Multiple trajectories of depressive symptoms among Chinese in middle and late life: Characterization and risk factors

Abstract

Background:

Depressive symptoms, which are continuously changing, are an essential manifestation of depression and can increase the risk of mental disorders and other diseases. Because the causes and cures for depression have not yet been identified, finding the characteristics, and risk factors of depressive symptom trajectories can help us identify at-risk populations early and reduce the related public disease burden.

Aims:

Herein we aimed to figure out the specific manifestations of depressive symptom trajectories among Chinese adults, explore the risk profiles of trajectory groups with higher depression burdens, and test the longitudinal associations between blood biomarkers with depressive symptoms.

Methods:

Trajectories of participants’ depressive symptoms measured by the Center for Epidemiologic Studies Depression scores were modeled with growth mixture models from 2011 to 2018. Multinomial logistic models tested associations of baseline covariates with trajectories. Generalized estimating equations were used to explore the longitudinal associations between blood data and depressive symptoms in two waves from 2011 to 2015.

Results:

Among the sample of 5,641 individuals aged 40 or over, four heterogeneous depressive symptom trajectories were defined: stable-low, high-decrease, stable-high, and low-increase. At baseline, demographic factors and health statuses such as gender, education, income, and self-reported health status were associated with trajectories. A significant association was found between high-density lipoprotein and depressive symptoms.

Conclusions:

These findings provide clues for predicting and identifying adults with elevated depression burdens in middle and late life and may facilitate the development of targeted preventive strategies for this population.

Keywords

Depressive symptoms longitudinal study risk factors trajectories

Introduction

Depressive symptoms are the main parts of depression, a chronic mental disorder that could affect individuals’ moods, thoughts, and physical health (Cui, 2015). Results from a meta-analysis suggested that the overall estimation of the current and lifetime prevalence of major depressive disorder in China was 1.6% and 3.3% (Gu et al., 2013). Depression has led to a heavy psychosocial and economic burden on families and society in China (Lu et al., 2021; Zhou et al., 2017). Previous research estimated that the loss of Disability Adjusted Life-Years (DALYs), according to depression was over 10 million in China in 2013, and the number was expected to increase by about 10% by 2025 (Charlson et al., 2016). Besides, depressive symptoms could also be present in other psychiatric and medical conditions, such as schizophrenia and personality disorders, as well as multiple sclerosis, rheumatic diseases, and a great variety of other conditions (Clarkin et al., 2019; Gascoyne et al., 2019; Harvey et al., 2019; Kwan et al., 2019), and lead to an increased risk of psychiatric disorders and cardiovascular disease (Rajan et al., 2020). It is crucial to figure out the nature of depressive symptoms, reveal the possible risk factors, and how to minimize it in the preclinical stage if we are to reduce this global burden.

Previous studies showed that individuals with elevated levels of depressive symptoms (Hill et al., 2014) and subthreshold depression (Balázs et al., 2013) were at high risk of concurrent and later psychopathology. In case, depression should be regarded as a continuum. Research (Musliner et al., 2016; Whalen et al., 2016) found that individuals with consistently high levels of depressive symptoms over time had similar characteristics, including female gender, low income, and low education. The longitudinal analysis also found that the long-term trajectories of depressive symptoms are heterogenous across the life course, with a notable minority experiencing persistent symptoms and most participants experiencing few or no symptoms. Longitudinal studies help further to understand the different depressive symptom trajectories over time and find their possible risk factors (Kwong et al., 2019). A study (Kuchibhatla et al., 2012) conducted in the USA identified four specific depressive symptoms trajectory groups (‘never’, ‘increasing’, ‘decreasing’, and ‘persistently moderate/high’). Previous studies have focused on examining the trajectories of depressive symptoms among older adults or children and adolescents. Previous studies on the trajectories of depressive symptoms in the general population have mainly been conducted in developed countries such as the United States. As a country with a large population and a high burden of depression (Lu et al., 2021), observing changes in and characteristics of depressive symptom trajectories among the general Chinese population helps identify and locate people at high risk of depression in later life.

To understand the risk profile of depressive symptom trajectories, it is essential to identify how different factors are related to various patterns of depressive symptoms. Previous evidence showed that trajectory groups with higher depression burdens were associated with demographic characteristics, including sex (Salk et al., 2017), socioeconomic position (Melchior et al., 2013), health statuses, such as physical health and disabilities (Liu et al., 1997), health behavior (Hu et al., 2019), social support (Bai et al., 2020), and initial depressive symptom severity (Montagnier et al., 2014). Besides demographic characteristics and health status, blood biomarkers, such as inflammatory and lipid markers, were also associated with depression. Previous studies found a relationship between low-grade systemic inflammation and depression (Frank et al., 2019). A study in Korea suggested that the higher level of depressive symptoms trajectory was associated with lipid abnormalities (Elovainio et al., 2010). Lack of study systematically explored the associations between different types of risk factors and depressive symptoms among community individuals, especially among Chinese in the community.

Figuring out whether different risk factors are associated with specific patterns of trajectories of depressive symptoms could offer more opportunities to target interventions for high-risk populations. In this study, we aimed to figure out the particular manifestations of trajectories of depressive symptoms among Chinese adults, explore the risk profiles of trajectory groups with higher depression burdens, and find the longitudinal associations between lipid markers and C-reactive protein (CRP) with depressive symptoms.

