Depression is one of the most common illnesses worldwide, with a prevalence of 3.8% among the general population and 5.7% among older adults aged 60 and over [1]. Depression is a severe health condition, especially if recurrent and of moderate to extreme intensity, which can significantly affect the person's quality of life [2]. Depression becomes a burden for the person, family, and society due to lowered productivity and higher healthcare costs [3, 4]. The prevalence of depression in older adults may be underestimated because it tends to be unrecognized and untreated compared to younger individuals [2]. In Indonesia, the number of older adults has increased over the past decades, with an estimated 16 million older adults (6.1% of the population) in 2019 [5]. This sizeable aging population faces various challenges, including disability, multiple chronic diseases, frailty, and especially depression [2, 6, 7]. According to Indonesia's basic health research, the prevalence of mental health disorders in Indonesia has increased from 6% in 2013 to 9.8% in 2018. In the same study, depression among persons older than 15 years old in 2018 was estimated at 6.1%, with only 9% receiving treatment for the condition in Indonesia [8].

Depression prevalence was lower in individuals with higher socioeconomic status, better self-reported health, more social capital, and functional independence in multiple studies [9-14]. In addition, engaging in physical activity was found to act as a protective factor against depression [15-17]. Various studies have shown that depression increases with multimorbidity, frailty, and dementia. Chronic diseases, including hypertension, diabetes, stroke, heart disease, kidney disease, and liver disease, confer a risk of depression over time [9, 10, 18]. People with visual and auditory impairment also have a higher prevalence of depression than the average population [19-21].

There is limited research on depression and its associated factors among the older adult population in Indonesia, and some studies only involve small samples [22-24]. Due to increasing trends and the high disease burden, more studies regarding depression in Indonesia have become necessary. The Indonesia Family Life Survey provides large representative data from various regions in Indonesia; data analysis on older people from these surveys might help create better-targeted strategies to prevent and improve depression among older Indonesian adults. This study aims to investigate factors associated with depression among older adults in Indonesia.

The prevalence of depression in our study was high (16.3%). This number was higher than that reported in the general population from the 2018 Indonesia Basic Health research (6,1%) [8]. Studies of older adults from a Portuguese national cohort (11.8%), the South Africa nation-wide survey (4%), and a meta-analysis by Lim et al. (12.9%), all showed lower prevalence. In comparison, studies from a nationwide survey in South Korea (27.8%) and a rural area in Thailand (18.5%) showed higher prevalence [9, 10, 35-37]. Different assessment methods of depression may cause variations between these studies, and alternatively, different cultures and sampling methods could be responsible for discrepancies. Both studies in South Korea and Thailand used Geriatric Depression Scale (GDS), but a different cut-off was established, with the South Korean study using ≥8 as a cut-off while the Thailand study using ≥13. This may explain the higher prevalence of depression in the South Korean study. The South African study used the World Mental Health Survey version of the Composite International Diagnostic Interview, and Indonesia basic health research used the Mini International Neuropsychiatry Interview; both showed lower prevalence than other studies. A meta-analysis by Lim et al. also showed that self-reported assessments tend to have a higher score than clinical diagnostic interviews. Diagnostic interviews adopted stricter criteria to differentiate individuals with clinical depression from those with only depressive symptoms.

Among demographic factors, there was an association with living in Java and other regions, while being older was not associated with depression. Many studies have shown that older age confers a higher risk, but this was not found in a meta-analysis by Maier et al. [18]. The insignificant result between older age and depression may be explained by the fact that older age is related to declining physical health and disability, which results in depression, and is not age-independent [10]. However, there was no sex difference in depression in our study. However, many studies have shown that older women tend to have depression more commonly than men [18, 36]. Given bio-psychosocial aspects, women have a higher concentration of serotonin that plays a role in anxiety and stress; they also tend to have a dysregulated hypothalamic-pituitary-adrenal axis in response to stress compared to men; women are also more likely to have experienced sexual and emotional abuse, and in the Asian and many other cultures, women’s role is restricted to home duty and child care, which are associated with depression. Women were also reported to be more likely poorer or have less control over economic resources in the household [5].

Depression was also more prevalent in participants with functional dependency as measured by IADL, having moderate and low socioeconomic status, low life satisfaction, and being self-perceived as unhealthy. These findings are consistent with many studies, including a meta-analysis [9, 16, 38, 39]. Having a functional disability could increase the prevalence of having clinical depression and worsen the existing depression [40]. The declining physical function creates uncertainty and hopelessness by being dependent in older adults [41, 42]. Being poor by the subjective judgment of the participant and/or from income/wealth aspects (objective poverty) predisposes older adults to depression. Having a low income may lead to poor living conditions and multiple serious life events, which increase the higher prevalence of depression. Men tend to be more affected by low socioeconomic status than women because, traditionally, men are the head of the family and responsible for family needs [42]. On the other hand, women usually outlive men and then live in poorer households, so they are more likely to experience poverty in later life [43]. However, as stated, we found no gender differences in depression. In a study done by Sasaki et al., subjective economic status had a stronger association than objective economic status. Some reasons for this phenomenon were health disparities and negative psychological consequences (stress-related dysregulation of the hypothalamic-pituitary-adrenal axis) due to differences in subjective economic status. Subjective economic status might represent a cognitive estimation of current economic status and extend it to the past and future prospects [44]. Having dissatisfaction in life and self-perception as being unhealthy probably have a bidirectional relationship with depression.

