The moderating role of workforce participation on the association between inflammatory diet and depression among US adults: findings from NHANES 2005-2018

doi:10.21203/rs.3.rs-4970979/v1

Download PDF

Research Article

The moderating role of workforce participation on the association between inflammatory diet and depression among US adults: findings from NHANES 2005-2018

https://doi.org/10.21203/rs.3.rs-4970979/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background Depression in the workforce is a highly prevalent and prominent public health problem with pro-inflammatory diet. We aimed to assess the association between inflammatory diet and depression, then to further examine the moderating role of workforce participation on the association of inflammatory diet with depression among US adults aged 20 years or above.

Methods Data from the National Health and Nutrition Examination Survey (NHANES) 2005–2018 was used. Binary logistic regression and moderating effect analysis were conducted to explore the associations between inflammatory diet, workforce participation, and depression. Restricted cubic spline (RCS) were employed to evaluate the dose–response association of dietary inflammatory index (DII) and hours worked per week with depression at different levels.

Results A total of 23,163 participants were enrolled with a depression prevalence of 7.4%. Both pro-inflammatory diet and not working were risk factors for depression. After adjusting for covariates, adults with pro-inflammatory diet (DII ≥ 0) were more likely to suffer from depression (OR = 1.30; 95%CI: 1.10–1.55). Workforce participation moderated the association between inflammatory diet and depression, especially moderating effect of not working and regular daytime working on the association of DII with depressive tendency was 0.12 and − 0.06. The protective effect of anti-inflammatory diet against depression was only found during the working period of 25 to 45 hours peer week. These effects were consistent across the analyses of subgroup and sensitivity.

Conclusions Not working may exacerbate the increase of pro-inflammatory diet to depression risk. Modest regular daytime working (25–45 hours per week) with anti-inflammatory diet may better in reducing depression risk.

Depression risk

Dietary inflammatory index

Labor force participation

Moderating effect

Cross-sectional study

Depression, is a common mental disorder associated with low mood, indifference, sadness, weariness, reduced motivation, guilt feelings, and even suicidal tendencies [1–3]. An estimated 5% of adults suffer from depression worldwide according to the latest statistics from World Health Organization (WHO) in March 2023 [4], and have accounted for 5.8% cause of death or injury from the 2021 Global Burden of Disease (GBD) study [5]. Depression among US adults increased from 8.7% in 2017–2018 to 10.6% in March 2020 [6], and was estimated to be one of the leading causes of disability by 2030 [7]. In addition, depression in the workforce is a highly prevalent and prominent public health problem [8], and dietary behavior has been considered as an important factors of depressive symptoms, however, the relationship between them is complex, multifaceted, and interactive [9, 10].

Much of existing studies have contributed to the association between diet and depression, while there are few established mechanisms to explain the relationship between dietary nutrients intake and depression, some research points out that one of the primary mechanisms is inflammation [10–12]. More precisely, inflammation may affect depression through the activation of cerebral hypothalamic-pituitary-adrenal axis, neurotransmitter system, and neural plasticity [13]. Correspondingly, dietary inflammatory index (DII) was developed to characterize the overall inflammatory potential of diet and measure the contribution of diet to inflammation [14, 15]. However, mixed results from existing studies about the association between DII and depression. For example, a recent research that included 1,743 participants evidenced that a pro-inflammatory diet was positively associated with increased risk of depressive symptoms [16], and another study that enrolled a total of 30,627 individuals highlighted this association between DII and depression was J-shaped [17], while there was a research presented a linear correlation between DII and depression in patients with chronic diseases and comorbidity [18].

Previous research has shown that the impact of workforce participation on depression among adults may vary, based on factors such as gender, age, work types or schedules, which could affect material living and health standards, the perceptions of workforce participation as well. For instance, prior research presented that self-employment was negatively associated with depression among aging workers (aged 50 years or above) [19], and a study used the data of UK biobank showed that night shift work was associated with an increased risk of depression regardless of genetic risk [20]. It has also been suggested that workforce participation and dietary behavior are related [21, 22], and some evidence for mediation by DII between shift work and depressive symptoms [9]. If workforce participation of adults affects their diet and depression risk, and depressive symptoms are influenced by DII, it would be necessary to further explore potential differences across various populations and determine the impact of specific status, types, schedules, and duration of work on the association between inflammatory diet and depression.

