Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Apr 18.
Published in final edited form as: Prev Med. 2022 Jan 21;155:106958. doi: 10.1016/j.ypmed.2022.106958

A diagnosis of diabetes and health behavior maintenance in middle-aged and older adults in the United States: The role of self-efficacy and social support

Weidi Qin 1,*
PMCID: PMC9015723  NIHMSID: NIHMS1784358  PMID: 35065973

Abstract

The present study aims to investigate the relationship between a diagnosis of diabetes and the maintenance of health behaviors, and whether self-efficacy and social support moderate the relationship. The study sample came from the 2006 to 2016 waves of the Health and Retirement Study in the United States (N = 13,143). A diagnosis of diabetes was ascertained by self-reported physician-diagnosed condition. Self-efficacy was measured using a 5-item scale. Social support from family and friends were measured separately by a 3-item scale. Three health behaviors were examined, namely alcohol consumption, smoking, and physical activity. Cox proportional hazards regression models were performed to test the study aims. Respondents who reported a diagnosis of diabetes were 1.50 times more likely to fail to maintain physical activity (95% CI = 1.26, 1.77). This relationship was moderated by social support from family, which was related to lower hazards of failure to maintain physical activity among individuals who had a diagnosis of diabetes compared to those without a diagnosis. The study suggests that a diagnosis of diabetes may be a stressful health event that negatively affects physical activity maintenance. In addition, the findings highlight the importance of incorporating strategies to mobilize social support from family, which may help individuals sustain their efforts to maintaining health-promoting behaviors after a diabetes diagnosis.

Keywords: Diabetes diagnosis, Health behavior maintenance, Self-efficacy, Social support, Older adults

1. Introduction

According to the national diabetes statistics, adults aged 65 or older have the highest prevalence of diabetes (26.8% or 14.3 million) (Centers for Disease Control and Prevention [CDC], 2020). In addition, the onset of diabetes occurs most often among middle-aged individuals aged 45–64 (CDC, 2020). It has been projected that the number of adults with diabetes will nearly triple by 2060 in the United States (Lin et al., 2018). The increasing burden of diabetes in older adults highlights the importance of health behaviors. Specifically, health-promoting lifestyle behaviors, such as physical activity, play an important role in the primary prevention and management of diabetes (Li et al., 2007).

According to the framework of behavioral maintenance, experiencing a failure to achieve desirable health outcomes can result in reduced motivation to maintain health-promoting behaviors (Rothman, 2000). A diagnosis of diabetes can be a stressful health event that may negatively affect health-promoting behavior maintenance among middle-aged and older adults. That is, individuals will tend to maintain their health-promoting behaviors if these behaviors lead to desirable health outcomes and prevent undesirable health situations. Conversely, experiencing a failure to achieve desirable health goals (e.g. a diagnosis of chronic illness) can result in reduced motivation to maintain healthy behaviors.

A wide array of research has documented health behavior changes following a diagnosis of a chronic disease. For example, using the Health and Retirement Study, previous evidence suggests that older adults tend to cease smoking and excessive drinking following a diagnosis of diabetes, and those who initiated behavior changes appear to maintain these changes over the long term (Newsom et al., 2012; Xiang, 2015). Similar findings on smoking cessation have also been reported in a study focusing on multiple chronic conditions including diabetes. Specifically, older adults are more likely to quit smoking following a new diagnosis of chronic condition (Choi and DiNitto, 2015). However, mixed findings are reported regarding the changes in physical activity following a diagnosis of diabetes. For example, recent evidence shows that older adults become less physically active and more sedentary following a diabetes diagnosis (Hackett et al., 2018). By contrast, other findings indicate that, after a diagnosis of diabetes, older women tend to engage in increased physical activity in terms of both frequency and intensity (Schneider et al., 2014). Despite the extensive evidence on the role of a diabetes diagnosis in health behavior changes, less is known about whether a diabetes diagnosis affects the maintenance of health-promoting behaviors, including moderate drinking, non-smoking, and engagement in physical activity.

Additionally, psychosocial resources are also important in influencing health-promoting behaviors maintenance after a diabetes diagnosis. These resources can be individual (e.g. self-efficacy) or environmental (e.g. social support). Extensive studies have documented the relationship between psychosocial resources and health-promoting behaviors maintenance. For example, lower self-efficacy is found to predict earlier relapse of alcohol consumption and smoking (Simmons et al., 2013; Tate et al., 2008). Further, recent findings suggests that low self-efficacy is a major barrier to health behavior maintenance, including adherence to physical activity, eye care, and preventive care service utilization (Fjeldsoe et al., 2011; Kelly et al., 2016). Evidence also indicates that intervention programs aimed at enhancing self-efficacy are more successful in long-term maintenance of engagement in physical activity (McAuley et al., 2011).

