthy eating behavior. Rehabilitation counselors can develop clients’ intentions by enhancing social cognitive factors and analyze positive influences on process and outcomes of VR. Also a counselor works with a client to find out accessible social support to implement an individualized healthy diet plan.

Given that people with MS could benefit from a healthy diet, it appears that a good understanding of the general risks of not eating healthily, the overall effects of a healthy diet, and a belief of being able to eat healthily might contribute to intention to engage in healthy dietary self-management. The above, potentially delivered via motivational interviewing, may be useful in developing intention, especially when people are receiving good social support. Furthermore, a strong intention would directly lead to making action and coping plans. In addition, social support will help with overcoming perceived barriers to performing healthy dietary management. However, action and coping planning had a relatively mild direct influence on healthy dietary behavior. This could be because there are insufficient variances in the current participants’ eating behavior as 80% of them reported that they ate healthily.

Action self-efficacy (a motivational stage) had a direct effect on maintenance self-efficacy (a belief bridging motivation to volition) and recovery self-efficacy (a volition stage), and increments in recovery self-efficacy increased performance of healthy eating. These three specific types of self-efficacy (i.e., the initial stage of thinking and engaging in practice trials, the middle stage of increasing frequency, and the final stage of self-regulating perseverance) all play a role in motivating, reinforcing, and perpetuating healthy dietary behaviors. An individual’s intention to eat healthily directly influences maintenance self-efficacy and action and coping planning of continuing healthy dietary behaviors from moderately to strongly. Maintenance self-efficacy has moderate to strong direct effects on recovery self-efficacy and action and coping planning. Action and coping planning had a moderate direct effect on eating healthily. When severity of MS strongly correlates with perceived barriers, which is reversely associated with recovery, it would be critical to implement coping planning. Coping planning includes the anticipation of barriers and the design of alternative actions that help to attain one’s goals in spite of barriers, whereas action planning pertains to the when, where, and how of doing intended action.

Both action planning and coping planning play a pivotal role in this motivational model of dietary self-management as it bridges the gap between intention and healthy dietary behavior. It appears that having a concrete plan and coping strategies (i.e., knowing how to deal with social events, obstacles and setbacks that interrupt a healthy eating routine) will help people better implement their action plan and engage in eating healthily on a regular basis. Clearly, having preaction self-efficacy and maintenance self-efficacy beliefs contributes to better recovery self-efficacy for relapse and action and coping planning for resumption and maintenance. Further better coping planning improves action planning, leading to the performance of healthy dietary behaviors. These plans of doing and problem solving also substantially support recovery self-efficacy.

Recovery self-efficacy plays a key role in helping people with MS deal with relapse and maintain appropriate eating with healthy intake. This finding further underscores that even though the nature of self-efficacy differs from the motivation phase to the volition phase, self-efficacy in each preceding phase forms the requisite competence for successfully completing the subsequent phase of the health dietary self-management process. The initial action self-efficacy at the motivational stage still has direct effects on recovery self-efficacy at the volitional stage. This explains the importance of action self-efficacy for performing healthy eating behaviors under all stages of changes in dietary health. Severity had a direct effect on perceived barriers, and perceived barriers affect recovery self-efficacy, rather than healthy eating behavior. This finding is not expected, given what is known about the association between barriers and accessing healthy food. This might be because most of the current participants have no problem having healthy food and have a basic understanding of healthy dietary knowledge, according to their social economic status and awareness of not eating healthily.

This study’s findings suggested that, in general, the HAPA model is an effective motivational model for predicting healthy dietary behavior. However, the original HAPA model may be underspecified for people with MS. In our study, we found both theoretical and empirical support to connect the path between action self-efficacy and recovery self-efficacy as well as the path between perceived barriers and recovery self-efficacy. For future studies, the HAPA model seems to have potential to develop a theoretical-driven self-management intervention empowering people with MS at high risk of secondary conditions to take care of their health, which is the fundamental prerequisite for working successfully on a job.

