Four-way decomposition of the effects of nutrient supplement and physical exercise on depression among older Chinese: a nationwide cross-sectional analysis

Abstract

Background

Studies on the associations between nutrient supplement, physical exercise, and depression among persons 65 years and older in China are still lacking. We aimed to investigate the association between nutrient supplement and depression and to explore the role of physical exercise in this association.

Methods

A total of 8,427 older Chinese adults from the Chinese Longitudinal Healthy Longevity Survey were included in this study. The Center for Epidemiologic Studies Depression Scale was used to assess participants’ depression, and nutrient supplement and physical exercise were evaluated using self-reports. Linear regression and logistic regression were performed to explore the association between nutrient supplement and depression, and a four-way decomposition method was used to explore the effects of nutrient supplement and physical exercise on depression.

Results

A total of 26.2% of participants suffered from depression. Prior to decomposition, multivariable linear regression and logistic regression models showed that older adults who received nutrient supplement had lower depression score (β: -0.704; 95% confidence interval [CI]: -1.004, -0.404) and a lower risk of depression (odds ratio: 0.710; 95% CI: 0.564, 0.894), especially for calcium and multivitamins. Four-way decomposition analysis suggested that physical exercise mediated the associations between nutrient (protein, calcium, iron, zinc, multivitamins, vitamin A/D, docosahexaenoic acid, and others) supplement and depression (all P < 0.05).

Conclusions

Calcium or multivitamin supplements were associated with inverse depression, and the reason why this association existed could be partly explained by the mediating effect of physical exercise.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-024-20995-8.

Introduction

A total of 280 million people were living with depression globally in 2019 [1], and only one-third of people with depression received formal mental health care [2]. In China, the 1-month prevalence of clinically significant depressive symptoms was 17.96% [3]. Among older adults, depressive disorder is the most common psychiatric disorder, and it has become increasingly prevalent [4]. Although effective prevention and treatment options exist, older adults are often overlooked and tend to experience residual symptoms even if treated [5]. Thus, more attention needs to be given to the prevention and treatment of depression in older adults.

Current antidepressant medications are purportedly only to have a modest benefit, and age can affect their efficacy [6]. Even some studies have suggested that medication can do more harm than good [7, 8], there are calls to revisit the impact that nutrition can have on mental health [9]. The exciting evidence indicated a potential link between nutritional factors and depression [10, 11], but studies have focused mainly on vitamins and paid less attention to older Chinese adults. Thus, the strength of the association between various nutrient supplements (including protein, calcium, etc.) and depression in older Chinese adults still needs to be further explored.

In addition, physical exercise has been demonstrated to be a protective factor against incident depression and to be a treatment for depression [12]. Studies have shown that nutrient supplement users are more likely to perform physical activity [13] and improve their physical capacity [14]. Studies have also shown that physical exercise can promote the absorption of nutrients [15]. An animal experiment revealed the protective effect of nutrient supplement (such as vitamin D) and treadmill exercise against depression [16]. However, it remains unclear whether the joint effect of nutrient supplement and physical exercise on depression is through mediation, interaction, or a combination of the two.

To address the knowledge gaps mentioned above, this study assessed the strength of the association between various nutrient supplements and depression in older Chinese adults and explored the effects of nutrient supplement and physical exercise on depression.

Materials and methods

Data and participants

The study data were obtained from the Chinese Longitudinal Healthy Longevity Survey (CLHLS). The CLHLS is a nationwide survey of older adults (≥ 65 years) conducted in a randomly selected half of the counties and cities in 23 of the 31 provinces and autonomous regions in China for eight follow-up surveys (1998, 2000, 2002, 2005, 2008, 2011, 2014, and 2018) [17]. The CLHLS study was approved by the Research Ethics Committee of Peking University (IRB00001052-13074), and all participants or their surrogate respondents provided a written informed consent.

This study used the most recent cross-sectional data from 2018 because it contained depression scales. A total of 15,874 samples were included, and 8427 valid samples were obtained by excluding any samples with missing values or ‘I do not know/do not know’ responses for the variable of interest. The flow chart of the participant selection process is shown in Fig. 1. This study was conducted in the form of a household survey where variables were collected through face-to-face interviews.