Having a low physical activity or sedentary life has a bidirectional relationship with depression. Depressive individuals tend to be less physically active, even formerly active individuals [17]. Having an active lifestyle has been associated with fewer depressive symptoms in other studies. The relation between depression and physical activity may be mediated by several variables, such as social relations and engagement, chronic conditions, and functional ability [45]. Older adults who have a disability may also be unable to do certain activities. Despite these findings, our study did not show a significant association between physical activity and depression. A meta-analysis by Huang et al. showed that older adults with dementia had an increased probability of depression compared to those with good cognitive function [46]. However, a systematic review by Maier et al. concludes that it is difficult to extrapolate these associations due to heterogeneous results between the studies [18]. In a study by Carta et al., performing moderate exercise improves cognitive function in healthy older adults. Due to bidirectional relationships between physical activity and cognitive function with depression, improvement in either variable might improve depression in older adults [47].

Insomnia was associated with depression in our study. A longitudinal study by Li et al. also confirmed our findings [48]. Two prospective cohort studies by Lee et al. and Baglioni et al. found that insomnia significantly increased depression in nondepressed older adults at baseline in the subsequent year [49, 50]. In the study by Lee et al., persistent insomnia increased depression risk rather than intermittent insomnia. A meta-analysis by Maier et al. also showed that insomnia, especially early insomnia, was significantly associated with depression [18]. Our assessment of insomnia has a limitation in that we could not differentiate insomnia into multiple types for further analysis. Insomnia and depression have a complex relationship. Depression diagnoses usually have criteria for sleep disturbances, while insomnia could cause alterations in emotional response, resulting in depression. Other studies also show that insomnia endorses maladaptive sleep beliefs that increase the individual’s sense of hopelessness, which also results in depression [51].

Our findings regarding stroke showed an increased prevalence of depression significantly, which has also been demonstrated by other studies [52-55]. The prevalence of post-stroke depression was around 31.1%, with more women affected by it [54]. The pathogenesis of poststroke depression itself is multifactorial, involving complex biological, social, and psychological mechanisms [56]. The higher number of depression in stroke patients may also be mediated by long-term hospitalization, isolation, and problems adapting to new conditions, such as significant disability and communication difficulties [55]. A longitudinal study by Thomas et al. showed that two-thirds of poststroke patients remained depressed at follow-up between 7 to 13 months and became even more depressed [53]. Lower perceived control of recovery and difficulty in communication contributed to this phenomenon. One of the problems is that depressive symptoms usually also involve somatic complaints, which can also be affected by the presence of stroke, so it is difficult to determine an accurate diagnosis [55].

In our study, more older people with arthritis suffer from depression than those without arthritis. According to Mortality and Morbidity Weekly Report (MMWR), in 2018, around 22.7% of US adults were diagnosed with arthritis, and 22.5% and 12.1% were suffering from anxiety and depression [57]. Based on age-specific prevalence, anxiety and depressive symptoms were more common in those between 18-44 years than in people over 65 years in our study. Visual and auditory impairment was significantly associated with depression in multiple studies, but in our study, only auditory impairment was associated with depression [19, 58, 59]. A study by Chou et al. found that visual impairment has a more robust impact than hearing impairment because it tends to be more obvious; it is also perceived as worse than hearing loss and wider restrictions [19]. A national cohort study by Kim et al. showed that mild to moderate hearing loss has little impact on depression, but in our study, we did not classify the hearing impairment because it was self-reported [60]. In treating patients with visual or hearing impairment, the healthcare providers often do not address the mental health aspects, including depression.

The prevalence of depression among older adults in Indonesia is still very high. Many factors contribute to depression, such as living in Java and other regions, being dependent, having moderate to low subjective economic status, having no life satisfaction, being self-perceived as unhealthy, and having experienced stroke, arthritis, hearing impairment, insomnia, dependency (IADL) and falls. The latter two could be indicative of frailty, a common syndrome in older people and associated with many of the health variables which have not been entered in multivariate regression analyses. Despite the inability to determine causality and the limitation that most of the assessments were self-reported, strong associations were found in this study. Predictors investigated in this study may help prevent and improve depression in Indonesian older adults, especially those living in Java. Improvement in healthcare, especially in the prevention and rehabilitation of stroke, arthritis, and hearing impairment, concurrent with early detection of depression in these chronic conditions, may help create a better quality of life among Indonesian older adults.