Limitations

There are several limitations that should be considered when applying or interpreting the results of the present study. First, a convenience sample was used for this study, and the participants completed the research instruments through an online survey. It is possible to have sampling bias because people who have access and ability to use the Internet may constitute a unique group of individuals with MS. Also, individuals who know how to use community resources are more likely to join the local chapters of the NMSS; they may be different from people who are less aware of community resources. Second, although we recruit participants from several chapters in the countries, we did not ask in what state they live. Therefore, we cannot compare the participants based on geographical differences. Third, most of the current participants had moderately severe dysfunction in ADLs. The survey focused on current health behaviors and secondary conditions, and did not specifically ask for their mobility functioning (i.e., ambulatory ability) and types of MS. However, it is very likely that severity of motor functioning (e.g., walking, chewing) and types of MS influence self-management of health behavior, such as shopping for healthy food, cooking healthy meals, and eating healthy food. Fourth, the numbers of items for a variable was considered for the length of survey. To save participant time and energy, this study used several two- to three-item short forms, which may cause measurement problems. Fifth, all of the measures used in this study were self-report, which may not correspond to objective indicators. Finally, the HAPA model accounts for only 15% of the dietary behavior of people with MS. Other variables may need to incorporate into this motivational model of dietary self-management. For example, empowerment and self-determination are important concepts in rehabilitation counseling (Degeneffe, Chan, Dunlap, Man, & Sung, 2011); incorporating these factors into the dietary self-management model may potentially improve the predictability of the current model.

Implications

An estimated 54 million persons in the United States, or nearly 20% of the population, currently live with a chronic illness or disability. People with disabilities are equally if not more susceptible to other chronic conditions relative to the general population and are at higher risk for health complications that result from their primary disability (Bishop et al., 2000). Individuals with disabilities are at substantially elevated risk for obesity (Rimmer, 1999), psychological distress (Turner & McLean, 1989), alcohol and other drug abuse (Bombardier, Rimmele, & Zintel, 2002), and smoking (Brawarsky, Brooks, Wilber, Gertz, & Walker, 2002). Health care research also indicates that persons with disabilities make up 20% of the US population but account for 47% of health care expenditures. Health promotion interventions have great potential to improve secondary health and mental health conditions, employment status, and quality of life of people with disabilities (Ravesloot, Seekins, & White, 2005). Rehabilitation counselors, with their training in psychosocial interventions for people with disabilities, are uniquely qualified to conduct theory-driven research, develop and validate interventions, and deliver efficacious and effective health promotion interventions for people with disabilities, including persons with MS.

In recent years, the CDC provided funding for research and development of several community-based health promotion programs for people with chronic illness and disability. Most notably, Seekins and his colleagues (Ravesloot et al., 2005; Seekins, Clay, & Ravesloot, 1994; Seekins et al., 1999) validated that health promotion interventions targeted at physical activity and healthy dietary behavior could reduce the limitation experienced by people with mobility impairments due to the most common secondary conditions. They developed a health promotion intervention program called Living Well With a Disability to teach people with disabilities the importance of physical activity, maintaining good mental health, seeking health information, and eating healthily. Seekins and his colleagues demonstrated in several studies that individuals who completed the intervention program rated themselves as less limited by secondary conditions than individuals who did not complete the intervention (Ravesloot, Seekins, & Young, 1998; Seekins et al., 1999). Most recently, Ravesloot et al. (2005) conducted a national study of the Living Well intervention implemented by staff recruited from community-based agencies located in geographically diverse areas around the United States with 188 persons with disabilities. They found that participants who completed the Living Well program had significant reduction in limitations from secondary conditions and that the intervention effect was maintained 2, 4, and 12 months after the conclusion of the intervention. Specifically, Ravesloot et al. (2005) concluded that the Living Well health promotion intervention was successful in reducing the average degree of limitation people reported due to secondary conditions, the number of symptom days they experienced, and their health care costs (US$807 per person due to reductions in health care utilization) while increasing their overall life satisfaction and the behaviors they used to improve health status. Importantly, they contributed to the evidence-based practice of healthy promotion interventions for people with disabilities, including dietary self-management. Ipsen et al.’s study (2010), comparing the secondary conditions and the rate of limitations due to secondary conditions between the VR clients and Living Well clients, proved that there are needs to provide health promotion interventions within VR service delivery.