To our knowledge, no study has addressed the association between inflammatory diet and depression with the role of workforce participation. Accordingly, more research is need to confirm the association between inflammatory diet, workforce participation, and depression among adults. Therefore, we conducted a cross-sectional study to explore and identify the association between inflammatory diet and depression with the moderating effect of workforce participation among US adults by using the National Health and Nutrition Examination Survey (NHANES) data. We further evaluated the dose–response association of DII and hours worked per week with depression risk at different levels.

Study population

Data of the study population came from the NHANES, an ongoing national survey conducted by the Centers for Disease Control and Prevention that focused on Americans' dietary nutrition and general health. Signed informed consent from all participants before participating in the study, and all study protocols were approved by the National Center for Health Statistics' ethical review board. Detailed information about the NHANES database should be visited at https://www.cdc.gov/nchs/nhanes/index.htm. Specifically, data for this study was gathered from the NHANES 2005–2018. Our study only included individuals aged 20 years or above (N = 39,749). Following elimination for missing data on working status or PHQ-9 items (N = 5,639), DII (N = 9,085), and demographics (N = 1,862), finally 23,163 eligible participants left for analysis (Fig. 1).

[Insert Fig. 1 here]

Assessment of depression (outcome)

The PHQ-9 (Patient Health Questionnaire, 9 items) was used by the NHANES to evaluate depression. The PHQ-9 consist of nine signs and symptoms for depression: appetite problems, attention problems, depressed mood, fatigue, interest blank, sleep troubles, somatization disorders, suicidal thoughts, and worthlessness feelings [23]. Each item on a scale from "0" (not at all) to "3" (nearly every day), total score could range from 0 to 27. A PHQ-9 score was further divided into binary categories: depression with a score ≥ 10, while no depression with a score < 10 [24]. The sensitivity and specificity for detecting major depression was 88% at a cut-off of 10 [25].

Assessment of DII (exposure)

We used the modified version for DII calculation developed by Shivappa et al., the further standardized and specific calculation method has been detailed in prior studies [15, 26]. In the present study, 28 of 45 dietary nutrients were incorporated due to the NHANES data limited: alcohol, carbohydrates, caffeine, carotene, cholesterol, energy, fiber, folic acid, iron, magnesium, monounsaturated fatty acids, niacin, n-3 fatty acids, n-6 fatty acids, polyunsaturated fatty acids, protein, saturated fatty acids, selenium, thiamine, total fat, vitamin A, vitamin B2, vitamin B6, vitamin B12, vitamin C, vitamin D, vitamin E, and zinc [17]. Notably, vitamin D data were not contained in the NHANES 2005–2006, it would be ignored. According to DII, participants were divided into pro-inflammatory diet (DII ≥ 0) and anti-inflammatory diet (DII < 0) [27, 28].

Assessment of workforce participation (moderator)

We used four measures of workforce participation at the individual level: work status, hours worked, work types, and shift work [29, 30]. According to prior research [31], work status (working vs not working) in the NHANES was captured by participants reporting whether they were working at a job or business last week. Furthermore, individual disclosed the number of hours they worked last week at all jobs or businesses, and we used hours worked per week as a continuous indicator for working. Private employee, government (including federal, state, and local) employee, self-employed, and unpaid family business or farm were covered as different types of workforce participation. Shift work was assessed with the following response options: regular daytime, evening shifts, night shifts, rotating shifts, or another schedule. Evening and night shifts were combined by referring the study [9].