Methods

Study sample

This study used data from the China Health and Retirement Longitudinal Study, CHARLS, a longitudinal cohort study that recruited participants residing in mainland China from 2011. The CHARLS baseline survey covered participants in 50 districts or countries, and 450 urban communities or villages to ensure the sample was representative of the mid-aged and older Chinese population (Zhao et al., 2011). The follow-up surveys were conducted in 2013, 2015, and 2018. This study used data from four waves from 2011 to 2018. After excluding participants without Blood data, a total of 10,120 participants completed the 2011 baseline survey (Figure 1). After excluding individuals without CES-D information at baseline, 5,660 participants aged 40 or older who completed four surveys from 2011 to 2018 were included in this analysis, and 5,641 participants included in the final analysis. Every participant provided written informed consent, and the CHARLS received ethical approval from the Peking University Institutional Review Board. This study is a secondary analysis conducted using established data sets.

Figure 1.

Flowchart of study participant selection.

Measures

Depressive symptoms

Depressive symptoms were assessed using the 10-item Center for Epidemiologic Studies Depression Scale (CES-D) short form in four waves from 2011 to 2018. CES-D is used worldwide (Mohebbi et al., 2018). Each negative item (e.g. ‘I felt fearful’) was scored from 0 (most or all the time) to 3 (rarely or none of the time), and each positive item (e.g. ‘I felt hopeful about the future’) was scored from 0 (rarely or none of the time) to 3 (most or all the time), and the total score of CES-D is 30. The 10-item CES-D Scale was demonstrated to have good validity and reliability among Chinese according to the previous psychometric analysis (H. Chen & Mui, 2014). Higher scores on the 10-item CES-D Scale indicate higher levels of depressive symptoms.

Baseline factors

Baseline factors were divided into demographic characteristics, health statuses, and lifestyles. The demographic factors included in this analysis were sex, age, educational background, marital status, Hukou, and income. The participants’ lifestyles included in this analysis were smoking (current), alcohol assumption (current), and physical activity (PA). The participant’s health status was represented by their BMI, functional limitations, self-reported health, and feeling of pain. Further information on these factors can be found in the eAppendix in Supplemental File 1.

Blood data

The high-sensitivity CRP and lipid marker (High-Density Lipoprotein Cholesterol, HDL; Low-Density Lipoprotein Cholesterol, LDL; Triglycerides, TG; and Total Cholesterol, TC) of the included participants were collected and tested following the ‘Blood Collection and Handling’ CDC manual. The details of collection and measurement are on the website of CHARLS: http://charls.pku.edu.cn/en. All test results for these quality control samples were within the target range (within two standard deviations of the mean quality control concentrations).

Statistical analysis

The missing data was imputed through the random forest method using the ‘missForest package’ (Stekhoven & Bühlmann, 2012) for R 3.5.1. Random forest analysis is an imputation method that can be used for mixed continuous and/or categorical data without requiring the specifications of the distributions of the variables (Hong & Lynn, 2020). We have attributed a single complete data set for analyses. Categorical variables are numbers and percentages, and continuous variables are median and interquartile range. χ² test was used to compare categorical characteristics across different trajectories, and Mann–Whitney tests were used for the continuous variables.

We used generalized estimating equations (GEE) to explore the longitudinal association between blood data and depressive symptoms in two waves from 2011 to 2015. GEE allows repeated measurement data from various distribution types to examine longitudinal associations. The GEE was adjusted by sex, age, educational background, income, marital status, medical history of dyslipidemia, emotional problems, and memory-related diseases.

Growth mixture modeling (GMM) was conducted in Mplus version 8 (Muthén and Muthén) to identify the latent trajectories of depressive symptoms using CES-D scores. GMM divides participants with different patterns into multiple heterogeneous trajectories. We used multinomial logistic regressions to test odds ratios (ORs) and the corresponding 95% Confidence intervals (CIs). Besides, we also applied several adjusted GMMs to test the different trajectory groups by adjusting various kinds of baseline factors (demographic factors, health status, and lifestyle) for sensitivity analysis. Further information on model fit and the steps we applied for the models can be found in the eAppendix in Supplemental File 1. Statistical analysis was performed using SPSS 22.0 and R 3.5.1 (www.r-project.org) with a p-value of <.05, which was considered statistically significant.

Results

A total of 5,641 participants aged 40 or older were included in this analysis. Individuals included in the final analysis were more likely to be female, married, to have agricultural hukou, and to have low educational levels. The median income for included participants was 10,950 yuan/year. Results from our GMM indicated that a four-class trajectory solution was best for the data (Supplemental Table 1).

Trajectories of depressive symptoms

Among the sample of 5,641 individuals, four heterogeneous trajectories of depressive symptoms were derived (Figure 2), and the trajectories of raw GMM are presented in Supplemental Figure 1. The results of sensitivity analyses are in Supplemental Figure 2. First, 3,462 individuals (61.4%) with consistently low levels of depressive symptoms through four waves were divided into the stable-low trajectory. Second, the high-decrease trajectory (1,108 individuals [19.6%]) included participants with high depressive symptoms at wave one and decreased through the follow-ups. Third, the stable-high trajectory (345 individuals [6.1%]) included individuals who had consistently high levels of depressive symptoms, and fourth, the low-increase trajectory (726 individuals [12.9%]) included individuals who started with low levels of depressive symptoms that increased during the follow-ups. The stable -high, high-decrease, and low increase trajectories were regarded as trajectories with higher depression burdens in this study since participants included in these trajectories had depressive symptoms at one or some of the four waves, compared to participants in the stable-low trajectory.

Figure 2.