The manner by which people with MS manage their physical health depends more on what they do themselves than on what is done to or for them. However, learning and practicing adaptive dietary self-management techniques can be challenging, and the behavior changes necessary for dietary self-management are unlikely to occur in the absence of significant motivation. This study demonstrated the importance of using a phase approach to increase motivation and self-efficacy to engage in healthy eating. Risk perceptions, outcome expectancy, and action self-efficacy are correlated with each other. These three social cognitive predictors contribute to developing intention. Planning bridges intention and healthy eating behavior, and it also helps cope with relapse. Predictors in the motivation phase, such as action self-efficacy and intention, further influence the predictors in the volition phase, such as maintaining self-efficacy and recovery self-efficacy. In the current study, we found that social support influences intention, and perceived barriers impact recovery self-efficacy inversely. The mechanism of these paths is different from the application of the HAPA on physical activity (Chiu, Lynch, Chan, & Berven, 2011).

The importance of conducting a cost-benefit analysis to assess pros and cons of engaging in healthy dietary behaviors and counseling to build the extent of action self-efficacy and social support would suggest that the motivational interviewing approach may be suitable for HAPA-based interventions. Interventions such as the Living Well program provide a foundation for developing evidence-based interventions, and the HAPA framework allows us to improve and refine our dietary/nutrition self-management training for people with MS. It appears that the use of the HAPA framework and a group motivational interviewing format to encourage the development of self-initiated health behaviors (e.g., healthy diet, physical activity, stress management, and social activity), personal responsibility, and commitment to a healthy lifestyle may be effective for people with MS. The HAPA framework provides the blueprint for developing the contents of a motivation-based health promotion intervention program for people with MS. The efficacy of such HAPA-based motivational interventions can be validated using randomized controlled trials. The current study has validated a theory-driven health promotion model for people with MS whose work functions and finances have been significantly impacted by MS.

Footnotes

Acknowledgements

The senior author would like to thank Dr. Malachy Bishop at the University of Kentucky; Dr. Phillip Rumril at Kent State University; Dr. John Fleming at the University of Wisconsin–Madison; Sara Bernstein, editor, Research Now, National MS Society; Kim Kinner, Programs and Advocacy director, Wisconsin Chapter, National MS Society; and Timothy Holtz, director of programs, Minnesota Chapter, National MS Society for their support and guidance as well as their kind assistance in recruiting participants for this study.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

This research was supported by the University of Wisconsin–Madison Population Health Dissertation Grant and the Robert Wood Johnson Foundation Health & Society Scholars Program (Project 133, No. PRJ21AF).

References

Ajzen

Fishbein

(1980). Understanding attitudes and predicting social behavior. Englewood Cliffs, NJ: Prentice Hall.

Anderson

E. S.

Winett

R. A.

Wojcik

J. R.

(2007). Self- regulation, self-efficacy, outcome expectations, and social support: Social cognitive theory and nutrition behavior. Annals of Behavioral Medicine, 34, 304–312.

Arbuckle

Wothke

(1999). AMOS 4 user’s reference guide. Chicago, IL: Smallwaters.

Bandura

(1997). Self-efficacy: The exercise of control. New York, NY: Freeman.

Becker

Stuifbergen

Sands

(1991). Development of a scale to measure barriers to health promotion activities among persons with disabilities. American Journal of Health Promotion, 5, 449–454.

Bentler

P. M.

Chou

C. P.

(1987). Practical Issues in Structural Modeling. Sociological Methods Research, 16, 78–117. doi:10.1177/004912418701600100410.1177/0049124187016001004

Bishop

Chou

Chan

C. C. H.

Rahimi

Chan

Rubin

S. E.

(2000). Wellness promotion for people with disabilities in private sector rehabilitation: A conceptual and operational framework. Journal of Rehabilitation Administration, 24, 57–72.