Fig. 1.

Flowchart of the selection of the study population from participants of the Chinese Longitudinal Healthy Longevity Survey

Measurements

Outcome, exposure and mediator measures

Depression as an outcome measure was assessed by the 10-item Center for Epidemiologic Studies Depression (CES-D-10) [18]. The CES-D-10 includes ten items using a four-point metric, ranging from 0 = ‘Rarely’ to 3 = ‘Most of the time’. The total score of the CES-D-10 range from 0 to 30, with higher score suggesting more symptoms of depression. A score ≥ 10 on the CES-D-10 was considered to indicate depression, which can be found in previous studies and has been fully verified in the depression measurement of older adults in China [19, 20]. In this study, the Cronbach alpha coefficient of CES-D-10 was 0.737.

Nutrient supplement was obtained by asking the question ‘Do you usually take nutrient supplement?’, which included protein, or calcium, or iron, or zinc, or multivitamins, or vitamin A/D, or docosahexaenoic acid (DHA), or others (such as lycopene, lecithin, lutein, coenzyme Q10). In addition, the number of nutrient supplements was calculated.

Physical exercise refers to purposeful fitness activities, such as walking, playing balls, running, qigong, etc. Physical exercise was divided into never exercising, past exercising, and current exercising by the questions ‘Do you exercise regularly now?’ and ‘Do you exercise regularly in the past?’. In the four-way decomposition analysis, physical exercise was grouped into non-exercise (never exercising or past exercising) and exercise.

Potential confounding factors

The accuracy of the results was improved by adjusting for confounding factors, including demographic characteristics (age and sex), socioeconomic characteristics (marital status, education, residence, co-residence, and self-rated economic status), lifestyle behaviors (smoking, drinking, and sleeping), and health status (body mass index [BMI], activities of daily living [ADL] and self-rated health status).

In the face-to-face interview, the researcher obtained the information about marital status (“married” and “unmarried and others”), education (“no schooling” and “primary school or higher”), residence (“urban” and “rural”), and co-residence (“with family and other” and “alone”) through the following questions: “Current marital status”, “Years of schooling”, “The current residential area of the interviewed elderly”, “co-residence of interviewee”, respectively. Self-rated economic status was obtained by asking ‘How does your life compare locally?’ and was divided into ‘very rich or rich’, ‘fair’ and ‘very poor or poor’ [21]. Participants were grouped into never smoking (or drinking), past smoking (or drinking) and current smoking (or drinking) groups according to their smoking (or drinking) status based on the questions ‘Do you smoke (or drink) in the past?’ and ‘Do you smoke (or drink) at present?’ [22]. Sleep duration was divided into short sleep duration (≤ 6 h) and long sleep duration (≥ 9 h) according to the question of how many hours you sleep every day [23]. BMI was calculated as weight (kg) divided by the squared height in meters (m²) and was categorized as underweight (< 18.5 kg/m²), normal weight (18.5–24.0 kg/m²), overweight (24.0–28.0 kg/m²), or obese (≥ 28.0 kg/m²) [24]. ADL disability assessed by the Katz index was defined as the need for an assistant to perform one or more of six activities (bathing, dressing, toileting, indoor transferring, continence, and feeding) [25]. Self-assessed health information was obtained by asking the question ‘How do you feel about your health?’ and was rated as ‘very good or good’, ‘fair’ or ‘very poor or poor’. Self-rated health status has been identified as a reliable predictor of health and has been widely used in previous health studies conducted in China [26].

Statistical analysis

The differences between groups were analyzed by chi-square test and t-test for the dichotomous variable and the continuous variables, respectively. Multivariable linear regression and logistic regression analyses were used to assess the association between nutrient supplement and depression, and model-adjusted factors included age, sex, marital status, education, residence, co-residence, self-rated economic status, smoking, drinking, physical exercise, sleeping, BMI, ADL, and self-rated health status.