Trajectories of depressive symptoms from a four-class solution. The model was adjusted for age, gender, marital status, educational level, hukou status, smoking and drinking status, income level, physical activity, BMI, ADL scores, IADL scores, often experiencing pain, and self-reported health status. Among the sample of 5,641 individuals, 3,462 individuals (61.4%) belonged to the stable-low trajectory, 1,108 individuals (19.6%) belonged to the high-decrease trajectory, 345 individuals (6.1%) belonged to the stable-high trajectory, 726 individuals (12.9%) belonged to the low-increase trajectory. CES-D indicates Center for Epidemiologic Studies Depression.

The class distributions without adjustment of covariates are shown in Supplemental Table 2. Compared to participants in the stable-low trajectory group, participants in other trajectory-groups were more likely to be female, to have low educational levels, to have agriculture hukou, to have self-reported good health status, and to feel pain. Participants in the stable-high trajectory group were more likely to be unmarried than those in the stable-low trajectory group. The class distributions with adjustment of different factors are shown in Supplemental Tables 3 to 5.

Association of baseline factors with trajectories of depressive symptoms

The results from the multivariate analysis of baseline factors with various trajectories of adjusted GMM are shown in Table 1, and the results of raw GMM are in Supplemental Table 6. Results of sensitivity analyses are presented in Supplemental Tables 7 to 9, which showed the related baseline factors of various trajectories of depressive symptoms of adjusted GMMs. The results between unadjusted and adjusted multivariate analyses are different. The OR in the related tables represents each trajectory compared with the stable-low trajectory.

Table 1.

Multivariate associations of all factors with adjusted trajectories of depressive symptoms.

	Multinomial OR [95% CI]
Risk factors	High-decrease versus Stable-low	Low-increase versus Stable-low	Stable-high versus Stable-low
Demographic characteristics
Gender
Male	0.486 [0.398, 0.593] * a	0.373 [0.300, 0.463] * a	0.321 [0.233, 0.441] * a
Female (Reference)	–	–	–
Education background
Middle school or below	2.061 [1.375, 3.089] * a	1.637 [1.110, 2.415] * a	3.661 [1.686, 7.950] * a
High school or above (Reference)	–	–	–
Age groups (years)
40–59	1.042 [0.516, 2.104]	2.684 [0.905, 7.962]	2.167 [0.638, 7.365]
60–79	0.960 [0.480, 1.923]	2.147 [0.727, 6.339]	1.692 [0.503, 5.684]
⩾80 [Reference]	–	–	–
Hukou
Agriculture	2.860 [2.089, 3.916] * a	2.879 [2.052, 4.041] * a	1.693 [1.074, 2.670] * a
Non-agriculture (Reference]	–	–	–
Marital Status
Married	0.308 [0.237, 0.400] * a	0.668 [0.492, 0.906] * a	0.310 [0.209, 0.458] ^a
Unmarried (Reference]	–	–	–
Income (yuan/year]
0–3,600	1.639 [1.266, 2.121] * a	1.702 [1.314, 2.203] * a	1.920 [1.293, 2.851] * a
3,601–10,950	1.840 [1.420, 2.386] * a	1.621 [1.243, 2.114] * a	1.571 [1.040, 2.374]*
10,951–21,900	1.611 [1.250, 2.076] * a	1.220 [0.936, 1.590]	1.698 [1.142, 2.525] * a
>21,900 (Reference]	–	–	–
Lifestyle
Smoke
Yes	0.919 [0.738, 1.144]	1.472 [1.138, 1.904] * a	0.970 [0.694, 1.356]
No (Reference]	–	–	–
Alcohol assumption
Yes	0.942 [0.764, 1.162]	0.794 [0.631, 0.998]*	1.049 [0.757, 1.455]
No (Reference)	–	–	–
PA (MET-min/week)
0	0.863 [0.585, 1.273]	1.080 [0.742, 1.570]	0.718 [0.392, 1.313]
0–599	0.760 [0.479, 1.206]	0. 639 [0.385, 1.061]	0.676 [0.335, 1.361]
600–2249	0.656 [0.499, 0.863] * a	0.783 [0.592, 1.036]	0.630 [0.414, 0.959]
⩾2250 (Reference)	–	–	–
Health status
BMI (kg/m²]	0.921 [0.898, 0.944] * a	0.945 [0.921, 0.968] * a	0.899 [0.866, 0.934] * a
ADL (scores)	1.347 [1.235, 1.468] * a	1.278 [1.160, 1.480] * a	1.359 [1.226, 1.506] * a
IADL (scores)	1.436 [1.348, 1.531] * a	1.132 [1.046, 1.225] * a	1.530 [1.418, 1.651] * a
Self-reported health status
Good	0.296 [0.247, 0.354] * a	0.475 [0.398, 0.567] * a	0.164 [0.117, 0.231] * a
Not good (Reference)	–	–	–
Feeling pain
Yes	7.537 [6.334, 8.969] * a	2.753 [2.293, 3.305] * a	12.727 [9.441, 17.157] * a
No (Reference)	–	–	–
Blood marker
Wave 1 (2011]
CRP (mg/L)	0.990 [0.976, 1.004]	1.003 [0.992, 1.015]	0.981 [0.956, 1.006]
TG (mg/dL)	0.999 [0.998, 1.001]	1.000 [0.999, 1.002]	1.001 [0.999, 1.003]
HDL (mg/dL)	1.002 [0.993, 1.012]	1.001 [0.992, 1.011]	1.007 [0.994, 1.019]
LDL (mg/dL)	1.001 [0.993, 1.009]	1.002 [0.994, 1.009]	1.000 [0.990, 1.010]
TC (mg/dL)	1.000 [0.992, 1.008]	1.000 [0.992, 1.007]	1.000 [0.990, 1.010]

Note. ADL = Activities of Daily Living; BMI = Body Mass Index; CRP = C-Reactive Protein; HDL = High Density Lipoprotein Cholesterol; IADL = Instrumental Activities of Daily; LDL = Low Density Lipoprotein Cholesterol; OR = Odds Ratio; PA = Physical Activity; TC = Total Cholesterol; TG = Triglycerides.

