Elderly Health Journal

1. Department of Health Policy and Community Health, Jiann-Ping Hsu College of Public Health, Georgia Southern University, Statesboro, Georgia, United States of America
2. Nutrition Program, Hunter College, City University of New York, New York, United States of America

Copyright © 2021 Elderly Health Journal. This is an open-access article distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/) which permits copy and redistribute the material just in noncommercial usages, provided the original work is properly cite.

Introduction
 

    Diabetes is a significant public health problem worldwide in terms of its incidence and prevalence, costs of care, and mortalities tied to its complications (1). According to an estimate in 2017, about 451 million adults were living with diabetes worldwide and it’s expected to increase to 693 million by 2045 (2). The health care spending on diabetes patients was estimated to be 850 billion United States (US) dollars in 2017 (2). In the US, diabetes has become the seventh leading cause of death disease (3) and, based on estimates from the blood samples collected in the 2011-2012 National Health and Nutrition Examination Surveys, a significant proportion of diabetes patients went undiagnosed (4). 
    Diabetes is especially alarming among older Americans. According to Laiteerapong and Huang (1), about 40% of adult diabetes patients in the US were adults aged 65 or older. The number of diabetes patients was projected to increase to 26.7 million or a 55% of prevalence rate by 2050 (5). Rosenquist and Fox (6) indicated that people with diabetes were 2 to 3 times more likely to die when compared to people without diabetes. A diagnosed diabetes patient stays in the disease for an average of 14.7 years and over half of them would live more than 10 years (6). With improvement in the care of diabetes, evidence showed that people with diabetes had survived longer (3). Consequently, both the social and economic burden of caring for diabetes has become a challenge to both their family caregivers and health care systems.
The diabetes among immigrants
    Older Americans with diabetes are a racial and ethnic heterogeneous group. Considerable variations in the prevalence of diabetes can be found across Asian, Hispanic, and non-Hispanic whites (7, 8). Significant socioeconomic disparities in diabetes-related mortality have also been documented (9, 10). Moreover, disparities in diabetes prevalence are also observed in the immigrant populations to the US (8). Higher prevalence rates were observed among non-Hispanic black, non-Hispanic Asian, and Hispanic population (4). Data from the 2011-2016 National Health Interview Survey found that, among US adults, 20.4 percent non-Hispanic blacks, and 22.1 percent Hispanics had diabetes. For non-Hispanic whites, the prevalence of diabetes was 12.1 percent. Ethnic differences between Hispanic and non-Hispanic whites would be even greater if the undiagnosed diabetes cases were included (7, 11). Ironically, with higher morbidity, lower levels of education, lower rates of health insurance coverage, and lower socioeconomic status (SES), there appears to be a “Hispanic paradox” or healthy immigrant advantage when compared to the non-Hispanic whites in the US (12, 14).
    Using data from national representative samples, studies have found Hispanic older adults live longer than their counterparts of non-Hispanic whites who had higher SES in average (15,16). Particularly, for adults aged 65 or older, foreign-born Hispanics had a substantially lower risk of mortality, compared to non-Hispanic whites. Self-selection (healthier individuals are more likely to migrate) and protection (strong social support and favorable health-related behaviors) have been proposed as potential reasons (16,17). In general, the study of the “immigrant health advantage” has been more consistently found in mortality. For other health status measures results have been mixed (15, 17).
Diabetes mortality and personal mastery
    Individuals with diabetes do have an increased risk of mortality compared to individuals without diabetes. It is estimated that people with diabetes have an 8-year shorter life expectancy compared to individuals without diabetes (6). The public health significance of diabetes is not reflected in the number of cases died of the disease. Researchers have argued that the number of deaths attributed to diabetes from death certificate data is likely an underestimate. The definitions of diabetes diagnosis have been changed over the past decades and the role of diabetes is often under-recognized when physicians attribute causes of death to a death certificate (6).
    Another reason why diabetes mortality may be under-represented is because diabetes patients are often at a higher risk for cardiovascular or other complications (1). With access to quality care and regular exercises, most diabetes patients could live with the disease for an extensive period to develop comorbidities. As the disease process unfolds, diabetes patients could be exposed to persistent stressors for a long period of time and a sense of control or personal mastery has long been documented to correlate with better health (18).
    Studies from the chronic stress literature have recognized that impacts of stressors on health and well-being might be reduced when the individuals have high levels of mastery (19). As a positive psychological resource, personal mastery is a conviction that a person has control over the important circumstances that are impinging on his or her life (18). High personal mastery is associated with better physical functioning, wellbeing, and longevity among older populations (19–21). This psychological resilience has been linked to positive health outcomes among people with chronic conditions, disabilities and their caregivers (22–24). Older adults with a better management of diabetes have to constantly pay attention to glycemic control. They are often told to exercise regularly and eat a healthy diet. As argued by Lachman and Firth (25), a plausible mechanism linking higher control beliefs to better health is engagement in beneficial health-related behaviors. Fewer studies have focused on the association between personal mastery and mortality, but evidence indicates that personal mastery is protective against both all-cause and cardiovascular mortality (26, 27).
    The stress associated with managing diabetes could have negative impacts on survival. Several studies have corroborated that there is a positive association between personal mastery or better coping styles and health outcomes among diabetes patients (10, 28, 29). However, few studies have specifically focused on the association between personal mastery and mortality among diabetes patients. Evidence has indicated that there are gender and ethnic variations in control beliefs and personal mastery and women, ethnic, and the less-educated tend to have lower sense of control than those of their counterparts (30). If personal mastery does have an association with lower risk of mortality, how might this protective effect differ across gender and immigration status. In this study, we used data collected from a representative sample of older Americans living with diabetes and linked them to their follow-up survival status. The purpose is to explore the association between personal mastery and all-cause mortality in a sample of diabetes patients from 2003 through the end of 2016.
 
