Factors of Successful Aging among Pre-Retirement Public Servants in Klang Valley, Malaysia

Background: Following the global trend,Malaysia is moving towards an aging population. With the change in age ratio, there will be more age-related diseases and challenges that need to be managed appropriately. This study aimed to determine the prevalence of successful aging (SA) among pre-retirement Malaysian public servants and the predictors. Methods: A cross-sectional study was carried out among 1,064 pre-retirement public servants (50-60 years old) working in nine federal government agencies within the Klang Valley. Multistage sampling was applied with simple random sampling for selecting four out of 23 ministries and two agencies under each ministry. Purposive sampling was used for selecting the state and respondents. The respondents completed a self-administered questionnaire and their cognitive function was assessed using the Mini-Mental State Examination (MMSE). SA was defined as being able to fulfil all three criteria i.e. absence of six major chronic diseases and having both good physical and psycho-cognitive functions. Results: The prevalence of SA was 37.5%. Multiple logistic regression showed that the factors with higher odds of having SA were younger age (50–54 years old) (adjusted odds ratio [aOR] 1.32, 95% confidence interval [CI] 1.01-1.73), being physically active (aOR 1.39, 95%CI 1.05 -1.84), non-obese (aOR 2.14, 95%CI 1.52-3.02) and good social support (aOR 1.78, 95%CI 1.30-2.43). Conclusion: A minority of public servants in this study have SA. Employers in various agencies should play roles in promoting SA aiming for healthy behaviours and providing healthy working environments.


Introduction
Many countries are now experiencing the global phenomenon termed the so-called 'aging population',whereby the older populationhas begun to outnumberthe young population aged under 15 years. The number of older people has been projected to rise more than double from 841 million in 2013 to more than 2 billion in 2050.
Meanwhile, Malaysia has been projected to have an aging population by 2020 [1].With the change in the age ratio, the government will facea higher burden of old age-related disabilities and socio-economic consequences that need to be managed properly,including in terms of healthcare services.
One strategy for reducing the impact is to ensure thata high proportion of the elderly, if not all,achieve successful aging (SA).
This concept of SA was popularized by Rowe and Kahn (1987) [2] , who applied biological theory in defining SA which comprisesthe absence of chronic disease and good physical, mental and social wellbeing among older adults.Meanwhile, various variables have been studied as predictors of SA including behavioural risk factors, physical functioning, social engagement, daily activities, cognitive functions, spiritual aspects [3]and genetics or hereditary factors [4].
A few studies in Malaysia have investigated this aspect, but none involved the younger population.
Therefore, this study was undertaken to determine the prevalence and associated factors of SA among pre-retirement age group public servants in the Klang Valley, Malaysia. It is hoped that the findings canassist public health personnel in carrying out strategies specifically for the pre-retirement age population to prepare it for SA.