Bold values indicates significant results according to p FDR.

p < .05; ^ap FDR < .05.

Being male was negatively associated with trajectories characterized by high depressive symptoms (high-decrease: OR, 0.486; 95% CI [0.398, 0.593]; low-increase: OR, 0.373; 95% CI [0.300, 0.463]; stable-high: OR, 0.321; 95% CI [0.233, 0.441]). Having low educational levels was associated with the low-increase trajectory (OR, 1.637; 95% CI [1.110, 2.415]) and the stable-high trajectory (OR, 3.661; 95% CI [1.686, 7.950]), and the high-decrease trajectory (OR, 2.061; 95% CI [1.375, 3.089]). Having agricultural hukou was associated with the low-increase trajectory (OR, 2.879; 95% CI [2.052, 4.041]), the high-decrease trajectory (OR, 2.860; 95% CI [2.089, 3.916]), and the stable-high trajectory (OR, 1.693; 95% CI [1.074, 2.670]). Feeling pains was also related to all trajectories characterized by high depressive symptoms (high-decrease: OR, 7.537; 95% CI [6.334, 8.969]; low-increase: OR, 2.753; 95% CI [2.293, 3.305]; stable-high: OR, 12.727; 95% CI [9.441, 17.157]).

Longitudinal association between lipid markers and CRP with depressive symptoms

The results of GEE are presented in Table 2, showing the longitudinal association between lipid markers and CRP with depressive symptoms. A significant association was found between HDL and depressive symptoms (OR, 1.028; 95% CI [1.010, 1.06]). However, CRP, LDL, TC, and TG did not have statistically significant associations with depressive symptoms.

Table 2.

Longitudinal association of lipid markers and CRP with depressive symptoms.^a

Risk factors	Depression scores OR (95% CI)	p
CRP (mg/L)	1.006 (0.983, 1.029)	.436
TG (mg/dL)	1.002 (0.999, 1.006)	.354
HDL (mg/dL)	1.028 (1.010, 1.046)	.001
LDL (mg/dL)	1.004 (0.989, 1.018)	.758
TC (mg/dL)	0.992 (0.978, 1.007)	.368

Note. CRP = C-Reactive Protein; HDL = High Density Lipoprotein Cholesterol; LDL = Low Density Lipoprotein Cholesterol; OR = Odds Ratio; TC = Total Cholesterol; TG = Triglycerides.

Bold values indicates significant results according to p FDR.

Adjusted for age, education background, gender, income, marital status, medical history of dyslipidemia, emotional problems, and memory-related diseases.

Discussion

This study identified four trajectories of depressive symptoms among Chinese in middle and late life. Almost half of the included participants were divided into groups with higher depression burdens (the stable-high, low-increase, and high-decrease trajectories). These trajectories were associated with demographic factors, lifestyle, and health status, but not the lipid markers and CRP at baseline. A longitudinal association was found between HDL and depressive symptoms from 2011 to 2015. These findings suggest that examining risk factors could help identify groups with severe and chronic depressive symptoms (stable-high and low-increase) who should be prioritized for early intervention.

There were differences in the demographic characteristics of included participants within each trajectory. Results showed that compared to participants who reported lower depressive symptom trajectories, individuals in other trajectories groups were more likely to be unmarried and female, and have low income and educational levels. Gender differences have consistently been related to trajectories of depressive symptoms. Our results found a strong association between being female with the high decrease, the low increase, and the stable-high trajectories, supporting the conclusion of a previous meta-analysis (Salk et al., 2017). It is generally reported that depression symptom levels tend to be higher and chronic in women. The possible underlying mechanism could be psychological, hormonal, neurochemical, genetic, personality factors, over-reporting of symptoms, and higher stress reactivity (Grigoriadis & Robinson, 2007). Developmental psychopathology theories present various pathways to gender differences in depression in a developmental context (Cicchetti & Rogosch, 2002). For example, the gender difference in neurodevelopmental changes and the confluence of hormonal during the pubertal transition might affect the gender difference in depression (Hyde et al., 2008).

Besides, our results found that low educational levels at baseline had detrimental effects on adults’ depressive symptoms, along with previous studies (Cho et al., 1998; W.-P. Wang, 2008). Some previous research found that individuals with high education background had a lower prevalence of depression, and education’s alleviating effect on depression increased with age. Specifically, a study by W.-P. Wang (2008) in Taiwan, China, found that individuals with more than 16 years of education had 40% lower depression levels than others. Cho et al. (1998) in Korea found that those with less than 13 years of education showed higher levels of depression. Yu et al. (2012) found that those with higher levels of education had relatively lower levels of depressive symptoms among Chinese older adults. Educational background may reduce the incidence of depression through two main mechanisms. One is the allocation function, in which education can increase the ability of individuals to resist depression by increasing the economic and social resources at their disposal. Specifically, previous studies have demonstrated that economic hardship, marital frustration, and unemployment significantly increase levels of depressive symptoms (Levecque et al., 2011). At the same time, those with higher education tend to be more likely to find good jobs, achieve higher income and social status, increase personal self-satisfaction, and be less likely to experience problems such as unemployment and divorce (Jalovaara, 2002; Kettunen, 1997). Thus, they will have relatively lower levels of depression. The second is the socialization function, that is, education can enhance individuals’ ability to resist depression. Firstly, the knowledge, skills, and concepts that individuals acquire through higher education help improve their control, cognitive function, and problem-solving abilities to better deal with work and life difficulties. Secondly, receiving a higher level of education helps individuals to develop healthy and good habits. Finally, people with higher education are more likely to create a well-developed social network, which could reduce the probability of depression (Ross, 2003).