Bombardier

C. H.

Rimmele

Zintel

(2002). The magnitude and correlates of alcohol and drug use before traumatic brain injury. Archives of Physical Medicine and Rehabilitation, 83, 1765–1773.

Bombardier

C. H.

Wadhwani

LaRotonda

(2005). Health promotion in people with multiple sclerosis. Physical Medicine and Rehabilitation Clinics of North America, 16, 557–570.

10.

Brawarsky

Brooks

D. R.

Wilber

Gertz

R. E.

Walker

D. K.

(2002). Tobacco use among adults with disabilities in Massachusetts. Tobacco Control, 11, 1129–1133.

11.

Byrne

B. M.

(2001). Structural equation modeling with AMOS. Basic concepts, applications, and programming. Mahwah, NJ: Lawrence Erlbaum.

12.

Centers for Disease Control and Prevention. (2011). Prevalence if sedentary leisure-time behavior among adults in the United States. Retrieved from http://www.cdc.gov/nchs/data/hestat/sedentary/sedentary.htm

13.

Chiu

C. Y.

Lynch

R. T.

Chan

Berven

(2011). The health action process approach as a motivational model for physical activity self-management for people with multiple sclerosis: A path analysis. Rehabilitation Psychology, 56, 171–181. doi:10.1037/a002458310.1037/a0024583

14.

Degeneffe

C. E.

Chan

Dunlap

Man

Sung

(2011). Development and validation of the caregiver empowerment scale: A resource for working with family caregivers of persons with traumatic brain injury. Rehabilitation Psychology, 56, 243–250. doi:10.1037/a002446510.1037/a0024465

15.

Fleming

J. O.

(2002). Diagnosis and management of multiple sclerosis. West Islip, NY: Professional Communications.

16.

Ford

E. S.

Ahluwalia

I. B.

Galuska

D. A.

(2000). Social relationships and cardiovascular disease risk factors: Findings from the Third National Health and Nutrition Examination Survey. Preventive Medicine, 30, 83–92.

17.

Gao

Mokhtarian

P. L.

Johnston

R. A.

(2008). Non-normality of Data in Structural Equation Models. University of California Transportation Center. Retrieved from http://www.escholarship.org/uc/item/11q0s48s

18.

Ghaderi

(2003). Structural modeling analysis of prospective risk factors for eating disorder. Eating Behaviors, 3, 387–396.

19.

Habek

Hojsak

Brinar

V. V.

(2010). Nutrition in multiple sclerosis. Clinical Neurology Neurosurgery, 112, 616–620.

20.

Hewson

D. C.

Phillips

M. A.

Simpson

K. E.

Drury

Crawford

M. A.

(1984). Food intake in multiple sclerosis. Human Nutrition Applied Nutrition, 38, 355–367.

21.

International Federation of Multiple Sclerosis Societies. (1984). Symposium on minimal record of disability for multiple sclerosis. Acta neurologica Scandinavica Supplementum, 101, 1–217.

22.

Ipsen

Seekins

Ravesloot

(2010). Building the case for delivering health promotion services within the vocational rehabilitation system. Rehabilitation Counseling Bulletin, 53, 67–77. doi:10.1177/003435520934823810.1177/0034355209348238

23.

Khan

Turner-Stokes

(2009). Effectiveness of vocational rehabilitation intervention on the return to work and employment of persons with multiple sclerosis. Cochrane Database of Systematic Reviews, 1, 1–27. doi:10.1002/14651858.CD007256.pub210.1002/14651858.CD007256.pub2

24.

Khurana

S. R.

Bamer

A. M.

Turner

A. P.

Wadhwani

R. V.

Bowen

J. D.

Leipertz

S. L.

Haselkorn

J. K.

(2009). The prevalence of overweight and obesity in veterans with multiple sclerosis. American Journal of Physical Medicine Rehabilitation, 88, 83–91.

25.

Kline

R. B.

(2005). Principles and practice of structural equation modeling (2nd ed.). New York, NY: Guilford.

26.

Limbert

(2010). Perceptions of social support and eating disorder characteristics. Health Care for Women International, 31, 170–178.

