This was a cross-sectional study conducted among older adults aged 60 years and above in four areas in Selangor, Malaysia. Two rural and two urban areas were selected randomly from the list of housing areas or villages. This study had been ethically approved by the Research Ethics Committee of the National University of Malaysia (FF-2018-532).
The study was conducted from 1 December 2018 to 30 April 2020, involving 1204 respondents who were randomly selected. Respondent information sheets and consent forms were given to the participants after the study has been thoroughly explained. The method of the interview was done in a quiet environment, face-to-face and initiated immediately after the respondents returned the consent forms. The interview was conducted by trained research assistants, ranging from 40 to 50 min to be completed. The study tools use the Bahasa Malaysian version of the Japan Gerontological Evaluation Study (JAGES-BM) questionnaire, with the aim to study the determinants of healthy ageing (12). The questionnaire contained multi-dimensional variables, including health factors, psychological factors, functional factors and social determinants of health (supplementary file).
The inclusion criteria are Malaysians who are 60 years old and above and can converse in the Malaysian language. All respondents were screened for the possibility of reduced cognitive function using the Abbreviated Mental Test. Those who scored less than 7 were excluded from the study.
The self-perceived happiness was subjectively determined by the respondent using a single question with a scale rated from 0 to 10 points. Thus, it was measured through the following question: To what degree do you feel you are currently happy? (Score ‘0’ for ‘Very unhappy’ and ‘10’ for ‘Very happy’). Based on this scale, happiness was defined with a self-rated score of 7–10 points, whereas unhappiness, 0–6 points. This level of the cut points was finalised using the previous survey among the elderly in Japan on their average happiness score, which was 6 points (11).
Sociodemographic variables included age, sex, marital status and location. Age was categorised into the following three categories: 60–74, 75–84 and >85 years. On the basis of the article by Steptoe et al., we were able to include five components, namely marital status, socioeconomic, education, social network and physical activities. Marital status was categorised into married living together, married living separately, widowed and divorced and never married. Location was divided into rural and urban.
Socioeconomic status includes household income, education level and current employment status. For household income classification, we were using the Income Structure 2019 by the Department of Statistic Malaysia. B40 refers to the base group or bottom 40%, with a monthly household income of less than RM4,850, whereas M40 refers to the middle-class group or middle 40%, with earnings of between RM4,851 and RM10,959 per month. Conversely, T20 is an upper-class group or top 20%, with monthly earnings of more than RM10,959(12).
In the study, we used social engagement and social support in measuring social network. Social network was measured with the frequency of engagement in group activities. The activities that have been measured were religious, volunteer, sports or club, hobbies, community meetings and political meetings or events. Regular participation of at least once a week is considered active engagement. Meanwhile, social support was measured by asking respondents if they have someone to talk to regarding concerns or complaints (emotional support) and who looked after them when they felt sick and were confined for a few days (instrumental support). These variables were also used by Park et al. in defining social support (10). We also asked if they listen to someone’s concerns or complaints and they look after someone when he/she is sick and confined for a few days.
Physical measurement comprises the body mass index (BMI) and handgrip strength (HGS). Weight and height were measured twice for BMI. The Malaysian BMI classification was used as a reference (13), in which underweight is defined as BMI < 18.5 kgm2, normal weight is BMI = 18.5–22.9 kg/m2, overweight is BMI > 23 kg/m2, pre-obese is BMI = 23.0–27.4 kg/m2, obese I is BMI = 27.5–34.9 kg/m2, obese II BMI = 35.0–39.9 kg/m2 and obese III is BMI > 40 kg/m2. HGS was measured using a handgrip dynamometer with the dominant hand (T.K.K. 5001 GRIP-A from Takei Scientific Instrument Co. Ltd. Japan). Each of the respondents was measured twice, and the mean was taken as reading.
Physical activities for older adults were evaluated through questions regarding the frequency of moderate exertion activities such as brisk walking, golf, dancing, farming, gardening or car washing. This later was translated to an average of 150 min of moderate-intensity aerobic physical activity through the week based on the WHO recommendation (16).
Respondents were asked about self-rated health status who were given four options, namely excellent, good, fair and poor. In addition, we used the Japanese version of the 15-item Geriatric Depression Scale (GDS) to evaluate depressive symptoms in older adults. The GDS score ranges from 0 to 15, with higher scores that indicate more severe symptoms. The score scale cut points namely those who were having GDS scores of 1–4 as normal, 5–8 as mild depression, 9–11 as moderate depression and ≥12 as severe depression.
Respondents were also asked about the diseases or comorbidity that they have. These include stroke, heart disease, diabetes, hyperlipidemia, respiratory disease, gastrointestinal, liver or gallbladder disease, kidney or prostate gland disease, musculoskeletal disease including osteoporosis and arthritis, traumatic injury, cancer, blood or immune disease, depression, dementia, Parkinson’s disease, eye and ear disease, malaria and HIV infection and gynaecological problem.
To determine the association among the study variables, chi-square and independent t-test were conducted. Simple logistic regression (SLogR) and multiple logistic regression (MLogR) were used to calculate the crude odds ratio and the adjusted odds ratio, respectively. A p-value of <0.05 was considered significant in all tests. Analyses were conducted using IBM SPSS version 21.0 (IBM Corp., Armonk, NY, USA).