Agriculture hukou was also a risk factor for trajectories with higher depression burdens. The previous research (Liang, 2020) also found that older adults in rural areas reported more prominent mental loneliness, low life satisfaction, and unhappiness. Individuals in rural areas could be defined as a group with substantial public health burdens. Besides, China’s urbanization movement has been intensified in recent years, and the economic gap between urban and rural areas has increased (S. Wang, 2014). An increased number of offspring of the rural population left the countryside to work in cities, resulting in less social support for the rural population, which also influenced the life satisfaction of rural individuals (Chen, 2018). Being female gender, having a low educational background or low-income level, or having agriculture hukou were risk factors for depressive symptoms in middle and late life. The public and health authorities should pay more attention to these specific groups in the early screening and intervention of depression.

Pains were all related to increased risk of trajectories with higher depression burdens. Previous studies also suggested that individuals who often suffer from pain are prone to depression. A 12-year longitudinal study from Amsterdam concluded that participants with elevated pain at baseline are at higher risk of developing depressive symptoms (Hilderink et al., 2012). Clinical research found that chronic pain is a stress state and could induce severe depression (Bair et al., 2003; von Knorring et al., 1983), with up to 85% of patients with chronic pain being affected by severe depression. Besides, patients with chronic pain are more likely to develop muscle soreness, which may lead to restlessness and irritability and subsequently to anxiety or depression (de la Vega et al., 2018; Pope et al., 2015). Pains also have a negative association with sleep duration, increasing the risk of depression (Amtmann et al., 2015).

The participants included in this analysis could be divided into four groups according to their characteristics of depressive symptoms trajectories. The stable-low trajectory included participants with consistently low levels of depressive symptoms, regarded as low depressive symptoms burdens; in this case, we compare the characteristics of other trajectories with higher depression burdens to make better intervention suggestions for the health authorities. Compared to individuals with the stable-low trajectory, participants included in trajectories with higher depression burdens tended to reach a higher level of depressive symptoms during the follow-ups. They had lower levels of educational background, had higher proportions of females, had higher proportions of agriculture hukou, had lower proportions of self-reported good health status, and had higher proportions of feeling pains. Early identification of people with higher depressive symptoms burdens is essential, so the health authorities should focus on the group who tend to be females, with low education background and agriculture hukou to determine the high-risk population through mental health screenings. Compared to individuals with the stable-low trajectory, participants included in the high-decrease trajectory started with high levels of depressive symptoms at baseline and decreased through the follow-ups, and participants in stable-high trajectory had consistently high levels of depressive symptoms through the follow-ups. It is important to find these two groups at an early stage. Early identification of people in these groups will enable them to receive early health promotion and health interventions to decrease depressive symptom levels and further accelerate the decline in depressive symptom levels, which could safeguard their long-term mental health. Besides, compared to participants in the stable-low trajectory, individuals included in the low-increase trajectory began with low levels of depressive symptoms and increased during the follow-ups. Early detection of this group enables health coaching and health promotion initiatives to address their risk lifestyles, such as smoking cessation campaigns, to reduce depressive symptom levels and prevent its exacerbation in the future.

Previous studies did not reveal a similar finding about the relationship between HDL levels and depressive symptoms as shown in our study. A previous review (Morris et al., 2021) summarized that lowered HDL is associated with schizophrenia, obsessive-compulsive disorder (OCD), major depressive disorder (MDD), and postpartum disorder (PPD). Two studies (Peng et al., 2016, 2017) from China presented the association between levels of HDL and the presence of psychopathology. One (Peng et al., 2016) reported that HDL levels are significantly higher in patients with MDD, and the other (Peng et al., 2017) reported that higher levels of HDL were associated with depression, and anxiety. Since the effects of HDL on depressive symptoms were small in this study, this study could only suggest that higher HDL levels could be related to higher depressive symptom trajectories. Further studies with larger samples and longer follow-ups should be made to explore the association between HDL levels and depressive symptoms worldwide.

There are limitations to this study. First, the present results could have selective bias by excluding participants without finishing all the follow-ups, limiting the current results’ representativeness. Second, this analysis did not include cognitive function, mental disorder history, recent stressful events, and early-life stressors as risk factors. Further research should be conducted to explore the relationship between these potential risk factors and depressive symptom trajectories.

This study identified four trajectories of depressive symptoms among Chinese in middle and late life. Results suggest factors including demographic characteristics, health status, and lifestyles are related to depressive symptom trajectories. It may be possible to provide the basis and clues for developing interventions and treatments for adults in middle and late life at most risk of more severe depressive symptoms and further prevent or reduce depression in later life.

Supplemental Material

sj-docx-1-isp-10.1177_00207640231164020 – Supplemental material for Multiple trajectories of depressive symptoms among Chinese in middle and late life: Characterization and risk factors

Supplemental material, sj-docx-1-isp-10.1177_00207640231164020 for Multiple trajectories of depressive symptoms among Chinese in middle and late life: Characterization and risk factors by Chao Li, Jin Liu, Yumeng Ju, Bangshan Liu and Yan Zhang in International Journal of Social Psychiatry

Footnotes

Acknowledgements

The authors thank all the participants in the survey design and data collection and the CHARLS research team for collecting high-quality, nationally representative data and making the data public.

