Prevalence and associated factors of depressive symptoms in Chinese diabetic patients: A study based on Andersen’s behavioral model

Abstract

BACKGROUND

Diabetes mellitus (DM) is a rapidly growing global health emergency of the 21^st century. Comorbidities, such as DM and depression, are common, presenting challenges to the healthcare system.

AIM

To investigate the prevalence of depression and its associated factors in patients with DM and to strengthen the management of depression in this patient group.

METHODS

Participants were selected from the 2018 China Health and Retirement Longitudinal Study. Depressive symptoms were assessed using the 10-item Center for Epidemiological Studies Depression Scale, with a score of 10 or more indicating depression. Group differences were compared using analysis of variance and χ² tests. Binary logistic regression was conducted to explore the odds ratios (ORs) of independent variables. Following Andersen’s behavioral model, predisposing, enabling, health need, and health behavior variables were introduced stepwise into the logistic model.

RESULTS

Of the 1673 patients with diabetes, 41.4% had depressive symptoms. Regarding the predisposing characteristics, patients who were male (OR 0.426, P < 0.05), married (OR 0.634, P < 0.05), and received a high school education or higher (OR 0.432, P < 0.05) reported fewer depressive symptoms. Healthcare needs, including better self-rated health (OR 0.458 for fair and OR 0.247 for good, P < 0.05) and more sleep (OR 0.642, P < 0.05), were associated with a lower likelihood of depressive symptoms. In contrast, pain (OR 1.440 for mild and OR 2.644 for severe, P < 0.05) and impairment in the basic activities of daily living (OR 1.886, P < 0.05) were inversely associated. Additionally, patients highly satisfied with healthcare services (OR 0.579, P < 0.05) were less likely to have depressive symptoms.

CONCLUSION

Nearly half of the patients with DM reported depressive symptoms, which were strongly associated with predisposing characteristics and healthcare needs, particularly physical pain and impairment in basic activities of daily living. Our study emphasizes the significance of enhanced screening and intervention for depression in diabetes care along with improved management of functional impairments.

Key Words: China; Comorbidity; Depressive symptoms; Diabetes mellitus; Health correlates

Core Tip: In this cross-sectional analysis, we present the prevalence of and factors associated with depressive symptoms in patients with diabetes mellitus (DM). Nearly half (41%) of the patients with DM reported depressive symptoms, which were strongly associated with predisposing characteristics and healthcare needs, particularly physical pain and impairment in basic activities of daily living. Enhanced screening and interventions for depression in diabetes care, along with improved management of functional impairment, should be implemented.

INTRODUCTION

Diabetes mellitus (DM) is a chronic metabolic disease characterized by elevated blood glucose levels. The International Diabetes Federation states that DM is one of the most rapidly growing global health emergencies of the 21^st century[1]. As of 2021, approximately 537 million people globally had DM, and this number is expected to rise to 643 million by 2030 and 783 million by 2045[1]. China has over 140 million people living with DM, and DM-related healthcare costs more than $160 billion[1]. Depression is a common mental disorder affecting approximately 300 million people worldwide[2]. It is characterized by persistent sadness and a lack of interest or pleasure in previously rewarding or enjoyable activities[3]. Depression is a major cause of disability in China, ranking fourth in years lived with disability in 2019[4]. According to a large epidemiological survey, the 12-month and lifetime prevalence of depression in the Chinese population is 3.6% and 6.8% respectively[5].

In addition to their individual contributions to the global disease burden, comorbidities of chronic conditions such as DM and depression are common[6-8]. Previous studies have reported a bidirectional relationship between DM and depression[9]. Roy and Lloyd[10] found that individuals with type 2 DM (T2DM) were three times more likely to experience depression than those without T2DM. Campayo et al[11] demonstrated that clinically significant depression was associated with a 65% increased risk of developing DM. According to a meta-analysis by Khaledi et al[12], 28% of patients with T2DM had varying degrees of depressive disorders. Geographically, the prevalence of depression in patients with DM was higher in Asia (32%) and Australia (29%) than in Africa (27%) and Europe (24%). Multiple cross-sectional studies in China have shown that the prevalence of depression among patients with DM ranges from 25.6% to 56.1%[13-15].