Furthermore, social support from network members can influence individuals' engagement in health-promoting behaviors (Antonucci et al., 2013; Berkman and Glass, 2000). Empirically, a study using the Health and Retirement Study shows that illness-related social support is related to better adherence to physical activity and healthy diets among older adults with diabetes (Nicklett and Liang, 2010). Consistent findings are also reported in a review study focusing on social support from family and friends (Kelly et al., 2016). However, there are mixed findings on the effects of social support on drinking behaviors. While frequent contact with network members and more social support longitudinally reduce alcohol intake (Peirce et al., 2000), other evidence suggests that social network influence can actually foster the spreading of alcohol intake behaviors if there are network members engaging in heavy drinking (Rosenquist et al., 2010). Additionally, research indicates that social support may buffer the harmful effects of smoking cessation-related withdrawal symptoms, thus facilitate the maintenance of non-smoking behaviors (Creswell et al., 2014). However, limited research has differentiated the relationship sources of social support (e. g. family or friends) in studying health behavior maintenance among individuals with diabetes.

While the direct effects of self-efficacy and social support on health-promoting behaviors maintenance are documented, it remains understudied as to whether these psychosocial resources influence the effects of a diabetes diagnosis on maintenance of health-promoting behaviors. Particularly, self-efficacy and social support may serve as stress coping resources that alleviate physiological and behavioral reactions to stressors (Cohen and Wills, 1985). A diagnosis of diabetes can be a stressful health event that adversely impacts subsequent health behaviors, but self-efficacy and social support may buffer the harmful effects of a diabetes diagnosis on health behavior maintenance. Given the possible buffering role of psychosocial resources in stressful events, it is important to explore the moderating effects of self-efficacy and social support in the relationship between diabetes diagnosis and health behavior maintenance. The findings can provide implications for health behavior interventions among middle-aged and older adults with newly diagnosed diabetes, such as mobilizing psychosocial resources to cope with a new health condition.

1.1. Study objectives

The current study aims to examine the maintenance of three health-promoting behaviors (i.e. no or moderate drinking, non-smoking, and physical activity) between middle-aged and older adults who have an onset of diabetes diagnosis and those without a diagnosis, as well as the moderating effects of self-efficacy and social support on diabetes diagnosis in predicting failure to maintain health-promoting behaviors. Specifically, the research questions are:

  1. Does a diagnosis of diabetes affect the health behavior maintenance among middle-aged and older adults?

  2. Do self-efficacy and social support moderate the effects of a diabetes diagnosis on health behavior maintenance in middle-aged and older adults?

2. Methods

2.1. Sample

The study sample for the current study came from the Health and Retirement Study (HRS), which is a longitudinal panel study that surveyed a nationally representative sample of adults aged 50 or older in the United States. The current study utilized six waves of data from wave 8 (2006) through wave 13 (2016) because these waves contained psychosocial data that allowed for testing the study aims. Additionally, because the psychosocial information was collected in each biennial wave from a rotating random 50% of all the participants (Smith et al., 2017), the six-wave data were concatenated into three waves with a four-year interval between each wave in order to obtain the psychosocial data for the complete sample.

A total of 17,691 respondents completed the interviews at baseline. After excluding individuals with diabetes at baseline and proxy respondents, the remaining sample included 13,143 respondents who had an average follow-up of 2.2 waves (8.8 years), and 928 respondents developed diabetes in the follow-up waves. Respondents with diabetes at baseline were excluded in order to ascertain an incidence of diabetes diagnosis during the study period. Further, because one purpose of this investigation is to explore predictors of health behavior maintenance, individuals who engage in unhealthy behaviors at baseline were excluded. Consequently, at baseline, N = 10,712 respondents reported no or moderate drinking, N = 10,689 respondents were not current smokers, and N = 9975 respondents engaged in recommended level of physical activity, which were included in the multivariable analyses.

2.2. Key variables

2.2.1. Event/Censoring

The event of interest (=1) was defined as the failure to maintain health-promoting behaviors since baseline wave (excessive drinking, current smoking, physical inactivity). For example, older adults who reported physical inactivity during follow-up waves were defined as event occurrence (failure to maintain being physically active). Similarly, individuals who reported smoking in a subsequent wave were defined as event. Censoring (=0) referred to respondents who did not report any event of interest throughout the selected study period, or had missing data due to death or attrition, thus the event of interest was unknown.

2.2.2. Survival time

Survival start point is an older adult's interview year at baseline wave of the selected study period. Survival endpoint is the end year of follow-up interviews for respondents who had not reported an event, died, or lost follow-up. For respondents who reported events of interest, time to event was defined as the period between interview start year and the year when older adults first indicated failure to maintain health-promoting behaviors, regardless of behavior changes in the subsequent waves after the event occurrence.

2.2.3. Health behaviors

The alcohol consumption was measured by the question, “in the last three months, on the days you drink, about how many drinks do you have?” The quantity of drinks was dichotomized into two categories (0 = no or moderate drinking, 1 = excessive drinking) based on the Dietary Guidelines for Americans 2020–2025 (U.S. Department of Agriculture and U.S. Department of Health and Human Services, 2020). According to this guideline, moderate drinking is defined as 2 or fewer drinks per day for men and 1 or fewer drinks per day for women. Excessive drinking was defined as consuming more than moderate level of drinking, which was used in previous research (Newsom et al., 2012).

Current smoking (0 = non-smokers, 1 = current smokers) was created by using two measures: 1) Have you ever smoked cigarettes? 2) Do you smoke cigarettes now? The response categories for both questions were “yes” or “no”. Respondents who were past smokers or never smoked were combined into one group of non-smokers.