The four-way decomposition method proposed by VanderWeele et al. [27] was used to explore the relationships among nutrient supplement, physical exercise, and depression. The total effect (TE) of an exposure on an outcome, in the presence of a mediator with which the exposure may interact, can be decomposed into four components, shown as the following equation: (i) the effect due to neither mediation nor interaction (controlled direct effect, CDE), (ii) the effect only due to interaction (reference interaction, INT_ref), (iii) the effect due to both mediation interaction (mediated interaction, INT_med), and (iv) the effect due to mediation alone (pure indirect effect, PIE).

$$\text{TE}=\text{CDE}+{\text{INT}}_{\text{ref}}+{\text{INT}}_{\text{med}}+\text{PIE}$$

Sensitivity analyses were conducted to assess the robustness of the model by converting the outcome variable (depression score) into a dichotomous indicator (depressed vs. nondepressed). The data were analyzed using R version 4.3.0 and Stata (v17). Two-tailed tests with P values < 0.05 were considered to indicate statistical significance.

Results

Basic characteristics of the study participants

A total of 8,427 older adults were included in the analysis. Among them, 3,975 (47.2%) were male and 4,452 (52.8%) were female, with a mean (standard deviation [SD]) age of 82.90 (11.22) years. The mean (SD) CES-D-10 score was 7.28 (4.43), and 2,208 (26.2%) participants suffered from depression. Participants who consumed nutrient supplement were more likely to be female, non-smoking, exercising, overweight or obese, live in urban areas, have more years of education and have a rich economic status (all P < 0.05) (as shown in Table 1).

Table 1.

Basic characteristics of study participants

Variables	Total (n = 8427)	Non-Nutrient supplement (n = 7773)	Nutrient supplement (n = 654)	Statistical tests	P
Age (years, mean ± SD)	82.90 ± 11.22	82.90 ± 11.27	82.88 ± 10.68	t = 0.050	0.960
Age (years)				$x^2$ = 0.271	0.603
≦80	3851 (45.7)	3559 (45.8)	292 (44.6)
>80	4576 (54.3)	4214 (54.2)	362 (55.4)
Sex (n, %)				$x^2$ = 9.102	0.003
Male	3975 (47.2)	3704 (47.7)	271 (41.4)
Female	4452 (52.8)	4069 (52.3)	383 (58.6)
Marital status (n, %)				$x^2$ = 1.912	0.167
Married	4052 (48.1)	3755 (48.3)	297 (45.4)
Unmarried and others	4375 (51.9)	4018 (51.7)	357 (54.6)
Education (years, mean ± SD)	3.72 ± 4.38	3.65 ± 4.33	4.64 ± 4.86	t = -5.050	< 0.001
Education				$x^2$ = 11.352	< 0.001
No schooling	3640 (43.2)	3399 (43.7)	241 (36.9)
Primary school or higher	4787 (56.8)	4374 (56.3)	413 (63.1)
Residence (n, %)				$x^2$ = 55.362	< 0.001
Urban	4847 (57.5)	4380 (56.3)	467 (71.4)
Rural	3580 (42.5)	3393 (43.7)	187 (28.6)
Co-residence (n, %)				$x^2$ = 1.460	0.227
With family and other	7015 (83.2)	6459 (83.1)	556 (85.0)
Alone	1412 (16.8)	1314 (16.9)	98 (15.0)
Self-rated economic status (n, %)				$x^2$ = 56.735	< 0.001
Very poor or poor	831 (9.9)	792 (10.2)	39 (6.0)
Fair	5923 (70.3)	5508 (70.9)	415 (63.5)
Very rich or rich	1673 (19.9)	1473 (19.0)	200 (30.6)
Smoking status (n, %)				$x^2$ = 11.561	0.003
Never smoking	5731 (68.0)	5272 (67.8)	459 (70.2)
Past smoking	1300 (15.4)	1184 (15.2)	116 (17.7)
Current smoking	1396 (16.6)	1317 (16.9)	79 (12.1)
Drinking status (n, %)				$x^2$ = 1.761	0.415
Never drinking	6072 (72.1)	5587 (71.9)	485 (74.2)
Past drinking	1005 (11.9)	930 (12.0)	75 (11.5)
Current drinking	1350 (16.0)	1256 (16.2)	94 (14.4)
Physical exercise (n, %)				$x^2$ = 147.293	< 0.001
Never exercising	4842 (57.5)	4613 (59.3)	229 (35.0)
Past exercising	586 (7.0)	523 (6.7)	63 (9.6)
Current exercising	2999 (35.6)	2637 (33.9)	362 (55.4)
Sleep duration (n, %)				$x^2$ = 1.612	0.447
≤ 6 h	3104 (36.8)	2875 (37.0)	229 (35.0)
6-8 h	3201 (38.0)	2938 (37.8)	263 (40.2)
≥ 9 h	2122 (25.2)	1960 (25.2)	162 (24.8)
Body mass index (n, %)				$x^2$ = 18.922	< 0.001
Underweight	1229 (14.6)	1129 (14.5)	100 (15.3)
Normal weight	4364 (51.8)	4076 (52.4)	288 (44.0)
Overweight or obese	2834 (33.6)	2568 (33.0)	266 (40.7)
Activities of daily living (n, %)				$x^2$ = 0.644	0.422
Normal	7008 (83.2)	6472 (83.3)	536 (82.0)
Disability	1419 (16.8)	1301 (16.7)	118 (18.0)
Self-rated health status (n, %)				$x^2$ = 1.895	0.388
Very poor or poor	1078 (12.8)	991 (12.7)	87 (13.3)
Fair	3240 (38.4)	3005 (38.7)	235 (35.9)
Very good or good	4109 (48.8)	3777 (48.6)	332 (50.8)
Depression score (mean ± SD)	7.28 ± 4.43	7.38 ± 4.43	6.19 ± 4.21	t = 6.900	< 0.001
Depression
No	6219 (73.8)	5687 (73.2)	532 (81.3)	$x^2$ = 20.464	< 0.001
Yes	2208 (26.2)	2086 (26.8)	122 (18.7)