The co-occurrence of DM and depression presents a dual burden on healthcare systems and is detrimental to socioeconomic development. Depression in people with DM leads to inadequate self-care, impaired self-management, and decreased quality of life[16-19]. It is also associated with increased healthcare utilization and costs[20-24]. Thus, evaluating the associations between DM and depression and identifying potential risk factors are crucial for early intervention. Researchers in China and abroad have identified several risk factors that increase the risk of depression in people with DM, including low socioeconomic status, adverse life events, health behaviors and status, and DM-related conditions and treatments[10,25-27]. In a community-based study in China, Sun et al[14] found that female sex, older age, lower education, single status, DM complications, and poor sleep quality were associated with an increased risk of depressive symptoms in patients with T2DM. Studies on DM and depression have steadily increased in China. Notably, many of these studies were confined to hospitals or limited geographical regions[28,29]. Moreover, studies utilizing national survey datasets have primarily focused on specific factors associated with depressive symptoms in patients with DM, such as rural-urban structure, physical function, and activities of daily living (ADL)[30,31]. However, systematic analyses of the correlates of depression in patients with DM from a theoretical perspective are rare. In this study, we examined the prevalence of and factors associated with depressive symptoms in Chinese patients with DM, using a nationally representative survey dataset. The analysis was framed within Andersen’s Behavioral Model of Health Service Use (Andersen’s behavioral model) to comprehensively understand the psychological dimensions of DM management.

MATERIALS AND METHODS

Data source and sample selection

In this cross-sectional study, we present the prevalence of and factors associated with depressive symptoms in patients with diabetes. Data were collected from the 2018 China Health and Retirement Longitudinal Study (CHARLS). CHARLS is a nationally representative survey organized by the National School of Development of Peking University and was conducted by the Institute of Social Science Survey of Peking University. A multi-stage stratified probability proportional to size sampling method was used in the CHARLS to select Chinese residents aged 45 years and older as respondents and collect their social, economic, and health-related information. In 2011, the CHARLS research team conducted a national baseline survey, interviewing 17708 respondents from 28 provinces, 150 countries/districts, and 450 villages/urban communities nationwide. The second, third, and fourth waves of the CHARLS follow-up survey were conducted in 2013, 2015, and 2018, respectively. In 2018, the fourth wave of the CHARLS increased the number of respondents to 19816. More information on the CHARLS, including its design, implementation, data release, and quality control, can be found on the official website (http://charls.pku.edu.cn).

In the present study, our inclusion criteria was a self-reported diagnosis of diabetes in the 2018 CHARLS. The exclusion criteria were: (1) Participants who did not complete the depression scale; and (2) Participants with missing data on predisposing characteristics, enabling resources, healthcare needs, and health behaviors. Ultimately, 1673 participants were eligible to participate in the survey (Supplementary Figure 1).

Theoretical framework

Andersen’s behavioral model, proposed by Andersen[32] in 1968, is a mainstream theoretical model used to study an individual’s health service utilization and behavior. In this model, factors related to individual health behaviors and outcomes are grouped into three categories: Predisposing characteristics, enabling resources, and need[33]. Predisposing factors are characteristics that make individuals more likely to use healthcare services. These factors include demographic characteristics, social structural factors, and health beliefs. However, enabling factors are the resources and abilities that allow individuals to access healthcare services. These factors include financial status, family resources, and social support. The need for healthcare services is influenced by predisposing and enabling factors, including perceived and evaluated needs. Additionally, a correlation exists between individual health behaviors (including health service utilization and lifestyle) and health outcomes. Andersen’s model presents a theoretical framework for a comprehensive understanding of the factors associated with depressive symptoms in patients with DM.

Dependent variables

The depression status of patients with DM was the dependent variable. It was assessed using the 10-item Center for Epidemiological Studies Depression Scale (CESD-10, Supplementary material). This scale has shown good reliability and validity in middle-aged and older Chinese populations[34]. The CESD-10 comprises 10 questions related to behaviors and feelings. It was designed to determine how frequently the respondent has experienced certain events in the past week. Positive feelings and behaviors that occur less than one day, 1-2 days, 3-4 days, and 5-7 days are scored as 3, 2, 1, and 0 points, respectively. Negative feelings or behaviors follow the opposite pattern. The CESD-10 yields a total score of 30 points. The severity of depressive symptoms is directly proportional to the score. Consistent with previous research[35], individuals scoring 10 or higher were defined as having depressive symptoms.