Methods
This cross-sectional study was conducted from July until December 2018and involved 1,064 public servants in thepre-retirement aged group (aged 50-60 years) working at nine government agencies within the Klang Valley (which includes Selangor state and the Kuala Lumpur federal territory).
Multistage sampling was applied, where simple random sampling was used for selecting five of 23available ministries and subsequently for selecting the agencies to represent each ministry.
Respondents were selected via purposive sampling until the required number was fulfilled.
The respondents were asked to complete a self-administered questionnaire and their cognitive function was subsequently assessed using the Mini-MentalState Examination (MMSE). Some of the questions were adapted from previous studiesand had undergone translation from the original English into Malay, as well as pre-testing, validity and reliability testingprior to usage. Cronbach's alpha test wasused and the minimum acceptable value was 0.7. Factor loadings for all the items in the questionnaire were > 0.4.
The dependent variable was operationalized based Rowe and Kahn's model (1987). Respondents were categorized as SAonly if they fulfilled all three criteria:1) no major chronic disease (self-reported to not have any of the following sixdiseases,i.e. diabetes mellitus, hypertension, stroke, chronic lung diseases, cancer and heart problems); 2) good physical functioning; and 3) good psycho-cognitive functioning.
For the physical function assessment, the respondents were required to rate their difficulty in performing nine physical tasks without using supportive equipment: walking 400 meters (1/4 mile);walking up 10 steps without resting(climbing)and standing for 2 hours; sitting for 2 hours; bending, bowingor kneeling; reaching or reaching something above the head; using fingers to hold or handle small objects; lifting or carrying an object weighing 4.5 kg; and pushing or pulling large objects. Ratings were based on a scale, i.e. unable to do it directly, very difficult, quite difficult,slightly difficult and not difficult at all. Those who answered not difficult at all, slight difficulty or quite difficult were categorized as having good physical function [5].
The respondents' psychological function was assessed using the Malay version 21-item Depression Anxiety Stress Scale (DASS-21) questionnaire and scoring [6]. Those with scoreresults that fell under normal and mild categorieswere deemed to have good psychological function.For assessing cognitive function,MMSE scores of ≥ 23 were considered to indicategood cognitive function [7].
In the present study, the behavioural variables were smoking status, alcohol consumption, substance usage in the last 12 months, physical activity, daily consumption of fruits and vegetables and body weight. The body mass index (BMI) was calculated as weight (kg)/height squared(m 2 ). Respondents who exercised at least 150 minutes per week (moderate intensity such as 30 minutes brisk walking at least five times per week) were categorized as physically active [8].
Respondents who answered the 8-itemDuke-UNC Functional Social SupportQuestionnaireusing a 5point Likert scale (ranging from 1 = much less than I would like, to 5 = as much as I would like) and who scored > 30 were considered to have good social support [9].
Respondents were considered to have barriers to healthcare if they answered 'yes' to at least one of the 11 items related to the cost of treatment, transportation problem, cost of transport, inadequate drugs or equipment at health facilities,inadequate health personnel skills, by health personnel, havingother personal commitments or work, not knowing where to go, thinking they were 'not sick enough', and denial of healthcare [8].
Respondents were assessed on their agreement with the statement 'I am well prepared for retirement' using a 4-point Likert scale (definitely false, mostly false, mostly true and definitely true).
Those who answered definitely true or mostly true were considered to have pre-retirement preparation.
Bivariate analysis was conducted using the Pearson chi-square test and the Yates correction test for certain variables. The association between predictors and SA was assessed using multiple logistic regression. The level of statistical significance for this study was p-value < 0.05. Foreigners, those who were already on long medical leave and Ministry of Health employees were excluded from this study. Table 1 shows the characteristics of the respondents. Of 1,064 respondents, the majority was married (90.9%), Malay (78.7%), female (72.9%), aged 50-54 years old (64.5%) and professional (74.2%) with bachelor degree qualification (55.4%). The mean age was 53.6±2.7 years. Most respondents had a monthly income of RM 5,600.00 and above (634, 59.6%) (high-income group). For the behavioural aspect, many respondents were physically inactive and had inadequate daily fruits and vegetables intake. In fact, 41.1% of the respondents were overweight. Eight hundred and thirteen respondents (76.4%) perceived that they had good social supportand almost all respondents claimed to have preretirement preparation (1,024, 96.2%). Only 57 respondents (5.4%) had experienced barriers to obtaining healthcare.  Fig. 1.

Results
Similarly, multiple logistic regression analysis proved that these variables were significant predictors for SA(