Association of nutrient supplement and depression

The association of nutrient supplement with depression is shown in Table 2. Nutrient supplement was associated with lower depression score, and this association remained after adjustment for confounders (β: -0.704; 95% confidence interval [CI]: -1.004, -0.404). Specifically, individuals who consumed calcium supplement (β: -0.596; 95% CI: -0.945, -0.246) or multivitamin supplements (β: -0.750; 95% CI: -1.420, -0.081) had lower depression score. In addition, the depression score decreased with increasing number of nutrient supplements (β: -0.191; 95% CI: -0.331, -0.050).

Table 2.

The association between nutrient supplement and depression score

Variables	Total (n = 8427)	Depression score	Unadjusted β (95% CI)	Adjusted β (95% CI)a
Nutrient supplement (n, %)
No	7773 (92.2)	7.38 ± 4.43	1 (reference)	1 (reference)
Yes	654 (7.8)	6.19 ± 4.21	-1.188 (-1.540, -0.835)*	-0.704 (-1.004, -0.404)*
Protein (n, %)
No	8301 (98.5)	7.29 ± 4.43	1 (reference)	1 (reference)
Yes	126 (1.5)	6.57 ± 4.11	-0.723 (-1.502, 0.056)	-0.212 (-0.868, 0.443)
Calcium (n, %)
No	7960 (94.5)	7.34 ± 4.42	1 (reference)	1 (reference)
Yes	467 (5.5)	6.29 ± 4.36	-1.051 (-1.463, -0.638)*	-0.596 (-0.945, -0.246)*
Iron (n, %)
No	8383 (99.5)	7.29 ± 4.43	1 (reference)	1 (reference)
Yes	44 (0.5)	6.34 ± 3.92	-0.948 (-2.259, 0.364)	0.177 (-0.925, 1.279)
Zinc (n, %)
No	8386 (99.5)	7.29 ± 4.43	1 (reference)	1 (reference)
Yes	41 (0.5)	6.05 ± 4.06	-1.241 (-2.599, 0.117)	-0.042 (-1.184, 1.099)
Multivitamins (n, %)
No	8306 (98.6)	7.31 ± 4.43	1 (reference)	1 (reference)
Yes	121 (1.4)	5.82 ± 3.82	-1.487 (-2.281, -0.693)*	-0.750 (-1.420, -0.081)*
Vitamin A/D (n, %)
No	8323 (98.8)	7.30 ± 4.43	1 (reference)	1 (reference)
Yes	104 (1.2)	6.12 ± 4.01	-1.183 (-2.039, -0.327)*	-0.675 (-1.397, 0.046)
Docosahexaenoic acid (n, %)
No	8399 (99.7)	7.29 ± 4.43	1 (reference)	1 (reference)
Yes	28 (0.3)	5.89 ± 3.73	-1.395 (-3.038, 0.247)	-0.330 (-1.709, 1.049)
Others (n, %)
No	8308 (98.6)	7.30 ± 4.42	1 (reference)	1 (reference)
Yes	119 (1.4)	6.07 ± 4.48	-1.234 (-2.035, -0.433)*	-0.574 (-1.248, 0.100)
Number of nutrients (mean ± SD)	-	-	-0.412 (-0.577, -0.246)*	-0.191 (-0.331, -0.050)*