Independent variables

We examined the potential relationship between depression status in patients with DM and four dimensions: Predisposing characteristics, enabling resources, health needs, and health behaviors using Andersen’s behavioral model. Predisposing characteristics were age, sex, education, and marital status. The enabling resources were the place of residence, personal income in the past year, months of living with children in the past year, and social activities. Healthcare needs were self-rated health, pain, impairment of basic ADL (BADL, assessed using the Basic ADL Scale), disability, and sleep duration. Health behaviors were outpatient visits in the past month, hospitalizations in the past year, satisfaction with healthcare service, DM treatment, blood glucose testing in the past year, and smoking and drinking habits. Figure 1 shows Andersen’s theoretical model of the factors influencing depression in patients with DM.

Figure 1 Andersen’s theoretical model of factors influencing depression status in diabetic patients.

Statistical analysis

Data processing and analysis were performed using Stata 16.0 (Stata Corp., LLC.; College Station, TX, United States). Continuous variables are presented as mean and standard deviation, and categorical variables are presented as the number of respondents and sample percentage. Analysis of variance was used for continuous variables, whereas the χ² test was used for categorical variables to show between-group differences. Binary logistic regression was used to investigate the odds ratios (ORs) of depressive symptoms correlates in patients with DM. The predisposing (Model 1), enabling (Model 2), need (Model 3), and health behavior variables (Model 4) were controlled stepwise to explicitly observe the effect of each factor. The significance level was set at 5%.

RESULTS

Predisposing characteristics and depressive symptoms of patients with DM

Of the 1673 patients with DM, 693 (41.4%) had depressive symptoms. The weighted mean ages of all patients, patients with DM and depressive symptoms, and patients with DM but without depressive symptoms were 63.80, 63.82, and 63.78 years, respectively. More than half of the patients were female (52.1%), had primary or secondary education (64.2%), and were married (87.1%). The proportions of females (67.1% vs 42.3%, P < 0.05), those with no formal education (22.2% vs 11.2%, P < 0.05), and single/divorced/widowed individuals (17.9% vs 9.7%, P < 0.05) were significantly higher in those with depressive symptoms than in patients without depressive symptoms (Table 1).

Table 1 Predisposing characteristics and depressive symptoms of patients with diabetes mellitus.

Variables	Total			Depressed			Not depressed			P value
	n	Unweighted	Weighted	n	Unweighted	Weighted	n	Unweighted	Weighted
Total	1673	100.0%	100.0%	693	100.0%	100.0%	980	100.0%	100.0%	-
Age (mean ± SD)	1673	63.14 ± 8.99	63.80 ± 9.28	693	63.25 ± 9.07	63.82 ± 9.30	980	63.06 ± 8.93	63.78 ± 9.27	0.936
Sex										< 0.001
Female	914	54.6%	52.1%	458	66.1%	67.1%	456	46.5%	42.3%
Male	759	45.4%	47.9%	235	33.9%	32.9%	524	53.5%	57.7%
Education										< 0.001
Not formally educated	310	18.5%	15.5%	173	25.0%	22.2%	137	14.0%	11.2%
Primary or middle school	1083	64.7%	64.2%	449	64.8%	67.9%	634	64.7%	61.8%
High school and above	280	16.7%	20.3%	71	10.2%	9.9%	209	21.3%	27.0%
Marital status										< 0.001
Single/divorced/widowed	231	13.8%	12.9%	131	18.9%	17.9%	100	10.2%	9.7%
Married/cohabitation	1442	86.2%	87.1%	562	81.1%	82.1%	880	89.8%	90.3%

Enabling resources and depressive symptoms of patients with DM

Regarding place of residence, 49.3% of the patients lived in rural areas, and 37.1% lived in urban areas. The weighted median income was ¥5471, and the weighted median number of months of living with children was 3.28 months for all patients. In the past month, 61.6% of the patients had participated in at least one social activity. Individuals with depressive symptoms had a higher proportion of living in rural areas (55.0% vs 45.5%, P < 0.05), a higher proportion of not participating in any social activities (43.8% vs 34.9%, P < 0.05), and a lower personal income (¥1839 vs ¥7835, P < 0.05) than in patients without depressive symptoms. The number of months of living with children was 3.34 and 3.24 months for DM patients with and without depressive symptoms, respectively, and the difference was not statistically significant (Table 2).