Physical inactivity was assessed in terms of vigorous, moderate and light activity, respectively. A dichotomized variable (0 = physical activity, 1 = physical inactivity) was created based on the physical activity guideline by the Department of Health and Human Services (DHHS), which recommended that older adults should engage in moderate-rigorous intensity physical activity 2 or more times a week (DHHS, 2018). Because light physical activities measured in the HRS (e.g. vacuum and home repairs) were defined as moderate-intensity activities in the guideline (DHHS, 2018), this measure was considered as moderate activity and included in creating the dichotomized variable of physical activity.

2.2.4. Diabetes diagnosis

The diagnosis of diabetes was measured by asking the respondents, “has a doctor ever told you that you had diabetes?” The response categories are 0 = No and 1 = Yes. Because respondents free of diabetes at baseline were selected for the study, this measure was used to calculate whether respondents had an incidence of diabetes diagnosis during the study period.

2.2.5. Self-efficacy

A five-item scale was used to measure self-efficacy. Respondents were asked to rate how much they agree or disagree with the following statements: 1) I can do just about anything I really set my mind to; 2) When I really want to do something, I usually find a way to succeed at it.3)Whether or not I am able to get what I want is in my own hands. 4) What happens to me in the future mostly depends on me. 5) I can do the things that I want to do. Each item was documented on a scale of 1 (strongly disagree) to 6 (strongly agree). An index of self-efficacy was created by averaging the items (Smith et al., 2017). The Cronbach's α for self-efficacy scale was 0.89.

2.2.6. Social support

A three-item scale was utilized to measure perceived social support from family. Specifically, respondents were asked to rate how they felt about their extended family to each of the following statements: 1) How much do they really understand the way you feel about things? 2) How much can you rely on them if you have a serious problem? 3) How much can you open up to them if you need to talk about your worries? Each item was recorded on a 1 (a lot)-to-4 (not at all) scale. Social support from friends was assessed by the same scale. An index of social support was created by averaging the items (Smith et al., 2017). The Cronbach's α was 0.86 for social support from family and 0.84 for social support from friends.

2.2.7. Control variables

Baseline socio-demographics and health indicators were controlled in the analyses. Demographic variables included age (in years), gender (male or female), race and ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, or others), education (in years), marital status (married/partnered or unmarried/un-partnered), and annual household income (in US dollars). Household income was measured continuously and was a combination of all incomes within a respondent's household (Bugliari et al., 2019). Household income was imputed to account for the missing values and was logarithmically transformed to reduce the positive skewness (Hurd et al., 2016). Health indicators included activities of daily living (ADL), instrumental activities of daily living (IADL), depressive symptoms, and a count of seven chronic conditions (high blood pressure, cancer, lung disease, heart disease, stroke, psychiatric problems, and arthritis).

2.3. Analysis strategy

Sample characteristics were compared between respondents who were censored and who reported an event of interest across the three health behaviors. Cox proportional hazards regression models using baseline covariates were conducted to examine the failure to maintain health-promoting behaviors (Collett, 2015). Although health behaviors are highly variable in nature, Cox models were used because multiple failure-time data were not available in the three-wave data. There were four steps in modeling each health behavior. In step 1, diagnosis of diabetes, self-efficacy, social support, and all control variables were entered to examine the main effects of predictors on each behavior. In step 2 to step 4, three interaction terms, 1) diabetes diagnosis × self-efficacy, 2) diabetes diagnosis × social support from family, and 3) diabetes diagnosis × social support from friends were added separately to test the moderating effects of self-efficacy and social support.

Cox regression models assume proportional individual hazard functions (Bradburn et al., 2003). In order to check the proportionality assumption, scaled Schoenfeld residuals and log-log plot of survival were examined (Cleves et al., 2016; Collett, 2015). Martingale residuals were generated and plotted to assess the nonlinearity of continuous variables (Bradburn et al., 2003). No significant violations of proportionality and linearity were identified. The HRS leave-behind sampling weights and design factors were adjusted to account for the complex survey design and to generate nationally representative estimates (Smith et al., 2017). Models' goodness of fit was assessed using the Akaike information criterion (AIC) and the global Wald test.

3. Results

Table 1 presents the descriptive statistics of respondents who reported health-promoting behaviors at baseline. For comparison purposes, respondents were grouped into two categories based on whether they were censored or reported an event of interest. At a bivariate level, respondents who developed diabetes were more likely to fail to maintain no or moderate drinking, non-smoking, and engagement in physical activity. Additionally, individuals who failed to maintain physically active tended to report lower self-efficacy and lower social support from friends.

Table 1.

Sample Characteristics by Maintenance of Health-promoting behaviors.