Author contributions

Miss Li conceptualized and designed the study, developed the data extraction instrument, collected data and carried out the initial analysis, and drafted and revised the manuscript. Professor Zhang and Dr. Liu (Bangshan) conceptualized the study, supervised data collection, and critically screened important intellectual contents of the manuscript. Dr. Liu (Jin) and Dr. Ju reviewed the manuscript. All authors have read and approved the manuscript as submitted and agree to be accountable for all aspects of the work.

Conflict of interest

The authors declare that they have no competing interests. Miss. Li wrote the first draft of the manuscript. No honorarium, grant, or other forms of payment was given to anyone to produce the manuscript.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was funded by the STI2030-Major Projects, 2021ZD0202000 to Yan Zhang. The data collection was supported by the Behavioral and Social Research Division of the National Institute on Aging of the National Institute of Health (grants 1-R21-AG031372-01, 1-R01-AG037031-01, and 3-R01AG037031-03S1) the Natural Science Foundation of China (grants 70773002, 70910107022, and 71130002), the World Bank (contracts 7145915 and 7159234), and Peking University. The funders had no role in study design, data collection and analysis, decision to publish, or manuscript preparation.

Statement of ethics

All methods were carried out following the Declaration of Helsinki. Each participant provided written informed consent, and the China Health and Retirement Longitudinal Study received ethical approval from the Peking University Institutional Review Board. The IRB approval number for the main household survey, including anthropometrics, is IRB00001052-11015; the IRB approval number for biomarker collection, was IRB00001052-11014. This study was a second analysis based on the data from the CHARLS study, so the ethics committee that approved the CHARLS study is noted to be Peking University.

Consent for publication

Not applicable.

ORCID iDs

Bangshan Liu

Yan Zhang

Data availability statement

All data generated or analyzed during this study are included in this article.

Supplemental material

Supplemental material for this article is available online.

References

Amtmann

Askew

R. L.

Kim

Chung

Ehde

D. M.

Bombardier

C. H.

Kraft

G. H.

Jones

S. M.

Johnson

K. L.

(2015). Pain affects depression through anxiety, fatigue, and sleep in multiple sclerosis. Rehabilitation Psychology, 60(1), 81–90. https://doi.org/10.1037/rep0000027

Bair

M. J.

Robinson

R. L.

Katon

Kroenke

(2003). Depression and pain comorbidity: A literature review. Archive of Internal Medicine, 163(20), 2433–2445. https://doi.org/10.1001/archinte.163.20.2433

Bai

Bian

Zhang

Cao

(2020). The impact of social support on the health of the rural elderly in China. International Journal of Environmental Research and Public Health,17(6), 2004. https://doi.org/10.3390/ijerph17062004

Balázs

Miklósi

Keresztény

Á.

Hoven

C. W.

Carli

Wasserman

Apter

Bobes

Brunner

Cosman

Cotter

Haring

Iosue

Kaess

Kahn

J. P.

Keeley

Marusic

Postuvan

Resch

, . . . Wasserman

(2013). Adolescent subthreshold-depression and anxiety: Psychopathology, functional impairment and increased suicide risk. Journal of Child Psychology and Psychiatry, 54(6), 670–677. https://doi.org/10.1111/jcpp.12016

Charlson

F. J.

Baxter

A. J.

Cheng

H. G.

Shidhaye

Whiteford

H. A.

(2016). The burden of mental, neurological, and substance use disorders in China and India: A systematic analysis of community representative epidemiological studies. Lancet, 388(10042), 376–389. https://doi.org/10.1016/S0140-6736(16)30590-6

Chen

(2018). Life satisfaction and related factors of rural older adults in the process of urbanization. Chinese Journal of Gerontology, 38(10), 4.

Chen

Mui

A. C.

(2014). Factorial validity of the Center for Epidemiologic Studies Depression Scale short form in older population in China. International Psychogeriatrics, 26(1), 49–57. https://doi.org/10.1017/s1041610213001701

Cho

M. J.

Nam

J. J.

Suh

G. H.

(1998). Prevalence of symptoms of depression in a nationwide sample of Korean adults. Psychiatry Research, 81(3), 341–352. https://doi.org/10.1016/s0165-1781(98)00122-x

Cicchetti

Rogosch

F. A.

(2002). A developmental psychopathology perspective on adolescence. Journal of Consulting and Clinical Psychology, 70(1), 6–20. https://doi.org/10.1037//0022-006x.70.1.6

10.

Clarkin

J. F.

Petrini

Diamond

(2019). Complex depression: The treatment of major depression and severe personality pathology. Journal of Clinical Psychology, 75(5), 824–833. https://doi.org/10.1002/jclp.22759

11.

Cui

(2015). Editorial: A systematic review of depression. Current Neuropharmacology, 13(4), 480. https://doi.org/10.2174/1570159X1304150831123535

12.

de la Vega

Groenewald

Bromberg

M. H.

Beals-Erickson

S. E.

Palermo

T. M.

(2018). Chronic pain prevalence and associated factors in adolescents with and without physical disabilities. Developmental Medicine and Child Neurology, 60(6), 596–601. https://doi.org/10.1111/dmcn.13705

13.

Elovainio

Pulkki-Råback

Kivimäki

Jokela

Viikari

Raitakari

O. T.

Telama

Keltikangas-Järvinen

(2010). Lipid trajectories as predictors of depressive symptoms: The Young Finns Study. Health Psychology, 29(3), 237–245. https://doi.org/10.1037/a0018875

14.