Table 2 Enabling resources and depressive symptoms of patients with diabetes mellitus.

Variables	Total			Depressed			Not depressed			P value
	n	Unweighted	Weighted	n	Unweighted	Weighted	n	Unweighted	Weighted
Total	1673	100.0%	100.0%	693	100.0%	100.0%	980	100.0%	100.0%	-
Residence										< 0.001
Rural areas	999	59.7%	49.3%	460	66.4%	55.0%	539	55.0%	45.5%
Urban areas	479	28.6%	37.1%	150	21.6%	29.7%	329	33.6%	41.9%
Combination zone	195	11.7%	13.7%	83	12.0%	15.3%	112	11.4%	12.6%
Personal income (mean ± SD)	1673	4881 ± 31882	5471 ± 38410	693	1957 ± 9270	1839 ± 9137	980	6949 ± 40803	7835 ± 48663	0.003
Months of living with children (mean ± SD)	1673	3.48 ± 6.14	3.28 ± 5.75	693	3.62 ± 6.32	3.34 ± 5.87	980	3.37 ± 6.02	3.24 ± 5.68	0.735
Social activity										0.003
No	675	40.3%	38.4%	317	45.7%	43.8%	358	36.5%	34.9%
Yes	998	59.7%	61.6%	376	54.3%	56.2%	622	63.5%	65.1%

Healthcare needs and depressive symptoms of patients with DM

In terms of self-rated health status, 41.0% of patients considered their health to be poor, and only 10.9% considered their health to be good. Body pain was common among patients with DM, with 43.2% having mild pain and 21.3% having severe pain. In addition, 22.6% and 41.4% of patients had BADL impairment and disability, respectively. The weighted mean sleep duration was 6.10 hours. Patients with depressive symptoms were more likely to have a poor self-rated health status (60.8% vs 28.2%, P < 0.05), severe pain (36.3% vs 11.6%, P < 0.05), BADL impairment (36.1% vs 13.8%, P < 0.05), disability (51.5% vs 34.7%, P < 0.05), and fewer hours of sleep (5.65 vs 6.39%, P < 0.05) than in those without depressive symptoms (Table 3).

Table 3 Healthcare needs and depressive symptoms of patients with diabetes mellitus.

Variables	Total			Depressed			Not depressed			P value
	N	Unweighted	Weighted	n	Unweighted	Weighted	n	Unweighted	Weighted
Total	1673	100.0%	100.0%	693	100.0%	100.0%	980	100.0%	100.0%	-
Self-rated health										< 0.001
Poor	733	43.8%	41.0%	445	64.2%	60.8%	288	29.4%	28.2%
Fair	753	45.0%	48.1%	219	31.6%	34.8%	534	54.5%	56.7%
Good	187	11.2%	10.9%	29	4.2%	4.4%	158	16.1%	15.1%
Pain										< 0.001
None	540	32.3%	35.5%	130	18.8%	20.2%	410	41.8%	45.4%
Mild	721	43.1%	43.2%	292	42.1%	43.5%	429	43.8%	43.0%
Severe	412	24.6%	21.3%	271	39.1%	36.3%	141	14.4%	11.6%
BADL impairment										< 0.001
No	1250	74.7%	77.4%	421	60.8%	63.9%	829	84.6%	86.2%
Yes	423	25.3%	22.6%	272	39.2%	36.1%	151	15.4%	13.8%
Disability										< 0.001
No	940	56.2%	58.6%	312	45.0%	48.5%	628	64.1%	65.3%
Yes	733	43.8%	41.4%	381	55.0%	51.5%	352	35.9%	34.7%
Sleep duration (mean ± SD)	1673	6.06 ± 1.99	6.10 ± 1.90	693	5.59 ± 2.16	5.65 ± 2.07	980	6.39 ± 1.78	6.39 ± 1.72	< 0.001

BADL: Basic activities of daily living.