Drinking
(Unweighted N = 10,712)
(Weighted N = 121,281,537)		 Smoking
(N = 10,689)
(Weighted N = 122,600,525)		 Physical Activity
(N = 9975)
(Weighted N = 117,453,829)
Censored Event P
value		 Censored Event P
value		 Censored Event P
value
Age in years (mean, SE) 65.07 (0.28) 59.98 (0.34) <0.001 64.65 (0.30) 59.13 (0.59) <0.001 62.32 (0.28) 65.59 (0.37) <0.001
Gender (%, SE) 0.611 0.098 0.141
 Male 45.90 (0.01) 44.87 (0.02) 45.00 (0.00) 53.80 (0.05) 45.83 (0.01) 43.56 (0.01)
 Female 54.10 (0.00) 55.13 (0.02) 54.99 (0.00) 46.20 (0.05) 54.17 (0.01) 56.44 (0.01)
Race (%, SE) 0.025 0.075 0.002
 NH White 83.34 (0.01) 82.58 (0.02) 84.76 (0.01) 76.78 (0.05) 86.18 (0.01) 83.60 (0.01)
 NH African American 9.56 (0.00) 7.52 (0.01) 7.97 (0.01) 10.66 (0.03) 7.09 (0.00) 9.60 (0.01)
 Hispanic 7.10 (0.01) 9.90 (0.02) 7.27 (0.01) 12.56 (0.04) 6.73 (0.01) 6.80 (0.01)
Marital status ((%, SE) 0.119 0.653 <0.001
 Married/Partners 65.65 (0.01) 69.47 (0.02) 68.38 (0.01) 66.00 (0.05) 69.68 (0.01) 64.76 (0.01)
 Unmarried/Unpartnered 34.35 (0.01) 30.53 (0.02) 31.62 (0.01) 34.00 (0.05) 30.32 (0.01) 35.24 (0.01)
Years of education (mean, (%, SE) 13.06 (0.07) 13.46 (0.12) 0.004 13.31 (0.07) 13.04 (0.32) 0.388 13.58 (0.07) 12.74 (0.08) <0.001
Income $ (mean, SE) 77,841.38 (2877.13) 109,321.40 (10,371.1) 0.005 88,756.25 (3095.20) 66,029.37 (7459.75) 0.006 94,455.68 (2838.04) 70,659.09 (7384.69) 0.002
ADL difficulty ((%, SE) <0.001 0.258 <0.001
 No 85.86 (0.00) 92.84 (0.01) 88.10 (0.01) 83.55 (0.04) 92.54 (0.00) 88.09 (0.01)
 Yes 14.14 (0.00) 7.16 (0.01) 11.90 (0.01) 16.45 (0.04) 7.46 (0.00) 11.91 (0.01)
IADL difficulty ((%, SE) <0.001 0.298 0.001
 No 87.89 (0.00) 95.61 (0.01) 90.13 (0.01) 86.75 (0.04) 93.72 (0.00) 90.66 (0.01)
 Yes 12.11 (0.00) 4.39 (0.01) 9.87 (0.01) 13.25 (0.04) 6.28 (0.00) 9.34 (0.01)
Depressive symptoms (mean, SE) 1.40 (0.03) 1.14 (0.07) 0.005 1.25 (0.02) 1.90 (0.30) 0.028 1.12 (0.03) 1.39 (0.05) <0.001
Count of chronic conditions (mean, SE) 1.61 (0.02) 1.18 (0.06) <0.001 1.51 (0.02) 1.67 (0.11) 0.177 1.36 (0.02) 1.64 (0.03) <0.001
Diagnosis of diabetes 0.024 0.002 <0.001
 No 91.94 (0.00) 88.39 (0.02) 92.14 (0.00) 80.67 (0.05) 93.40 (0.00) 87.86 (0.01)
 Yes 8.06 (0.00) 11.61 (0.02) 7.86 (0.00) 19.33 (0.05) 6.60 (0.00) 12.14 (0.01)
Self-efficacy (mean, SE) 4.74 (0.01) 4.94 (0.05) <0.001 4.78 (0.01) 4.89 (0.12) 0.394 4.88 (0.01) 4.72 (0.02) <0.001
Perceived social support from family (mean, SE) 2.88 (0.01) 2.90 (0.40) 0.574 2.88 (0.01) 2.74 (0.12) 0.247 2.90 (0.01) 2.87 (0.02) 0.272
Perceived social support from friends (mean, SE) 3.06 (0.01) 3.05 (0.04) 0.741 3.07 (0.01) 2.94 (0.11) 0.222 3.11 (0.01) 3.04 (0.02) <0.001
% (SE) in the population 93.34 (0.00) 6.66 (0.00) 98.81 (0.00) 1.19 (0.00) 78.55 (0.00) 21.45 (0.00)
Open in a new tab

Note. HRS Leave-Behind survey weights were applied to adjust for the complex survey design. SE = standard error; NH = non-Hispanic; ADL = activities of daily living, IADL = instrumental activities of daily living.

Table 2 presents results from the weighted Cox regression models predicting health behavior maintenance. No main effects of diabetes diagnosis or moderating effects of self-efficacy and social support were found in predicting the failure to maintain moderate drinking and non-smoking. However, a higher level of self-efficacy was related to increased hazards of failure to maintain non-smoking. In terms of physical activity, respondents who reported a diagnosis of diabetes were 50% more likely to fail to maintain being physically active compared to those without a diagnosis of diabetes (p < 0.001). One significant interaction was found between diabetes diagnosis and perceived social support from family. The significant interaction is plotted in Fig. 1. Specifically, stronger social support from family was associated with lower hazards of failure to maintain physical activity among older adults who developed a new diagnosis of diabetes, but not related among those without a diagnosis.

Table 2.

Hazard Ratios From Cox Proportional Hazards Regressions Predicting Probability of Maintaining Health-promoting behaviors.