Discussion
Successful aging is not a new concept among researchers and many of them have continued to study it to seek the most ideal definition and to identify any additional factors related to SA.Here, we aimed to determine the prevalence and associated factors of SA among public servants in the Klang Valley.
As we had adapted the Rowe and Kahn's model (1987), thuswe compared our findings with previous studies that applied a similar model. In terms of prevalence, some studies have shown that 10.1% of respondents aresuccessful agers [10], andthe prevalence may be up to 50% or more [11]. Meanwhile, the prevalence of SA in the present study was concordant with those studies and within that range, i.e. 37.5%. When comparing our findings with previous local studies, the prevalence of SA among the older people (aged 60-80 years) was much lower at 13.8% [12]. This demonstratesthe decreasing trend in the prevalence of SA as age increases. This is supported by a study that involved the younger age group population as the respondents. In that study, the prevalence of SA was 50.1% for those aged 50-54 years, 46.2% for those aged 55-59 years, 42.0% for those aged 60 years and above and only 37.2% for those aged 65 years and above [11]. The differences in the findings could be due to the difference in the theories or models used to study SA, differences between age groups or studied populations, as well as influences from the culture.
In agreement with previous studies, we found that younger age, non-obese, being physically active and good social support were the four significant predictors for SA.Having normal BMI, regular exercise and social support distinguished people who continued to age successfully 4 years later from those who did not have it [13].
Apart from gender and occupation, age was a predictor for SA in one biomedical model [(14)].Even though there was a large-scale review of SA among younger people, some studies produced the opposite findings. Despite experiencing late-life disability, some people still felt that they had aged successfully. This is because they tend to use adaptation and coping strategies to align their perception of SA with their experiences [15]. This is supported by a study that documented that older age was associated with lower likelihood of objectively-defined SA), but with a greater likelihood of self-ratedSA [16].
In the present study,non-obese respondents had 2.1 times higher odds of having SA, which is the highest odds ratio of all the predictors analysed. This association is consistent with that of a prior study [17].Obesity has a negative impact on health-related quality of life. Even a small amount of weight loss (5-10% of the initial weight) is beneficial for both young and old people to prevent the adverse effects of obesity [18]. Thus, optimizing body weight and dietary intake are proposed as nutritional strategies towards reducing the risk of age-related chronic diseases.
Previous studies have indicated that good social support is a significant predictor for SA [19].As predicted, our results revealed similar findings. Among the various possible predictors analysed, good social support was the second strongest predictor for SA. Good social support can be obtained from a spouse, family members, relatives, friends or neighbours. Higher levels of social support were beneficial for preventing depressive symptoms, thus maintaining or improving life satisfaction [20],which has a significantly positive impact on successful later years [21]. People who are able to visit their relatives and friends are more likely to be successful agers [22].
Similarlyto previous studies, being physically active was a significant predictor for SA in the present study. Physically active respondents are more likely to be rated as successful agers [23]. Being physically active continued to be an important significant determinant of self-perceived health into very late adulthood [24].This is very important for improving balance, mobility and for preserving independence [25].
Healthy lifestyles, including during midlife, have been proven to be associated with good overall health during aging.However, we could not determinethe associations between remaining sociodemographic and behaviour variables and SA in the present study. This is inconsistent with previous studies [26][27].A possible explanation for the discrepancy is that the majority of our respondents were female and it is not the norm for Malaysian women to smoke or drink alcohol due to socio-cultural environmentalinfluence [28].
We observed that the most difficult criteria of SA for our respondents to fulfil was 'having no major chronic disease', which put them into the non-successful agers category.Therefore, efforts should be taken to highlight the importance of preventing chronic diseases to everyone, even though studies had agreed on the possibility for chronic illness and SA to coexist within the same individual [29][30].
People should be clear that we are not saying this to criticise or stigmatize whomever does not meet the Rowe and Kahn's definition of SAbut to emphasizethat every pre-retirement age group adult should have awareness and knowledge of their own health or medical status. Subsequently, action should be taken to improve it by having regular health screenings and practicing healthy lifestyles.
Employers play a role in providing a healthy working environment and carrying out promotional activities on SA. Apart from the roles of health professionals, the goals towards higher prevalence of SA in the future can be achieved if everyone plays their part in improving their wellbeing.
Involving the pre-retirement population groupas respondents could yield better understanding of SA.
Based on the findings of this study, more focused intervention can be implemented in this group.Meanwhile, the limitation of this study is that people with underlying medical problems who were already on long medical leave and who had retired early from government service had already been excluded from the outset at the sampling stage. Therefore, there is possibility of over or underestimate of the total percentage successful agers.

Conclusion
The prevalence of SA in this study is 37.5% and the associated factors identified were younger age, non-obese, being physically active and good social support.A health-conscious attitude and the practice of healthy lifestyles should be inculcated among pre-retirement public servants to prepare them for SA. Apart from that, it would be best to apply bio-psychosocial theories, spiritual aspect and layperson's perspective on aging in studying SA to obtain a broader picture in thisregard in general population.