Health behaviors and depressive symptoms of patients with DM

The weighted median number of outpatient visits and hospitalizations was 0.58 and 0.5, respectively, and nearly one-fifth (19.8%) of the patients were dissatisfied with healthcare services. The majority (82.0%) of patients were receiving treatment for DM, and the median number of blood glucose tests was 25.2. Regarding smoking and drinking, 58.4% and 68.3% of patients did not smoke or drink alcohol, respectively. Patients with depressive symptoms had more outpatient visits (0.77 vs 0.46, P < 0.05) and hospitalizations (0.65 vs 0.41, P < 0.05), whereas the proportion of satisfaction with healthcare services (29.0% vs 38.9%, P < 0.05), smoking (33.1% vs 47.2%, P < 0.05), and drinking (24.8% vs 36.2%, P < 0.05) was lower than in patients without depressive symptoms. The proportion of DM treatment (82.5% vs 81.7%, P > 0.05) and blood glucose testing times (24.9 vs 25.4, P < 0.05) were not significantly different between patients with and without DM (Table 4).

Table 4 Health behaviors and depressive symptoms of patients with diabetes mellitus.

Variables	Total			Depressed			Not depressed			P value
	n	Unweighted	Weighted	n	Unweighted	Weighted	n	Unweighted	Weighted
Total	1673	100.0%	100.0%	693	100.0%	100.0%	980	100.0%	100.0%	-
Outpatient visit (mean ± SD)	1673	0.57 ± 1.63	0.58 ± 1.54	693	0.77 ± 2.08	0.77 ± 1.96	980	0.44 ± 1.19	0.46 ± 1.18	< 0.001
Hospitalization (mean ± SD)	1673	0.50 ± 1.11	0.50 ± 1.06	693	0.66 ± 1.39	0.65 ± 1.30	980	0.39 ± 0.83	0.41 ± 0.86	< 0.001
Health service satisfaction										0.002
Dissatisfied	321	19.2%	19.4%	159	22.9%	24.1%	162	16.5%	16.3%
Neutral	744	44.5%	45.6%	308	44.4%	46.9%	436	44.5%	44.8%
Satisfied	608	36.3%	35.0%	226	32.6%	29.0%	382	39.0%	38.9%
DM treatment										0.724
No	341	20.4%	18.0%	142	20.5%	17.5%	199	20.3%	18.3%
Yes	1332	79.6%	82.0%	551	79.5%	82.5%	781	79.7%	81.7%
Blood glucose test (mean ± SD)	1673	20.2 ± 60.0	25.2 ± 69.8	693	22.7 ± 76.4	24.9 ± 79.5	980	18.4 ± 44.9	25.4 ± 62.7	0.885
Smoking										< 0.001
No	999	59.7%	58.4%	457	65.9%	66.9%	542	55.3%	52.8%
Yes	674	40.3%	41.6%	236	34.1%	33.1%	438	44.7%	47.2%
Drinking										< 0.001
No	1186	70.9%	68.3%	534	77.1%	75.2%	652	66.5%	63.8%
Yes	487	29.1%	31.7%	159	22.9%	24.8%	328	33.5%	36.2%

Variables associated with depressive symptoms in patients with DM using logistic regression analysis

Binary logistic regression analysis of depression status is shown in Table 5. Model 1 was controlled for predisposing variables and showed that age, sex, and education were significantly correlated with depression status in patients with DM. Notably, marital status was no longer significant in the multivariable analysis. Model 2 examined predisposing and enabling variables. The results showed that a patient’s place of residence and participation in social activities were significantly associated with depression. Model 3 introduced predisposing, enabling, and healthcare need factors and showed that self-rated health, pain, BADL, and sleep duration were significantly associated with depression. However, the variables, including place of residence and participation in social activities, were no longer significant in Model 3.

Table 5 Variables associated with depressive symptoms in patients with diabetes mellitus in the logistic regression analysis.