Excessive Drinking Current Smoking Physical Inactivity
Main effects Interaction effects
HR (95% CI) p value HR (95% CI) p value HR (95% CI) p value HR (95% CI) p value
Diabetes diagnosis 1.09 (0.72, 1.64) 0.685 1.90 (0.94, 3.85) 0.074 1.50 (1.26, 1.77) <0.001 2.67 (1.52, 4.66) 0.001
Self-Efficacy 1.10 (0.98, 1.24) 0.089 1.34 (1.02, 1.75) 0.034 0.95 (0.91, 0.98) 0.003 0.94 (0.91, 0.98) 0.003
Perceived social support from family 1.01 (0.90, 1.13) 0.897 0.93 (0.66, 1.30) 0.658 0.97 (0.91, 1.03) 0.274 0.99 (0.92, 1.07) 0.866
Perceived social support from friend 0.91 (0.78, 1.05) 0.200 0.82 (0.57, 1.20) 0.304 0.95 (0.89, 1.02) 0.135 0.95 (0.89, 1.01) 0.118
Perceived social support from family × Diabetes diagnosis – – – 0.82 (0.68, 0.98) 0.032
Age 0.95 (0.94, 0.96) <0.001 0.93 (0.90, 0.95) <0.001 1.02 (1.02, 1.03) <0.001 1.02 (1.02, 1.03) <0.001
Women (ref. = men) 1.15 (0.95, 1.38) 0.143 0.78 (0.49, 1.25) 0.298 0.96 (0.85, 1.09) 0.523 0.96 (0.85, 1.08) 0.478
Race (ref. = white)
 African American 0.83 (0.60, 1.15) 0.259 1.09 (0.48, 2.48) 0.832 1.18 (1.01, 1.38) 0.034 1.19 (1.02, 1.39) 0.028
 Hispanic 1.43 (1.02, 2.00) 0.039 1.43 (0.58, 3.51) 0.425 0.81 (0.65, 1.01) 0.059 0.82 (0.66, 1.01) 0.065
Non-married (ref. = married) 1.19 (0.92, 1.56) 0.187 0.98 (0.54, 1.80) 0.956 0.91 (0.81, 1.03) 0.131 0.91 (0.81, 1.02) 0.118
Year of education 0.98 (0.95, 1.02) 0.382 0.99 (0.90, 1.08) 0.803 0.94 (0.93, 0.96) <0.001 0.94 (0.93, 0.92) <0.001
Income (log transformed) 1.13 (0.98, 1.32) 0.094 0.88 (0.68, 1.13) 0.304 0.83 (0.78, 0.88) <0.001 0.83 (0.78, 0.88) <0.001
ADL difficulty (ref. = no) 1.05 (0.68, 1.62) 0.815 1.27 (0.60, 2.69) 0.526 1.20 (1.01, 1.43) 0.037 1.20 (1.00, 1.43) 0.047
IADL difficulty (ref. = no) 0.55 (0.35, 0.87) 0.012 1.21 (0.52, 2.82) 0.649 1.05 (0.83, 1.32) 0.694 1.05 (0.83, 1.33) 0.667
Depressive symptoms (CES-D) 0.99 (0.94, 1.04) 0.748 1.11 (0.94, 1.32) 0.217 1.02 (0.98, 1.06) 0.281 1.02 (0.98, 1.06) 0.261
Count of chronic conditions 0.94 (0.84, 1.04) 0.246 1.22 (0.99, 1.52) 0.066 1.05 (1.00, 1.09) 0.033 1.04 (1.00, 1.09) 0.035
Model specifics
 Wald χ2 (df) 201.27 (15) <0.001 65.96 (15) <0.001 432.91 (15) <0.001 437.67 (16) <0.001
 AIC 9467.11 1688.16 36,759.82 36,757.06
 # of persons 9224 9255 8727 8727
Open in a new tab

Note. HR = hazard ratio. CI = confidence interval. ADL = activities of daily living; IADL = instrumental activities of daily living; AIC = Akaike information criterion. Person-level leave-behind survey weights were applied to adjust for the complex survey design. Results of non-significant interactions were not shown.

Fig. 1.

Fig. 1.

Open in a new tab

Predicted probability of maintaining physically active over time by diabetes diagnosis and perceived social support from family. Predictions were plotted based on estimates of interaction term from Cox proportional hazard regression model, with all covariates adjusted and LB sampling weights applied. Survival was defined as no physical inactivity by the end of the study. Each data point represented the proportion of respondents who continued engaging in physical activity with recommended intensity and frequency.

4. Discussion

The study findings extend the previous work by revealing the stressful and negative role of a diabetes diagnosis in the maintenance of physical activity. Furthermore, the study identified the buffering effects of social support from family on the relationship between an adverse health event and health behavior maintenance. These associations highlight the importance of effective utilization of psychosocial coping resources in successful diabetes self-management.