 
Methods
Data and samples
    The data were taken from the Diabetes Study 2003, an off-year survey of the Health and Retirement Study (HRS) (31). Questionnaires were mailed to the 2,381 HRS respondents who responded “yes” to the question,” Have a doctor ever told you that you have diabetes?” in the 2002 HRS core interview survey. A total of 1,901 respondents (79.7%) returned their questionnaire. The survival information of the respondents was obtained from the 2016 cross-wave tracker file created by the HRS (31) . The death date of deceased participants in the HRS was obtained from both the tracker file. After excluding 122 cases with missing information on items covering the construct of personal mastery, a total of 1,779 participants were available for the current analysis. All identifiers in the data files had been removed before they became available on the HRS website for downloading. 
Measures
    After linking the participants of the 2003 Diabetes Study to the cross-wave tracker file, survival time since 2003 were calculated in years by subtracting 2003 from the year of death recorded in the tracker file for each of the deceased. If a participant died in 2003, a 0.5 year of survival time would be assigned to them. The observations for respondents who were still alive after the end of 2016 were treated as censored.
Personal mastery was measured by a 7-item scale with a 5-point Likert scale (“strongly disagree” = 1 to “strongly agree” = 5), developed by Pearlin (21). Three items in the scale were positively worded while the remaining 4 were worded negatively. To keep the direction of scoring in the same direction for all items, scoring was reversed for negatively-worded four items  “I have little control over the things happen to me,” “Sometimes I feel that I am being pushed around in life,” “I live life one day at a time and don’t think much about the future,” and “I often feel hopeless about the future.” A summary score of personal mastery was created by summing all 7 items. A higher score implies a stronger personal mastery. In order to reduce the number of cases dropped from current analysis, missing data for this measure were imputed for the 118 respondents who responded to 6 of the 7-item scale by multiplying the score for 6 items by a factor of 1.167 (the quotient of 7 divided by 6). However, the observations that missed 2 or more items on the 7-item scale were deleted.
    Covariates related respondent’s immigrant and ethnic status included in this study were participants’ nativity (1 = “born in the US”; 0 = “foreign-born”), Hispanic status (1= “Hispanics” and 0 = “non-Hispanic”), and African-American status (1 = “yes” and 0 = “otherwise”). Participants’ chronological age in the year 2003 was used in this study and gender was coded as (1 = male) and (2 = female). Health insurance is critical in gaining access to diabetes care. In the questionnaire, participants were asked: “at the time your diabetes was first diagnosed, were you covered by health insurance?” and “at the time your diabetes was first diagnosed, did you have any coverage for prescription drugs?” Only participants who answered both items with a “yes,” were classified as (having insurance = 1; otherwise = 0). Years of schooling was used an indicator for respondent’s SES. Finally, the Total Illness Burden Index (TIBI) was included to predict the mortality of the diabetes patients over the study period. The TIBI is a composite measure assessing comorbid illness in older adults using self-reported symptoms and has been proved a valuable tool in evaluating the severity of comorbidities (30).
Data analysis
    Proportional Hazard Models were used to evaluate the significance of each of the predictor variables, which was entered into the models hierarchically. First, respondents’ personal mastery score, insurance status, nativity (foreign or US-born), minority status (Hispanic or not and Black or not) were entered into the Cox regression models (Model 1),  then, chronological age, gender, years of schooling, and TIBI were entered to predict mortality over the 13-year period (Model 2). Furthermore, the distribution in several predictors included in the survival analyses significantly differ across gender. Separated analysis on the male and female sub-samples were also conducted to examine if there were different association patterns across gender.
Ethical consideration
    All identifiers of the survey respondents had been removed before the data became available for downloading from the HRS website. Therefore, there should be no ethical concerns in terms of breaching privacies or confidentialities.  
 