Variables	Model 1		Model 2		Model 3		Model 4
	OR	SE	OR	SE	OR	SE	OR	SE
Predisposing characteristic
Age	0.985a	0.006	0.984a	0.006	0.976b	0.008	0.984	0.009
Sex: Male	0.447b	0.051	0.436b	0.051	0.522b	0.069	0.426b	0.106
Education1
Primary or middle	0.830	0.128	0.912	0.138	0.980	0.183	0.929	0.219
High school and above	0.335b	0.070	0.416b	0.087	0.459b	0.112	0.432a	0.141
Marital status: Married	0.954	0.147	0.883	0.143	0.860	0.151	0.634a	0.136
Enabling resource
Residence2
Urban areas	-	-	0.673a	0.114	0.877	0.155	1.004	0.191
Combination zone	-	-	1.128	0.241	1.268	0.233	1.536	0.361
Personal income	-	-	0.986	0.020	1.006	0.024	1.020	0.024
Months of living with children	-	-	0.977	0.025	0.966	0.028	1.002	0.034
Social activity: Yes	-	-	0.814a	0.083	0.920	0.108	0.846	0.121
Healthcare need
Self-rated health3
Fair	-	-	-	-	0.475b	0.063	0.458b	0.074
Good	-	-	-	-	0.379b	0.109	0.247b	0.070
Pain4
Mild	-	-	-	-	1.204	0.206	1.440a	0.249
Severe	-	-	-	-	2.372b	0.424	2.644b	0.532
BADL impairment: Yes	-	-	-	-	2.112b	0.270	1.886b	0.323
Disability: Yes	-	-	-	-	1.267	0.158	1.332	0.201
Sleep duration	-	-	-	-	0.676b	0.101	0.642a	0.118
Health behaviors
Outpatient visit	-	-	-	-	-	-	1.081	0.050
Hospitalization	-	-	-	-	-	-	1.006	0.063
Health service satisfaction5
Neutral	-	-	-	-	-	-	0.902	0.174
Satisfied	-	-	-	-	-	-	0.579b	0.121
DM treatment: Yes	-	-	-	-	-	-	0.912	0.162
Blood glucose test	-	-	-	-	-	-	1.005	0.053
Smoking: Yes	-	-	-	-	-	-	1.200	0.258
Drinking: Yes	-	-	-	-	-	-	1.255	0.252

^aP < 0.05, SE (weighted) in parentheses.

^bP < 0.01, SE (weighted) in parentheses.

¹Reference group: Not formally educated.

²Reference group: Rural areas.

³Reference group: Poor.

⁴Reference group: None.

⁵Reference group: Dissatisfied.

BADL: Basic activities of daily living; OR: Odds ratio.

Model 4 revealed that satisfaction with healthcare services was significantly associated with depression. Specifically, patients who were more satisfied with healthcare services (OR 0.579, P < 0.05) were less likely to have depressive symptoms. Model 4 also indicated that patients who were male (OR 0.426, P < 0.05), had a high school education or higher (OR 0.432, P < 0.05), and were married (OR 0.634, P < 0.05) were less likely to report depressive symptoms. Better self-rated health status (OR 0.458 for fair and OR 0.247 for good, P < 0.05) and longer sleep duration (OR 0.642, P < 0.05) decreased the probability of depressive symptoms in patients with DM, whereas body pain (OR 1.440 for mild and OR 2.644 for severe, P < 0.05) and impaired BADL (OR 1.886, P < 0.05; OR 1.332, P < 0.05) increased this probability (Table 5).

DISCUSSION

Summary of findings

About 41% of the patients with DM reported depressive symptoms. Depressive symptoms were significantly associated with predisposing characteristics and healthcare needs. Among these factors, physical pain and limitations in BADL were strongly associated with depression.

Comparison with literature

We found a 41% prevalence of depressive symptoms among patients with DM. This figure is consistent with previous studies in China[15,36] but higher than those reported in Brazil (22%) and the United States (25%)[37,38]. Differences in social and cultural contexts, as well as in the accessibility of mental health services in different countries, could have contributed to this phenomenon. Additionally, differences in the depression scales used in these studies may also explain the differences in prevalence. Despite the high prevalence of depressive symptoms in patients with DM, depression remains underdiagnosed and undertreated in China. According to a meta-analysis, the treatment rate for major depressive disorders in China is only 19.5%[39], suggesting a need to improve depression screening and mental healthcare accessibility.