OR odds ratio, CI confidence interval

^*indicates P < 0.05

^aAdjusted for age, sex, marital status, education, residence, co-residence, self-rated economic status, smoking, drinking, physical activity, sleeping, body mass index, activities of daily living, and self-rated health status

Four-way decomposition analysis

To further clarify the mechanism by which nutrient supplement was associated with depression, a four-way decomposition analysis was used to assess the effect of nutrient supplement on depression score due to mediation and interaction with physical exercise (Table 3, Fig. 2). According to the model accounting for the presence of a TE between nutrient supplement and physical exercise, nutrient supplement reduced depression score (TE: -1.170, P < 0.001). A mediating effect of physical exercise on the relationship between nutrient supplement and depression score was observed (PIE_: -0.307, P < 0.001). For calcium and multivitamin supplements, we also found a mediating effect of physical exercise (all P < 0.001), with no interaction observed. Specifically, physical exercise mediated the effects of 26.2% of nutrient supplement on depression, 29.9% of calcium supplement on depression, and 30.0% of multivitamin supplements on depression.

Table 3.

The effect of nutrient supplement on depression score due to mediation and interaction with physical exercise

	Total Effect		CDE		INTref		INTmed		PIE
	Est (SE)	P	Est (SE)	P	Est (SE)	P	Est (SE)	P	Est (SE)	P
Nutrient supplement	-1.170 (0.177)	< 0.001	-0.816 (0.256)	0.001	-0.029 (0.111)	0.795	-0.018 (0.071)	0.795	-0.307 (0.039)	< 0.001
Protein	-0.853 (0.397)	0.032	-0.124 (0.561)	0.825	-0.247 (0.255)	0.332	-0.157 (0.165)	0.343	-0.326 (0.076)	< 0.001
Calcium	-1.074 (0.207)	< 0.001	-0.646 (0.300)	0.032	-0.065 (0.131)	0.618	-0.043 (0.086)	0.618	-0.321 (0.044)	< 0.001
Iron	-0.996 (0.691)	0.150	0.779 (1.137)	0.493	-0.611 (0.459)	0.184	-0.630 (0.493)	0.201	-0.533 (0.121)	< 0.001
Zinc	-1.428 (0.712)	0.045	0.356 (1.180)	0.763	-0.591 (0.476)	0.215	-0.637 (0.529)	0.229	-0.556 (0.123)	< 0.001
Multivitamins	-1.399 (0.394)	< 0.001	-1.315 (0.644)	0.041	0.185 (0.268)	0.488	0.151 (0.220)	0.491	-0.420 (0.080)	< 0.001
Vitamin A/D	-1.154 (0.427)	0.007	-0.990 (0.692)	0.153	0.127 (0.289)	0.661	0.094 (0.216)	0.662	-0.385 (0.085)	< 0.001
Docosahexaenoic acid	-1.293 (0.925)	0.162	1.112 (1.417)	0.433	-0.996 (0.575)	0.083	-0.928 (0.610)	0.128	-0.481 (0.157)	0.002
Others	-1.160 (0.405)	0.004	-0.610 (0.629)	0.332	-0.104 (0.266)	0.697	-0.075 (0.193)	0.698	-0.371 (0.080)	< 0.001
Number of nutrients	-0.409 (0.091)	< 0.001	-0.229 (0.133)	0.085	-0.027 (0.055)	0.616	-0.008 (0.016)	0.617	-0.145 (0.021)	< 0.001

Adjusted for age, sex, marital status, education, residence, co-residence, self-rated economic status, smoking, drinking, sleeping, body mass index, activities of daily living, and self-rated health status

CDE controlled direct effect, INT_med mediated interaction, INT_ref reference interaction, PIE pure indirect effect, Est estimate, SE standard error

Fig. 2.