Frank

Kaushal

Poole

Lawes

Chalder

Cadar

(2019). Systemic low-grade inflammation and subsequent depressive symptoms: Is there a mediating role of physical activity? Brain, Behavior, and Immunity, 80, 688–696. https://doi.org/10.1016/j.bbi.2019.05.017

15.

Gascoyne

C. R.

Simpson

Jr Chen

van der Mei

Marck

C. H.

(2019). Modifiable factors associated with depression and anxiety in multiple sclerosis. Acta Neurologica Scandinavica, 140(3), 204–211. https://doi.org/10.1111/ane.13132

16.

Grigoriadis

Robinson

G. E.

(2007). Gender issues in depression. Annals of Clinical Psychiatry, 19(4), 247–255. https://doi.org/10.1080/10401230701653294

17.

Xie

Long

Chen

Pan

Yan

Liang

Tan

Xie

Wei

(2013). Epidemiology of major depressive disorder in mainland china: A systematic review. PLoS One, 8(6), e65356. https://doi.org/10.1371/journal.pone.0065356

18.

Harvey

P. D.

Deckler

Jones

M. T.

Jarskog

L. F.

Penn

D. L.

Pinkham

A. E.

(2019). Autism symptoms, depression, and active social avoidance in schizophrenia: Association with self-reports and informant assessments of everyday functioning. Journal of Psychiatric Research, 115, 36–42. https://doi.org/10.1016/j.jpsychires.2019.05.010

19.

Hilderink

P. H.

Burger

Deeg

D. J.

Beekman

A. T.

Oude Voshaar

R. C.

(2012). The temporal relation between pain and depression: Results from the longitudinal aging study Amsterdam. Psychosomatic Medicine, 74(9), 945–951. https://doi.org/10.1097/PSY.0b013e3182733fdd

20.

Hill

R. M.

Pettit

J. W.

Lewinsohn

P. M.

Seeley

J. R.

Klein

D. N.

(2014). Escalation to major depressive disorder among adolescents with subthreshold depressive symptoms: Evidence of distinct subgroups at risk. Journal of Affective Disorders, 158, 133–138. https://doi.org/10.1016/j.jad.2014.02.011

21.

Hong

Lynn

H. S.

(2020). Accuracy of random-forest-based imputation of missing data in the presence of non-normality, non-linearity, and interaction. BMC Medical Research Methodology, 20(1), 199. https://doi.org/10.1186/s12874-020-01080-1

22.

Martikainen

(2019). Rural-urban disparities in age trajectories of depression caseness in later life: The China Health and Retirement Longitudinal Study. PLoS One, 14(4), e0215907. https://doi.org/10.1371/journal.pone.0215907

23.

Hyde

J. S.

Mezulis

A. H.

Abramson

L. Y.

(2008). The ABCs of depression: Integrating affective, biological, and cognitive models to explain the emergence of the gender difference in depression. Psychological Review, 115(2), 291–313. https://doi.org/10.1037/0033-295X.115.2.291

24.

Jalovaara

(2002). Socioeconomic differentials in divorce risk by duration of marriage. Demographic Research, 7(16), 537–564.

25.

Kettunen

(1997). Education and unemployment duration. Economics of Education Review, 16(2), 163–170. https://doi.org/10.1016/S0272-7757(96)00057-X

26.

Kuchibhatla

M. N.

Fillenbaum

G. G.

Hybels

C. F.

Blazer

D. G.

(2012). Trajectory classes of depressive symptoms in a community sample of older adults. Acta Psychiatrica Scandinavica, 125(6), 492–501. https://doi.org/10.1111/j.1600-0447.2011.01801.x

27.

Kwan

Marzouk

Ghanean

Kishwar

Anderson

Bonilla

Vitti

Touma

(2019). Assessment of the psychometric properties of patient-reported outcomes of depression and anxiety in systemic lupus erythematosus. Seminars in Arthritis and Rheumatism, 49(2), 260–266. https://doi.org/10.1016/j.semarthrit.2019.03.004

28.

Kwong

A. S. F.

López-López

J. A.

Hammerton

Manley

Timpson

N. J.

Leckie

Pearson

R. M.

(2019). Genetic and environmental risk factors associated with trajectories of depression symptoms from adolescence to young adulthood. JAMA Network Open, 2(6), e196587. https://doi.org/10.1001/jamanetworkopen.2019.6587

29.

Levecque

Van Rossem

De Boyser

Van de Velde

Bracke

(2011). Economic hardship and depression across the life course: The impact of welfare state regimes. Journal of Health and Social Behavior, 52(2), 262–276. https://doi.org/10.1177/0022146510394861

30.

Liang

(2020). Heterogeneity in the trajectories of depressive symptoms among elderly adults in rural China: The role of housing characteristics. Health & Place, 66, 102449. https://doi.org/10.1016/j.healthplace.2020.102449

31.

Liu

C. Y.

Wang

S. J.

Teng

E. L.

Fuh

J. L.

Lin

C. C.

Lin

K. N.

Chen

H. M.

Lin

C. H.

Wang

P. N.

Yang

Y. Y.

Larson

E. B.

Chou

Liu

H. C.

(1997). Depressive disorders among older residents in a Chinese rural community. Psychological Medicine, 27(4), 943–949. https://doi.org/10.1017/s0033291797005230

32.

Huang

Yin

Wang

Huang

Yan

Wang

Xiao

Zhou

Zhang

Chen

Zhang

, . . . Zhang

(2021). Prevalence of depressive disorders and treatment in China: A cross-sectional epidemiological study. The Lancet Psychiatry, 8(11), 981–990. https://doi.org/10.1016/s2215-0366(21)00251-0

33.