27.

Luszczynska

Sutton

(2006). Physical activity after cardiac rehabilitation: Evidence that different types of self-efficacy are important in maintainers and relapsers. Rehabilitation Psychology, 51, 314–321. doi:10.1037/0090-5550.51.4.31410.1037/0090-5550.51.4.314

28.

Lynch

R. T.

Chiu

C. Y.

(2009). Health promotion interventions for people with chronic illness and disability. In Chan

Cardoso

Chronister

(Eds.), Psychosocial interventions for people with chronic illness and disability: A handbook for evidence-based rehabilitation health professionals (pp. 277–305). New York, NY: Springer.

29.

National Multiple Sclerosis Society. (2011). About MS. Retrieved from http://www.nationalmssociety.org/about-multiple-sclerosis/index.aspx

30.

Nigg

Hellsten

Norman

Braun

Breger

Burbank

Williams

(2005). Physical activity staging distribution: Establishing a heuristic using multiple studies. Annals of Behavioral Medicine, 29, 35–45.

31.

Payne

(2001). Nutrition and diet in the clinical management of multiple sclerosis. Journal of human nutrition and dietetics, 14, 349–357.

32.

Physical Activity Guidelines Advisory Committee. (2008). 2008 Physical activity guidelines for Americans. Washington, DC: U.S. Department of Health and Human Services.

33.

Prochaska

J. O.

DiClement

C. C.

Norcross

J. C.

(1992). In search of how people change: Application to addictive behaviors. American Psychologist, 47, 1102–1114. doi:10.1037/0003-066X.47.9.110210.1037/0003-066X.47.9.1102

34.

Quintana

S. M.

Maxwell

S. E.

(1999). Implications of recent developments in structural equations modeling for counseling psychology. Counseling Psychologist, 27, 485–527.

35.

Ravesloot

Ruggiero

Ipsen

Traci

Boehm

Ware-Backs

Rigles

(2011). Health behavior change within the disability context. American Public Health Association. Retrieved from http://www.apha.org/membergroups/sections/aphasections/disability/about/healthbehaviorandchange.htm

36.

Ravesloot

Ruggiero

Ipsen

Traci

Seekins

Boehm

. . . Rigles

(2011). Disability and health behavior change. Disability and Health Journal, 4, 19–23.

37.

Ravesloot

Seekins

White

(2005). Living Well with a Disability health promotion intervention: Improved health status for consumers and lower costs for health care policymakers. Rehabilitation Psychology, 50, 239–245. doi:10.1037/0090-5550.50.3.23910.1037/0090-5550.50.3.239

38.

Ravesloot

Seekins

Young

Q. R.

(1998). Health promotion for people with chronic illness and physical disabilities: The connection between health psychology and disability prevention. Clinical Psychology & Psychotherapy, 5, 76–85.

39.

Renner

Kwon

Yang

B.-H.

Paik

Ki-C.

Kim

. . . Schwarzer

(2008). Social-cognitive predictors of dietary behaviors in South Korean men and women. International Journal of Behavioral Medicine, 15, 4–13.

40.

Renner

Schwarzer

(2005). Risk and health behaviors. Documentation of the scales of the research project: Risk appraisal consequences in Korea. International University Bermen & Freie Universitat Berlin. Retrieved from http://www.gesundheitsrisiko.de/docs/RACKEnglish.pdf

41.

Rimmer

J. H.

(1999). Health promotion for person with disabilities: The merging paradigm shift from disability prevention to prevention of secondary conditions. Physical Therapy, 79, 495–502.

42.

Schwarzer

(1992). Self-efficacy in the adoption and maintenance of health behaviors: Theoretical approaches and a new model. In Schwarzer

(Ed.), Self-efficacy: Thought control of action (pp. 217–243). Washington, DC: Hemisphere.

43.

Schwarzer

(2008). Modeling health behavior change: How to predict and modify the adoption and maintenance of health behaviors. Applied Psychology: An International Review, 57, 1–29.

44.