Four-way decomposition diagram of nutrient supplement, physical exercise and depression score. a nutrients; b protein; c calcium; d iron; e zinc; f multivitamins; g vitamin A/D; h DHA; i other nutrients. The model was adjusted for age, sex, marital status, education, residence, co-residence, self-rated economic status, smoking, drinking, sleeping, body mass index, activities of daily living, and self-rated health status. DHA: docosahexaenoic acid

Sensitivity analyses

The sensitivity analysis results showed that nutrient supplement (odds ratio [OR]: 0.710; 95% CI: 0.564, 0.894) were still associated with a lower risk of depression (Table S1). The association between nutrient supplement and depression was still partly mediated by physical exercise (Table S2).

Discussion

Summary of findings

In this study, which included a total of 8,427 participants, people who consumed nutrient supplement had lower depression score and a lower risk of depression. Nutrient supplement involved in reducing depression score and risk mainly include calcium and multivitamins. Further four-way decomposition analysis showed that physical exercise played a mediating role in the association between nutrient supplement and depression. Our findings highlight the importance of nutrient supplement (calcium or multivitamin supplements) and physical exercise in combating depression in older adults, which may provide evidence in support of the theory basis for public health policy-making.

Association between nutrient supplement and depression

Our study revealed that nutrient supplement use was inversely associated with depression score and the prevalence of depression, and the same results have been reported in previous studies [10, 11]. Moreover, this study revealed that the score and prevalence of depression were also related to the number of nutrient supplements used; that is, the greater the number of nutrient supplements used, the lower the score and risk of depression. Among many nutrient supplements, calcium and multivitamin supplements were closely associated with the score and prevalence of depression in this study. One study revealed a benefit of mineral intake against depression, with an adjusted OR (95% CI) of 0.64 (0.47–0.88) for calcium in the highest versus lowest tertile [28], similar to our results. Multivitamins (e.g., vitamins A, C, D, E, and K and the B vitamins) are groups of chemically diverse organic compounds that are essential for maintaining normal metabolism. Studies have shown that depression is linked to a deficiency of vitamin C and vitamin B [29]. Vitamin D has been found to have antioxidant, anti-inflammatory, pro-neurogenic, and neuromodulatory properties and thus may act in a similar manner to classic antidepressants [30]. A randomized controlled study showed that vitamin D supplement could improve depression score in persons aged 60 and over [31]. Another systematic review of randomized controlled trials revealed that intervention with B vitamins (e.g., B₁, B₂, B₆, B₉, and B₁₂) and/or vitamin D may be an effective and well-tolerated adjuvant strategy for improving the symptoms of depression [32]. Studies have also shown that supplemental intake of calcium plus vitamin D can decrease the risk of depressive symptoms in older adults and postpartum women [33, 34]. Therefore, current evidence suggests that the use of nutrient supplement helps reduce the risk of depression and reduce depressive symptoms and that a combination of supplements (such as B vitamins along with vitamin D, or calcium plus vitamin D) could be even more effective.

Mediation of physical exercise

In this study, we also found that physical exercise played a mediating role between the use of nutrient supplement and the score and prevalence of depression. Studies have shown that physical exercise itself can reduce the risk of depression or alleviate symptoms of depression [12, 35]. There are many possible mechanisms. For example, exercise can affect the levels of neurotrophic factors in brain cells, thereby promoting neuronal health and functional recovery in circuits and enhancing drug treatment response [36]. Second, exercise reduces the level of inflammation in the nervous system by reducing the levels of transcriptional co-activators of pro-inflammatory cytokine synthesis and release, as well as increasing the levels of anti-inflammatory cytokines [12, 37].