Melchior

Chastang

J. F.

Head

Goldberg

Zins

Nabi

Younès

(2013). Socioeconomic position predicts long-term depression trajectory: A 13-year follow-up of the GAZEL cohort study. Molecular Psychiatry, 18(1), 112–121. https://doi.org/10.1038/mp.2011.116

34.

Mohebbi

Nguyen

McNeil

J. J.

Woods

R. L.

Nelson

M. R.

Shah

R. C.

Storey

Murray

A. M.

Reid

C. M.

Kirpach

Wolfe

Lockery

J. E.

Berk

(2018). Psychometric properties of a short form of the Center for Epidemiologic Studies Depression (CES-D-10) scale for screening depressive symptoms in healthy community dwelling older adults. General Hospital Psychiatry, 51, 118–125. https://doi.org/10.1016/j.genhosppsych.2017.08.002

35.

Montagnier

Dartigues

J. F.

Rouillon

Pérès

Falissard

Onen

(2014). Ageing and trajectories of depressive symptoms in community-dwelling men and women. International Journal of Geriatric Psychiatry, 29(7), 720–729. https://doi.org/10.1002/gps.4054

36.

Morris

Puri

B. K.

Bortolasci

C. C.

Carvalho

Berk

Walder

Moreira

E. G.

Maes

(2021). The role of high-density lipoprotein cholesterol, apolipoprotein A and paraoxonase-1 in the pathophysiology of neuroprogressive disorders. Neuroscience and Biobehavioral Reviews, 125, 244–263. https://doi.org/10.1016/j.neubiorev.2021.02.037

37.

Musliner

K. L.

Munk-Olsen

Eaton

W. W.

Zandi

P. P.

(2016). Heterogeneity in long-term trajectories of depressive symptoms: Patterns, predictors and outcomes. Journal of Affective Disorders, 192, 199–211. https://doi.org/10.1016/j.jad.2015.12.030

38.

Peng

Y. F.

Xiang

Wei

Y. S.

(2016). The significance of routine biochemical markers in patients with major depressive disorder. Scientific Reports, 6, 34402. https://doi.org/10.1038/srep34402

39.

Peng

Y. F.

Zhong

S. M.

Yu-Hua

Hui-Dan

Wu-Jun

(2017). The relationship between major depressive disorders and glucose parameters: A cross-sectional study in Chinese population. Advances in Clinical and Experimental Medicine, 26(4), 665–659. https://doi.org/10.17219/acem/63023

40.

Pope

C. J.

Sharma

Mazmanian

(2015). A systematic review of the association between Psychiatric Disturbances and endometriosis. Journal of Obstetrics and Gynaecology Canada, 37(11), 1006–1015. https://doi.org/10.1016/s1701-2163(16)30050-0

41.

Rajan

McKee

Rangarajan

Bangdiwala

Rosengren

Gupta

Kutty

V. R.

Wielgosz

Lear

AlHabib

K. F.

H. U.

Lopez-Jaramillo

Avezum

Seron

Oguz

Kruger

I. M.

Diaz

Nafiza

M. N.

Chifamba

, . . . Yusuf

(2020). Association of symptoms of depression with cardiovascular disease and mortality in low-, middle-, and high-income countries. JAMA Psychiatry, 77(10), 1052–1063. https://doi.org/10.1001/jamapsychiatry.2020.1351

42.

Ross

C. E.

(2003). Education, social status, and health Macsource press.

43.

Salk

R. H.

Hyde

J. S.

Abramson

L. Y.

(2017). Gender differences in depression in representative national samples: Meta-analyses of diagnoses and symptoms. Psychological Bulletin, 143(8), 783–822. https://doi.org/10.1037/bul0000102

44.

Stekhoven

D. J.

Bühlmann

(2012). MissForest—non-parametric missing value imputation for mixed-type data. Bioinformatics, 28(1), 112–118. https://doi.org/10.1093/bioinformatics/btr597

45.

von Knorring

Perris

Eisemann

Eriksson

Perris

. (1983). Pain as a symptom in depressive disorders. II. Relationship to personality traits as assessed by means of KSP. Pain, 17(4), 377–384. https://doi.org/10.1016/0304-3959(83)90169-0

46.

Wang

(2014). Study on the impact of urbanization on urban-rural income gap. Yunnan University of Finance and Economics.

47.

Wang

W.-P.

(2008). Education and depression in Taiwan: Aging trajectories cohort variations, mechanisms of divergence, and Resource Substitution. The University of Texas at Austin.

48.

Whalen

D. J.

Luby

J. L.

Tilman

Mike

Barch

Belden

A. C.

(2016). Latent class profiles of depressive symptoms from early to middle childhood: Predictors, outcomes, and gender effects. Journal of child psychology and psychiatry, and allied disciplines, 57(7), 794–804. https://doi.org/10.1111/jcpp.12518

49.

Cuijpers

(2012). Prevalence and correlates of depressive symptoms in Chinese older adults: A population-based study. International Journal of Geriatric Psychiatry, 27(3), 305–312. https://doi.org/10.1002/gps.2721

50.

Zhao

Peifeng (Perry) Hu

E. C.

Jung

K. I.

Strauss

J. M.

Yang

Yin

Wang

(2011). China health and retirement longitudinal study 2011-2012 National Baseline Blood Data Users’ Guide. Peking University.

51.

Zhou

Wang

Zhu

Zhang

Ding

Tang

(2017). Prevalence and associated factors of anxiety and depression among patients with chronic respiratory diseases in eight general hospitals in Jiangsu Province of China: A cross-sectional study. Psychiatry Research, 251, 48–53. https://doi.org/10.1016/j.psychres.2017.01.070

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.35 MB