As suggested by the framework of behavioral maintenance, individuals may feel discouraged and tend to discontinue engagement in health-promoting behaviors if the health outcomes fail to meet the expectations (Schwarzer, 2008). The present study findings indicate that a diabetes diagnosis may be a stressful event that discourages individuals' motivation to maintain physically active. It is also possible that older adults with diabetes tend to experience significant physical and psychological barriers to physical activity (Rasinaho et al., 2007). However, there are no significant effects of diabetes diagnosis on drinking and smoking behavior. The findings suggest that older adults who are moderate drinkers and non-smokers are likely to continue their behaviors after a diagnosis of diabetes. In fact, existing evidence suggests that, for health-risking behaviors such as heavy drinking and smoking, a diagnosis of diabetes can serve as a teachable moment that prompts individuals to cease engagement in these behaviors (Hackett et al., 2018; Keenan, 2009). It is possible that, compared to physical inactivity, individuals perceive drinking and smoking as more harmful and associated with more severe health consequences, thus they have higher motivation to avoid health-risking behaviors after a diabetes diagnosis. Based on evidence from the current study and previous research, a diagnosis of diabetes may be a stressor that undermines the maintenance of health-promoting behaviors (e.g. physical activity), but a teachable moment that initiates the reduction and cessation of health-risking behaviors (e.g. drinking and smoking) (Newsom et al., 2012; Xiang, 2015).

Further, higher self-efficacy is associated with better physical activity maintenance, suggesting that confidence and motivation play beneficial roles in individuals' compliance with physical activity. This finding echoes existing evidence that stronger self-efficacy is predictive of improved engagement in physical activity (Hahn et al., 2015; King et al., 2010). Surprisingly, self-efficacy is related to increased hazards of failure to maintain non-smoking. It is possible that the maintenance of non-smoking is heavily attributed to satisfaction with the health outcomes of cessation (Baldwin et al., 2006). That is, while self-efficacy leads to motivation to initiate health-promoting behaviors, adherence to health-promoting behaviors depends on whether one achieves desired health goals. Other evidence also suggests that a very high level of self-efficacy has counteracting effects in the long-term success of smoking cessation, as individuals can overestimate their capacity to maintain non-smoking and neglect the coping skills to engage in risk-reducing behaviors (Staring and Breteler, 2004). Nevertheless, previous research suggests that diabetes-specific self-efficacy might be more relevant and useful in influencing a specific lifestyle behavior among individuals with diabetes (King et al., 2010). Given that a global measure of self-efficacy is assessed in the HRS, future research needs to consider using diabetes-specific self-efficacy in studying health behavior maintenance.

In terms of the moderating effects of psychosocial resources, one significant interaction is found between social support from family and diabetes diagnosis in predicting physical activity maintenance. This finding lends evidence to the buffering and protective role of social support in adverse situations (Cohen and Wills, 1985). While a diagnosis of diabetes can negatively affect physical activity maintenance, strong social support from family may buffer the harmful effects and help individuals sustain the efforts in maintenance. This finding suggests that social support from family may be an important stress-coping resource that protects against physical inactivity following a diabetes diagnosis. In particular, family members can provide encouragement after a diabetes diagnosis and help older adults cope with the new condition. However, compared to older adults who developed diabetes, those without diabetes have significantly lower hazards of failure in physical activity maintenance, regardless of the level of social support from family. It is possible that satisfaction with desirable health outcomes (e. g. absence of diabetes diagnosis) enhances individuals' motivation to continue the efforts to behavior maintenance (Baldwin et al., 2006; Schwarzer, 2008).

The study findings should be interpreted in light of limitations. First, no causal inferences can be drawn from the findings of the present study because it is not possible to determine the order of timing between a diabetes diagnosis and failure to maintain a behavior due to the scope of the analytic strategy. Nevertheless, previous evidence has pointed to the initiation and maintenance of health behaviors over time following a diagnosis (Newsom et al., 2012; Xiang, 2015). Second, self-reported diagnosis of diabetes automatically excludes individuals with undiagnosed diabetes. It has been estimated that 3.6% of middle-aged adults and 5.4% of older adults have undiagnosed diabetes (CDC, 2020). Exclusion of this sample can lead to a biased estimation of the effects of diabetes diagnosis on hazards of failure. Third, the measures on the frequency of physical activity across the three intensity levels are crude. A finer scale is necessary to measure whether older adults meet the recommended 150 min of moderate-intensity activity or 75 min of vigorous-intensity activity per week (DHHS, 2018). Last, three waves of data preclude using analysis approach that can account for multiple failure-time data on health behavior changes, which are highly variable. As more waves of data become available, additional analysis is needed to further examine health behavior maintenance.

5. Conclusions

The study highlights the negative role of diabetes diagnosis in failure to maintain physical activity. Health interventions need to consider that a diagnosis of diabetes may be a stressful health event that adversely affects older adults' health behavior maintenance, such as physical activity. Further, given the buffering role of social support from family in the adverse effects of diabetes diagnosis on health behavior maintenance, social support can be an important psychosocial resource that help individuals cope with newly diagnosed diabetes. Diabetes educators can help older adults identify and mobilize support from family members, and involve family members in developing health behavior interventions and diabetes management plans.

Acknowledgements

This study was developed from the author’s doctoral dissertation. The author would like to thank her dissertation committee (Ann Nguyen, PhD; David Miller, PhD; David Hussey, PhD; Siran Koroukian, PhD; and Adam Perzynski, PhD) for their guidance on the dissertation. The author would also like to thank Dr. Pingfu Fu for his help with the interpretation of survival data.

Funding

The preparation of this article was supported in part by the National Institute on Aging of the National Institutes of Health (T32AG000221). The content is solely the responsibility of the author and does not necessarily represent the official views of the National Institutes of Health. The Health and Retirement Study (HRS) is sponsored by the National Institute on Aging of the National Institutes of Health (U01AG009740) and is conducted by the University of Michigan.