In addition, in this study, we found a mediating effect of physical exercise, which mediated the effects of 29.9% calcium supplement and 30.0% multivitamin supplements on depression. A previous study reported that individuals who consumed multiple-vitamin supplement were more likely to perform physical exercise (OR: 1.30; 95% CI: 1.09, 1.55), participate in team sports (OR: 1.42; 95% CI: 1.19, 1.70), and participate in other organized sports (OR: 1.32; 95% CI: 1.10, 1.59) than individuals who did not consume multiple-vitamin supplement were [13]. A recent study revealed that vitamin D3 had a positive impact on aerobic and anaerobic performance, and vitamin D3 supplement at a daily dose of 6000 IU for eight weeks was sufficient to improve physical capacity [14]. A randomized, double-blind, controlled study demonstrated the potential of calcium β-hydroxy-β-methylbutyrate and vitamin D3 supplements to enhance muscle strength and physical functionality in older adults [38]. The above findings suggest that nutrient supplement can help to enhance physical exercise. Study has also reported that physical exercise promoted the absorption of nutrients in the body [15], which could regulate the association between nutrient supplement and depression at another level. This suggested that taking nutrient supplement would increase physical exercise and that increased physical exercise would promote the absorption of supplement and reduce the risk of depression. Therefore, physical exercise not only directly affects the occurrence of depression but also indirectly affects nutrient availability and depression.

Strengths and limitations

First, the data for this study were obtained from the CLHLS, a nationally representative survey study. Second, this study was one of the few studies based on a large sample to investigate the association between the use of nutrient supplement and depression in an elderly Chinese population. Third, this study explored the mediating role of physical exercise based on four-way decomposition analysis and provides a theoretical basis for the mechanism by which nutrient supplement reduces the risk of depression. This study also has the following shortcomings. First, this was a cross-sectional study, and the conclusions obtained need to be verified in a prospective cohort study. In addition, it is unfortunate that the continuous variables with detailed information on the duration and frequency of physical exercise as well as the dosage of nutrient supplement were not provided in the CLHLS project, which therefore not included in the statistical analyses of our present study. Finally, the sample size of 8427 cases included in this study may be relatively small for national data and the survey has not been carried out annually. We will continue to focus on the possible changes of this project in the future, including the sample size, survey frequency and the indices included in the questionnaire.

Conclusions

In this national epidemiological study, we found that the use of nutrient supplement, especially calcium and multivitamins, was associated with a lower risk of depression. Furthermore, we also found that physical exercise played a mediating role in the association between nutrient supplement and depression. Therefore, moderate consumption of calcium or multivitamin supplements should be advocated among the older adults, and that the policy could contribute to improve physical exercise and reduce depressive symptoms. Furthermore, besides advocating appropriate intake of calcium or multivitamin supplements, physical exercise as a mediating variable should also encourage the public to actively participate in physical exercise, which will help to further reduce the risk of depression. However, the mechanism of supplement's risk of depression at the bottom of the well requires further study.

Abbreviations

CI

Confidence interval

CLHLS

Chinese Longitudinal Healthy Longevity Survey

CES-D-10

10-Item Center for Epidemiologic Studies Depression

DH

Docosahexaenoic acid

BMI

Body mass index

ADL

Activities of daily living

TE

Total effect

CDE

Controlled direct effect

INT_ref

Reference interaction

INT_med

Mediated interaction

PIE

Pure indirect

OR

Odds ratio

SD

Standard deviation

Authors’ contributions

Shuna Huang, Jianhui Guo and Renquan Jiang: methodology, formal analysis, writing—original draft, visualization. Ke Ma and Fuxin Lin: methodology, writing—original draft. Siying Wu, Dezhi Kang and Huangyuan Li: conceptualization, methodology, writing—review & editing. All authors read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (No. 82473659; No. 82173553), the Technology Platform Construction Project of Fujian Province (No. 2021Y2001), and the Fujian Provincial Health Technology Project (No. 2021QNA030).

Data availability

Data are from the 2018 Chinese Longitudinal Healthy Longevity Survey, which is a public, open access repository (10.18170/DVN/WBO7LK). The datasets used are available from the first author on reasonable request.

Declarations

Ethics approval and consent to participate

The data from CLHLS survey already obtained the ethical approval and informed consent and was approved by the Ethics Committee of Peking University (IRB00001052-13074). All participants or their surrogate respondents provided a written informed consent.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.