Results
    The descriptive characteristics of the sample and study variables are shown in Table 1. It consisted of 861 male and 918 female diabetes patients interviewed in 2003. By the year 2016, among the 1,779 diabetic participants, 948 of them had died (488 males and 460 females). Compared to male diabetes patients, females had more severe diabetes comorbidities, measured the TIBI, and a lower average score of personal mastery in life.
    When the male and female diabetes patients were combined (Model 1 of Table 2), having health insurance, being foreign-born, Hispanic status, and higher personal mastery score were significant predictors of mortality over the 13 years. One unit increase in personal mastery score would lead to a six-percent decrease in all-cause mortality (Hazard Ratio = 0.94, p < .001) over the study period. Being a Hispanic (Hazard Ratio = 0.76, p = 0.037) and those who had insurance coverage (Hazard Ratio = 0.82, p = 0.002) were also less likely to die (Model 1 of Table 2). In contrast, diabetes patients who were born in the US were more likely to die (Hazard Ratio = 1.31, p = 0.049) over the same period, compared to the foreign-born diabetes patients.
    In the Model 2 of Table 2, when other predictors were added, the significance of health insurance and Hispanic status in predicting mortality disappeared. As expected, age, gender, and education level were strongly associated with mortality. Older and male diabetes patients were more likely to die over time, so were the patients with lower levels of education (years of schooling) and more severe diabetes comorbidities. One year increase in schooling would have a two-percent reduction in mortality (Hazard Ratio = 0.98, p =  0.02) and one-unit increase in the TIBI score would have a one-percent increase in the risk of death (Hazard Ratio = 1.01, p < 0.001) over the study period.
When the same hierarchical modeling was applied to gender-specific sub-samples, different association patterns were identified for the male and female diabetes patients. Looking at the male sub-sample from Model 1 in Table 3, personal mastery was highly predictive of mortality and so were insurance and Hispanic status over the study period. But, when other covariates were entered, the full model (Model 2) in Table 3 revealed that the significance of health insurance and Hispanic status were absorbed and replaced with age and TIBI. However, personal mastery was still a significant predictor of mortality among male diabetes patients.
    Among the female diabetes patients, both insurance coverage and personal mastery score were significant predictors of mortality (Model 1 in Table 4). In the full model, being foreign-born, older, more years of schooling, and severity of diabetes comorbidity were significantly associated with survival. Meanwhile, personal mastery score was still a significant predictor of mortality over the 13-year period after taking other covariates into account.
    Finally, using the median of personal mastery score as a cut-off, survival functions for participants with higher and lower than median master scores were plotted in Figure 1. Participants with higher mastery scores had a better chance to survive to the end of 2016, when compared to those with lower mastery scores.