Predisposing characteristics, including sex, education, and marital status, were significantly associated with depressive symptoms in Model 4. We found that male patients were less likely to have depressive symptoms than female patients were, consistent with previous research identifying female sex as a risk factor for depression in individuals with DM[10,14,25,29]. Patients who were highly educated and married were found to have a lower likelihood of reporting depressive symptoms in this study, which is consistent with previous studies[14,40]. Individuals with lower levels of education or without a spouse may have limited access to health care services, a lack of family care, or an unhealthy lifestyle, contributing to poor health outcomes. Therefore, these patients may be more vulnerable to depressive symptoms. Models 1-3 showed a decreased likelihood of older patients having depressive symptoms, except in Model 4. Bai et al[41] reported a similar relationship. Their study showed that the prevalence of depression was lower in the age groups of 60-74 and ≥ 75 years than in those aged 45-59. This may be because those aged 45-59 years have higher expectations and face increased life and work stress levels. However, some studies have also suggested an increased likelihood of older patients with depressive symptoms, which is inconsistent with our findings[14,37]. This difference may be attributed to the specific demographic focus of the present study, which included only patients with DM aged 45 years or older. Further studies are required to understand these issues thoroughly.

Regarding enabling resources, Model 2 showed a significant correlation between the place of residence, social activities, and depressive symptoms. However, this association was no longer significant in Models 3 and 4. This suggests that some of the variables included in Models 3 and 4 may be covariates of these two factors.

Healthcare needs, including self-rated health, physical pain, BADL, and sleep duration, were associated with depressive symptoms in Model 4. Self-rated health is an individual’s perception of their health and has been recognized as a reflection of the actual state of health. It has also been suggested as a predictor of the onset of mental disorders[42]. In the present study, patients with better self-rated health status were less likely to report depressive symptoms. This is consistent with previous studies showing a negative association between self-rated health and depression[43-45]. Greater healthcare needs, including severe pain and impaired BADL, were associated with an increased risk of depressive symptoms in patients with DM, consistent with previous studies[46,47]. This is because the burden of greater body mass in patients with DM increases pain, leading to higher depression levels. Moreover, DM-related medical problems increase functional impairment, which also leads to higher depression levels[46]. We also found that the associations between severe pain, impaired BADL, and depressive symptoms were particularly robust. This suggests that longer sleep duration may act as a protective factor against depression in patients with DM. However, a U-shaped relationship between sleep duration and depressive symptoms has been reported in previous studies[48,49]. Therefore, healthcare professionals and community workers should pay attention to health status, pain, BADL, and sleep quality assessments when managing patients with DM, to reduce the likelihood of depressive symptoms.

This study revealed a significant association between satisfaction with healthcare services and depression. Patients with DM who reported higher levels of satisfaction with healthcare services were less likely to report depressive symptoms. Depressive symptoms may increase uncertainty about DM treatment, which can negatively impact patients' perceived service quality and lead to decreased overall satisfaction with medical services. Previous studies in different contexts have shown that depression is associated with increased healthcare utilization[20-22,24]. However, the present study did not report a similar association between depression and the number of outpatient visits or hospitalizations. Several reasons likely exist for this discrepancy, and further research is needed to examine the impact of depression on healthcare utilization among individuals with DM in China.

Limitations

Our study, with its large sample size and theoretical approach, provides a useful foundation for formulating strategies in research and clinical practice to prevent, control, and intervene regarding depression among individuals with diabetes. However, our study has several limitations. First, although data were obtained from CHARLS, a large nationally representative sample survey, the results of this study should be interpreted with caution. CHARLS surveyed community-dwelling individuals aged 45 years and older and did not specifically target individuals with DM or depressive symptoms. In addition, the survey included only middle-aged and older Chinese residents, which might have limited the generalizability of the findings to younger patients with diabetes. Second, the depression scale used in this study was the CESD-10, which applies to the assessment of individual depressive symptoms but cannot diagnose depression clinically. This might have slightly overestimated the prevalence of depression in patients with DM. Third, the development of depression may be influenced by clinical factors related to diabetes, such as duration and severity. However, the present analysis did not include these factors because the CHARLS dataset lacks clinical information on diabetes. Given the limitations of the available data, we acknowledge the importance of these issues and emphasize that caution should be exercised when interpreting our findings.

CONCLUSION

We found that nearly half of individuals with diabetes reported depressive symptoms, which were strongly associated with predisposing characteristics and healthcare needs, particularly physical pain and impairment of daily activities. Our study emphasizes the significance of enhanced screening and interventions for depression in diabetes care, along with improved management of functional impairments.

ACKNOWLEDGEMENTS

The authors appreciate the CHARLS and the field team and every respondent in the survey for their contributions.