Footnotes

Disclosure of ethical compliance

This study was based on a publicly available de-identified data, and thus exempt from ethical compliance.

Declaration of Competing Interest

None.

References

  1. Antonucci T, Ajrouch KJ, Birditt KS, 2013. The convoy model: explaining social relations from a multidisciplinary perspective. The Gerontologist 54 (1), 82–92. 10.1093/geront/gnt118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Baldwin AS, Rothman AJ, Hertel AW, Linde JA, Jeffery RW, Finch EA, Lando HA, 2006. Specifying the determinants of the initiation and maintenance of behavior change: an examination of self-efficacy, satisfaction, and smoking cessation. Health Psychol. 25 (Issue 5), 626–634. American Psychological Association. 10.1037/0278-6133.25-5.626. [DOI] [PubMed] [Google Scholar]
  3. Berkman LF, Glass T, 2000. 7. Social integration, social networks, social support, and health. In: Social Epidemiology. Oxford University Press, pp. 137–173. [Google Scholar]
  4. Bradburn MJ, Clark TG, Love SB, Altman DG, 2003. Survival analysis part III: multivariate data analysis – choosing a model and assessing its adequacy and fit. Br. J. Cancer 89 (4), 605–611. 10.1038/sj.bjc.6601120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bugliari D, Campbell N, Chan C, Hayden O, Hayes J, Hurd Michael Karabatakis A, Main R, Mallett J, McCullough C, Meijer E, Moldoff M, Pantoja Philip, Rohwedder Susann, St Clair P, 2019. RAND HRS Longitudinal File 2016 (V1) Documentation. https://www.rand.org/content/dam/rand/www/external/labor/aging/dataprod/randhrs1992_2016v1.pdf. [Google Scholar]
  6. Centers for Disease Control and Prevention, 2020. National Diabetes Statistics Report 2020. https://www.cdc.gov/diabetes/pdfs/data/statistics/national-diabetes-statistics-report.pdf.
  7. Choi NG, DiNitto DM, 2015. Role of new diagnosis, social isolation, and depression in older adults’ smoking cessation. The Gerontologist 55 (5), 793–801. 10.1093/geront/gnu049. [DOI] [PubMed] [Google Scholar]
  8. Cleves M, Gould W, Marchenko YV, 2016. An Introduction to Survival Analysis Using Stata (Third). Stata Press. [Google Scholar]
  9. Cohen S, Wills TA, 1985. Stress, social support, and the buffering hypothesis. Psychol. Bull 98 (Issue 2), 310–357. American Psychological Association. 10.1037/0033-2909.98.2.310. [DOI] [PubMed] [Google Scholar]
  10. Collett D, 2015. Modelling Survival Data in Medical Research (Third). CRC Press. [Google Scholar]
  11. Creswell KG, Cheng Y, Levine MD, 2014. A test of the stress-buffering model of social support in smoking cessation: is the relationship between social support and time to relapse mediated by reduced withdrawal symptoms? Nicotine Tob. Res 17 (5), 566–571. 10.1093/ntr/ntu192. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. DHHS, 2018. Physical Activity Guidelines for Americans, 2nd edition https://health.gov/our-work/physical-activity/current-guidelines.
  13. Fjeldsoe B, Neuhaus M, Winkler E, Eakin E, 2011. Systematic review of maintenance of behavior change following physical activity and dietary interventions. Health Psychol. 30 (Issue 1), 99–109. American Psychological Association. 10.1037/a0021974. [DOI] [PubMed] [Google Scholar]
  14. Hackett RA, Moore C, Steptoe A, Lassale C, 2018. Health behaviour changes after type 2 diabetes diagnosis: findings from the English Longitudinal Study of Ageing. Sci. Rep 8 (1), 16938. 10.1038/s41598-018-35238-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hahn EA, Burns JL, Jacobs EA, Ganschow PS, Garcia SF, Rutsohn JP, Baker DW, 2015. Health literacy and patient-reported outcomes: a cross-sectional study of underserved English- and Spanish-speaking patients with type 2 diabetes. J. Health Commun. 20 (sup2), 4–15. 10.1080/10810730.2015.1061071. [DOI] [PubMed] [Google Scholar]
  16. Hurd MD, Meijer E, Moldoff M, Rohwedder S, 2016. Improved Wealth Measures in the Health and Retirement Study: Asset Reconciliation and Cross-Wave Imputation. RAND Corporation PP; - Santa Monica, CA. 10.7249/WR1150. [DOI] [Google Scholar]
  17. Keenan PS, 2009. Smoking and weight change after new health diagnoses in older adults. Arch. Intern. Med 169 (3), 237–242. 10.1001/archinternmed.2008.557. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kelly S, Martin S, Kuhn I, Cowan A, Brayne C, Lafortune L, 2016. Barriers and facilitators to the uptake and maintenance of healthy behaviours by people at mid-life: a rapid systematic review. PLoS One 11 (1), e0145074. 10.1371/journal.pone.0145074. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. King DK, Glasgow RE, Toobert DJ, Strycker LA, Estabrooks PA, Osuna D, Faber AJ, 2010. Self-efficacy, problem solving, and social-environmental support are associated with diabetes self-management behaviors. Diabetes Care 33 (4). 10.2337/dc09-1746, 751 LP–753. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Li C, Ford ES, Mokdad AH, Jiles R, Giles WH, 2007. Clustering of multiple healthy lifestyle habits and health-related quality of life among U.S. adults with diabetes. Diabetes Care 30 (7), 1770–1776. 10.2337/dc06-2571. [DOI] [PubMed] [Google Scholar]