Variables	Male (N = 861)	Female (N = 918)
Number of cases died by end of 2016	488	460
Mean age in 2003	70.2 (8.3)	69.0 (9.5)
Years of schooling	12.1 (3.5)	11.1 (3.5)
Born in the United States (yes = 1)	791 (91.9)	819 (89.2)
Black (yes = 1)	127 (14.8)	206 (22.4)
Hispanic (yes = 1)	69 (8.0)	114 (12.4)
Insurance status when diagnosed with diabetes (yes = 1)	558 (64.8)	558 (60.8)
Personal mastery score	24.5 (4.4)	24.0 (4.4)
Total Illness Burden Index	33.6(18.3)	38.2 (18.3)

Predictors	Model 1	Model 2
Health insurance covering prescription (Yes = 1)	0.82 (0.72, 0.93)**	0.98 (0.86, 1.12)
Black (yes = 1)	0.86 (0.72, 1.01)	0.91 (.76, 1.09)
Born in the United States (yes = 1)	1.31 (1.00, 1.71)*	1.37 (1.05, 1.80)*
Hispanic (Yes = 1)	0.76 (0.59, 0.98)*	0.79 (0.60, 1.03)
Personal mastery score	0.94 (0.92, 0.95)***	0.96 (0.94, 0.97)***
Gender (female versus male)	---	0.80 (.71, .92)**
Age in 2003	---	1.08 (1.07,1.09)***
Years of schooling	---	0.98 (0.96, 1.00)*
Total Illness Burden Index	---	1.01(1.01,1.02)***
- 2 Log Likelihood (degree of freedom)	13500.47(5)	13077.38 (9)

Variables	Model 1	Model 2
Health insurance covering prescription (Yes = 1)	0.80 (0.67, 0.96)*	0.94 (0.78, 1.14)
Black (yes = 1)	0.86 (0.66, 1.12)	0.93 (0.71, 1.22)
Born in the US (yes = 1)	1.20 (0.82, 1.75)	1.14 (0.77, 1.68)
Hispanic (yes = 1)	0.67 (0.45, 0.99)*	0.74 (0.49, 1.11)
Personal mastery score	0.93 (0.91, 0.95)***	0.95 (0.93, 0.98)***
Age in 2003	---	1.08 (1.07,1.09)***
Years of schooling	---	0.99 (0.96, 1.02)
Total Illness Burden Index	---	1.02 (1.01,1.02)***
- 2 Log Likelihood (degree of freedom)	6208.92 (5)	6009.61 (8)

Variables	Model 1	Model 2
Health insurance covering prescription (Yes = 1)	0.82 (0.68,0.99)*	1.00 (.83, 1.21)
Black (yes)	0.91 ( .72, 1.13)	0.89 (0.70,1.13)
Born in the US (yes = 1)	1.42 (0.98, 2.07)	1.57 (1.08, 2.27)*
Hispanic (Yes = 1)	0.88 (0.63, 1.24)	0.78 (0.54, 1.13)
Personal mastery score	0.94 (0.92, 0.96)***	0.96 (0.94, .98)***
Age in 2003	---	1.08 (1.07,1.10)***
Years of schooling	---	0.96 (0.93, 0.99)**
Total Illness Burden Index	---	1.01 (1.01,1.02)***
- 2 Log Likelihood (degree of freedom)	5967.18 (5)	5751.32 (8)