Schwarzer

Luszczynska

Ziegelmann

J. P.

Scholz

Lippke

(2008). Social-cognitive predictors of physical exercise adherence: Three longitudinal studies in rehabilitation. Health Psychology, 27, S54–S63.

45.

Schwarzer

Renner

(2000). Social-cognitive predictors of health behavior: Action self-efficacy and coping self-efficacy. Health Psychology, 19, 487–495.

46.

Schwarzer

Schulz

(2000). The Berlin social support scales. Retrieved from http://www.coping.de

47.

Schwarz

Leweling

(2005). Diet and multiple sclerosis. Nervenarzt, 76, 131–142.

48.

Seekins

Clay

Ravesloot

(1994). A descriptive study of secondary conditions reported by a population of adults with physical disabilities served by three independent living centers in a rural state. Journal of Rehabilitation, 60, 47–51.

49.

Seekins

White

G. W.

Ravesloot

Young

Q.-R.

Golden

Lopez

. . . Szalda-Petree

(1999). Developing and evaluating community-based health promotion programs for people with disabilities. In Simeonson

R. J.

McDevitt

L. N.

(Eds.), Issues in disability & health: The role of secondary conditions and quality of life (pp. 221–238). Chapel Hill: University of North Carolina.

50.

Shaikh

A. R.

Yaroch

A. L.

Nebeling

Yeh

M. C.

Resnicow

(2008). Psychosocial predictors of fruit and vegetable consumption in adults a review of the literature. American Journal of Preventive Medicine, 34, 535–543.

51.

Slawta

J. N.

Wilcox

A. R.

McCubbin

J. A.

Nalle

D. J.

Fox

S. D.

Anderson

(2003). Health behaviors, body composition, and coronary heart disease risk in women with multiple sclerosis. Archives of Physical Medicine and Rehabilitation, 84, 1823–1830.

52.

Sniehotta

F. F.

Schwarzer

Scholz

Schuz

(2005). Action planning and coping planning for long-term lifestyle change: Theory and assessment. European Journal of Social Psychology, 35, 565–576. doi:10.1002/ejsp.25810.1002/ejsp.258

53.

Solari

Amato

M. P.

Bergamaschi

Logroscino

Citterio

Bochicchio

Filippini

(1993). Accuracy of self-assessment of the minimal record of disability in patients with multiple sclerosis. Acta Neurologica Scandinavica, 87, 43–46. doi:10.1111/j.1600-0404.1993.tb04073.x10.1111/j.1600-0404.1993.tb04073.x

54.

Steptoe

Perkins-Porras

Rink

Hilton

Cappuccio

F. P.

(2004). Psychological and social predictors of changes in fruit and vegetable consumption over 12 months following behavioral and nutrition education counseling. Health Psychology, 23, 574–581.

55.

Stuifbergen

A. K.

Becker

Blozis

Timmerman

Kullberg

(2003). A randomized clinical trial of a wellness intervention for women with multiple sclerosis. Archives of Physical Medicine and Rehabilitation, 84, 467–477.

56.

Stuifbergen

A. K.

Roberts

G. J.

(1997). Health promotion practices of women with multiple sclerosis. Archives of Physical Medicine and Rehabilitation, 78, 3–9.

57.

Turner

R. J.

McLean

P. D.

(1989). Physical disability and psychological distress. Rehabilitation Psychology, 34, 225–242. doi:10.1037/h009172710.1037/h0091727

58.

U.S. Department of Agriculture and U.S. Department of Health and Human Services. (2010). Dietary Guidelines for Americans, 2010. Retrieved from http://www.cnpp.usda.gov/Publications/DietaryGuidelines/2010/PolicyDoc/PolicyDoc.pdf

59.

Weber

E. U.

Blais

A. R.

Betz

(2002). A domain-specific risk-attitude scale: Measuring risk perceptions and risk behaviors. Journal of Behavioral Decision Making, 15, 263–290. doi:10.1002/bdm.41410.1002/bdm.414

60.

Weston

Gore

P. A.

Chan

Catalano

(2008). An introduction to using structural equation models in rehabilitation psychology. Rehabilitation Psychology, 53, 340–356.