  21. Lin J, Thompson TJ, Cheng YJ, Zhuo X, Zhang P, Gregg E, Rolka DB, 2018. Projection of the future diabetes burden in the United States through 2060. Popul. Health Metrics 16 (1), 9. 10.1186/s12963-018-0166-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. McAuley E, Mullen SP, Szabo AN, White SM, Wojcicki TR, Mailey EL, Gothe NP, Olson EA, Voss M, Erickson K, Prakash R, Kramer AF, 2011. Self-regulatory processes and exercise adherence in older adults: executive function and self-efficacy effects. Am. J. Prev. Med 41 (3), 284–290. 10.1016/j.amepre.2011.04.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Newsom JT, Huguet N, McCarthy MJ, Ramage-Morin P, Kaplan MS, Bernier J, McFarland BH, Oderkirk J, 2012. Health behavior change following chronic illness in middle and later life. J. Gerontol. Series B 67B (3), 279–288. 10.1093/geronb/gbr103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Nicklett EJ, Liang J, 2010. Diabetes-related support, regimen adherence, and health decline among older adults. J. Gerontol. Series B 65B (3), 390–399. 10.1093/geronb/gbp050. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Peirce RS, Frone MR, Russell M, Cooper ML, Mudar P, 2000. A longitudinal model of social contact, social support, depression, and alcohol use. Health Psychology 19 (Issue 1), 28–38. American Psychological Association. 10.1037/0278-6133.19.1.28. [DOI] [PubMed] [Google Scholar]
  26. Rasinaho M, Hirvensalo M, Leinonen R, Lintunen T, Rantanen T, 2007. Motives for and barriers to physical activity among older adults with mobility limitations. J. Aging Phys. Act 15 (1), 90–102. 10.1123/japa.15.1.90. [DOI] [PubMed] [Google Scholar]
  27. Rosenquist JN, Murabito J, Fowler JH, Christakis NA, 2010. The spread of alcohol consumption behavior in a large social network. Ann. Intern. Med 152 (7), 426. 10.7326/0003-4819-152-7-201004060-00007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Rothman AJ, 2000. Toward a theory-based analysis of behavioral maintenance. Health Psychol. 19 (Issues 1, Suppl), 64–69. American Psychological Association. 10.1037/0278-6133.19.Suppl1.64. [DOI] [PubMed] [Google Scholar]
  29. Schneider KL, Andrews C, Hovey KM, Seguin RA, Manini T, LaMonte MJ, Margolis KL, Waring ME, Ning Y, Sims S, Ma Y, Ockene J, Stefanick ML, Pagoto SL, 2014. Change in physical activity after a diabetes diagnosis: opportunity for intervention. Med. Sci. Sports Exerc. 46 (1), 84–91. 10.1249/MSS.0b013e3182a33010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Schwarzer R, 2008. Modeling health behavior change: how to predict and modify the adoption and maintenance of health behaviors. Appl. Psychol 57 (1), 1–29. 10.1111/j.1464-0597.2007.00325.x. [DOI] [Google Scholar]
  31. Simmons VN, Litvin EB, Jacobsen PB, Patel RD, McCaffrey JC, Oliver JA, Sutton SK, Brandon TH, 2013. Predictors of smoking relapse in patients with thoracic cancer or head and neck cancer. Cancer 119 (7), 1420–1427. 10.1002/cncr.27880. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Smith J, Ryan L, Fisher G, Sonnega A, Weir D, 2017. HRS Psychosocial and Lifestyle Questionnaire 2006–2016. [Google Scholar]
  33. Staring ABP, Breteler MHM, 2004. Decline in smoking cessation rate associated with high self-efficacy scores. Prev. Med 39 (5), 863–868. 10.1016/j.ypmed.2004.03.025. [DOI] [PubMed] [Google Scholar]
  34. Tate SR, Wu J, McQuaid JR, Cummins K, Shriver C, Krenek M, Brown SA, 2008. Comorbidity of substance dependence and depression: role of life stress and self-efficacy in sustaining abstinence. Psychol. Addict. Behav 22 (Issue 1), 47–57. American Psychological Association. 10.1037/0893-164X.22.1.47. [DOI] [PubMed] [Google Scholar]
  35. U.S. Department of Agriculture, & U.S. Department of Health and Human Services, 2020. Dietary Guidelines for Americans 2020–2025. Ninth Edition. https://www.dietaryguidelines.gov/sites/default/files/2020-12/Dietary_Guidelines_for_Americans_2020-2025.pdf. [Google Scholar]
  36. Xiang X, 2015. Chronic disease diagnosis as a teachable moment for health behavior changes among middle-aged and older adults. J. Aging Health 28 (6), 995–1015. 10.1177/0898264315614573. [DOI] [PubMed] [Google Scholar]

RESOURCES