1. Laiteerapong N, Huang ES. Diabetes in older adults. in: diabetes in America [Internet]. 3rd Edition. National Institute of Diabetes and Digestive and Kidney Diseases; 2018. p. 16-1–26. Available from: https://www.niddk.nih.gov/about-niddk/strategic-plans-reports/diabetes-in-america-3rd-edition
2. Cho NH, Shaw JE, Karuranga S, Huang Y, da Rocha Fernandes JD, Ohlrogge AW, et al. IDF diabetes atlas: global estimates of diabetes prevalence for 2017 and projections for 2045. Diabetes Research and Clinical Practice. 2018; 138: 271–81.
3. Johnson NB, Hayes LD, Brown K, Hoo EC, Ethier KA. CDC national health report: Leading causes of morbidity and mortality and associated behavioral risk and protective factors-United States, 2005-2013. Morbidity and Mortality Weekly Report. 2014; 63(4): 3-26.
4. Menke A, Casagrande S, Geiss L, Cowie CC. Prevalence of and trends in diabetes among adults in the United States, 1988-2012. The Journal of the American Medical Association. 2015; 314(10): 1021-9.
5. Caspersen CJ, Thomas GD, Boseman LA, Beckles GLA, Albright AL. Aging, diabetes, and the public health system in the United States. American Journal of Public Health. 2012; 102(8): 1482–97.
6. Rosenquist KJ, Fox CS. Mortality trends in Type 2 diabetes. In: Diabetes in America [Internet]. 3rd Edition. National Institute of Diabetes and Digestive and Kidney Diseases; 2018. p. 36 -1–14. Available from: https://www.niddk.nih.gov/about-niddk/strategic-plans-reports/diabetes-in-america-3rd-edition
7. Cheng YJ, Kanaya AM, Araneta MRG, Saydah SH, Kahn HS, Gregg EW, et al. Prevalence of diabetes by race and ethnicity in the United States, 2011-2016. The Journal of the American Medical Association. 2019; 322(24): 2389-98.
8. Engelman M, Leafia ZY. The immigrant health differential in the context of racial and ethnic disparities: the case of diabetes. Advances in Medical Sociology. 2019; 19: 147–71.
9. Miech RA, Kim J, McConnell C, Hamman RF. A growing disparity in diabetes-related mortality. American Journal of Preventive Medicine. 2009; 36(2): 126–32.
10. Saydah SH, Imperatore G, Beckles GL. Socioeconomic status and mortality: contribution of health care access and psychological distress among U.S. adults with diagnosed diabetes. Diabetes Care. 2013; 36(1): 49–55.
11. Goldman, Noreen. Will the Latino mortality advantage endure?. Research on Aging. 2016; 38(3): 263–82.
12. McDonald JA, Paulozzi LJ. Parsing the paradox: Hispanic mortality in the US by detailed cause of death. Journal of Immigrant and Minority Health. 2019; 21(2): 237–45.
13. Mehta NK, Elo IT, Engelman M, Lauderdale DS, Kestenbaum BM. Life expectancy among US-born and foreign-born older adults in the United States: estimates from linked social security and medicare data. Demography. 2016; 53(4): 1109–34.
14. Boen CE, Hummer RA. Longer—but harder—lives?: the Hispanic health paradox and the social determinants of racial, ethnic, and immigrant–native health disparities from midlife through late life. Journal of Health and Social Behavior. 2019; 60(4): 434–52.
15. Lariscy JT, Hummer RA, Hayward MD. Hispanic older adult mortality in the United States: new estimates and an assessment of factors shaping the Hispanic paradox. Demography. 2015; 52(1): 1–14.
16. Treas J, Gubernskaya Z. Immigration, aging, and the life course. In: Ferraro K, George L, Editors. Handbook of aging and social sciences. 8nd ed. Elsevier; 2016.
17. Pearlin LI. The life course and the stress process: some conceptual comparisons. The Journal of Gerontology Series B Psychological Sciences and Social Sciences. 2010; 65B(2): 207–15.
18. Thoits PA. Stress and health: major findings and policy implications. Journal of Health and Social Behavior. 2010; 51: S41–53.
19. Mitchell UA, Ailshire JA, Brown LL, Levine ME, Crimmins EM. Education and psychosocial functioning among older adults: 4-year change in sense of control and hopelessness. The Journal of Gerontology. Series B, Psychological Sciences and Social Sciences. 2018; 73(5): 849-59.
20. Pearlin LI, Lieberman M, Menaghan EG, Mullan JT. The stress process. Journal of Health and Social Behavior. 1981; 22(4): 337–56.
21. Szabó Á, Klokgieters SS, Kok AAL, Tilburg TGV, Huisman M. Psychological resilience in the context of disability: a study with Turkish and Moroccan young-old immigrants living in the Netherlands. The Gerontologist. 2020; 60(2): 259–69.
22. Savla J, Wang Z, Zhu J, Brossoie N, Roberto KA, Blieszner R. Mastery and longevity in spousal caregivers of persons with dementia. The Journal of Gerontology. Series B, Psychological Sciences and Social Sciences. 2020; 75(7): 1558-62.
23. Cooper R, Huisman M, Kuh D, Deeg DJH. Do positive psychological characteristics modify the associations of physical performance with functional decline and institutionalization? findings from the longitudinal aging study Amsterdam. The Journal of Gerontology. Series B, Psychological Sciences and Social Sciences. 2011; 66(4): 468–77.
24. Lachman ME, Firth KM. The adaptive value of feeling in control during midlife. In: How healthy are we? Chicago: University of Chicago Press; 2004. p. 320–49.
25. Surtees PG, Wainwright NWJ, Luben R, Khaw KT, Day NE. Mastery, sense of coherence, and mortality: evidence of independent associations from the epic-Norfolk prospective cohort study. Health Psychology: Official Journal of the Division of Health Psychology, American Psychological Association. 2006; 25(1): 102–10.
26. Infurna FJ, Ram N, Denis G. Level and change in perceived control predict 19-year mortality: findings from the Americans' changing lives study. Developmental Psychology. 2013; 49(10): 1833–47.
27. Roepke, SK, Grant I. Toward a more complete understanding of the effects of personal mastery on cardiometabolic health. Health Psychology. 2011; 30(5): 615–32.
28. Yi-Frazier JP, Smith RE, Vitaliano PP, Yi JC, Mai S, Hillman M, et al. A person-focused analysis of resilience resources and coping in diabetes patients. Stress Health. 2010; 26(1): 51–60.
29. Lachman ME, Neupert SD, Angrigoroaei S. The relevance of control beliefs for health and aging. In: Handbook of the psychology if aging. Amsterdam: Elsevier/Academic Press; 2011. p. 175–90.
30. Servais M. Overview of HRS public data files for cross-sectional and longitudinal analysis [Internet]. Ann Arbor, Michigan: Institute for Social Research, University of Michigan; 2010. Available from: https://hrs.isr.umich.edu/sites/default/files/biblio/OverviewofHRSPublicData.pdf
31. Kerr, EA, Smith, DM, Kaplan, SH, Hayward, RA. The association between three different measures of health status and satisfaction among patients with diabetes. Medical Care Research and Review. 2003; 60(2): 158–77.
32. Lin H-W, Tung H-J. Using changes in life satisfaction and health to predict the survival status among older men and women in Taiwan. Journal of Women & Aging. 2013; 25(3): 227–41.
    
33. Ross CE, Mirowsky J. Gender and the health benefits of education. Sociol Q. 2010; 51(1): 1–19.
34. Hummer RA, Hayward MD. Hispanic older adult health & longevity in the United States: current patterns & concerns for the future. Daedalus. 2015; 144(2): 20–30.
35. Mirowsky J, Ross CE. Education, social status, and health. 1th ed. New York: Routledge; 2003.