Assessment of covariates

Based on previous studies [9, 16, 18], demographics, health behaviors, and health status-related covariates that may affect the association between DII, workforce participation, and depression were included in our study. Demographics included age (≤ 39 years, 40–64 years, ≥ 65 years), gender (male, female), race (non-Hispanic White, non-Hispanic Black, Mexican American, other Hispanic, Other/Multiracial), education (less than high school, high school, some college, college or above), marital status (married/living with partner, widowed/divorced/separated, never married), and poverty income ratio (PIR, poor, not poor). According to the studies [28, 32, 33], health behaviors included drinking status (non-drinker, 1–5 drinks/month, 5–10 drinks/month, 10 + drinks/month), smoking status (never smoker, former smoker, current smoker), sleeping time (≤ 6 hours, 7–8 hours, ≥ 9 hours), physical activities (inactive, moderate, vigorous). Health status included body mass index (BMI, underweight, normal, overweight, obese), diabetes (yes, no), hypertension (yes, no), and self-reported perceived health (fair or poor, good, very good, excellent), by referring the research [32, 34]. Generally, a PIR value < 1 was considered as poor, a PIR value ≥ 1 was defined as not poor [23]. For detailed BMI classes: underweight (BMI < 18.5), normal (18.5 ≤ BMI < 25), overweight (25 ≤ BMI < 30), and obese (BMI ≥ 30) [32]. Self-reported diabetes, glycosylated hemoglobin (HbA1c) ≥ 6.5%, or fasting plasma glucose level ≥ 126 mg/dl were considered diabetes [35]. Self-reported high blood pressure told by doctors, the last two times measured average systolic/diastolic blood pressure of at least 140/90 mmHg, or antihypertensive medication being used were considered hypertension [7].

Statistical analysis

In our study, the statistical software R (version 4.4.1) and Storm Statistical Platform (https://www.medsta.cn/software) were used for all statistical analysis. A two-side P-value < 0.05 was considered statistically significant. We adopted the recommendations for accurate reporting in medical research statistics [36], using multiple imputation to fill in missing covariate data. In moderating analysis models, depression (Y, outcome) and DII (X, exposure) were entered as continuous variables, and workforce participation (M, moderator) was entered as a categorical variable [37, 38]. Model "2-way" from the R package "bruceR::PROCESS" was used to examine the moderating effect, with a 95% confidence interval (CI) evaluated by 1000 bootstrap resampling and 12345 seed. We employed restricted cubic spline (RCS) for the binary logistic models to assess the dose–response association of DII and hours worked per week with depression at different levels. In addition, we recruited subgroup analysis and interactions for categorical variables, gender, age, race, education, marital status, PIR, drinking status, smoking status, sleeping time, physical activities, BMI, diabetes, hypertension, and perceived health among covariates. Furthermore, in sensitivity analysis, data of non-multiple imputation and all adult participants were included, also the propensity score matching (PSM) methods were used to test our results stability.

Participants' characteristics

Table 1 presents the characteristics of the weighted study samples. Our analysis enrolled a total of 23,163 adults aged 20 years or above, of whom 7.4% with depression. Depressed adults were mainly clustered in the 40 to 64 years age group (53.4%), and were less distributed in the age group of 65 years or above (13.0%). The percentage of depressed adults was 65.1% female, while the percentage of non depressed adults was 48.7% male. Adults who with depression were more likely to be non-Hispanic White (65.6%), have some college degree (33.7%), be married or living with partner (49.1%), be not poor (71.2%), drink 1 to 5 times per month (54.9%), be current smoker (40.3%), sleep 6 hours or less (54.2%), do inactive physical activities (64.3%), be obese (48.8%), diabetes-free (79.3%), have fair or poor self-reported health (51.9%), have pro-inflammatory diet (85.3%), and not working (60.7%).

Table 1

Characteristics of participants with and without depression [mean (SE) / n (%)], weighted.^a
Characteristics	Overall (n = 23,163)	With depression (n = 2,004)	Without depression (n = 21,159)	P-value
Gender				< 0.001
Male	11,143 (47.7)	713 (34.9)	10,430 (48.7)
Female	12,020 (52.3)	1,291 (65.1)	10,729 (51.3)
Age (years)				< 0.001
≤ 39	7,742 (34.9)	629 (33.5)	7,113 (35.0)
40–64	9,903 (46.7)	1,049 (53.4)	8,854 (46.1)
≥ 65	5,518 (18.4)	326 (13.0)	5,192 (18.9)
Race				< 0.001
Non-Hispanic White	10,898 (71.3)	902 (65.6)	9,996 (71.7)
Non-Hispanic Black	4,806 (10.4)	441 (13.6)	4,365 (10.1)
Mexican American	3,462 (7.5)	293 (7.6)	3,169 (7.5)
Other Hispanic	2,038 (4.7)	242 (6.9)	1,796 (4.5)
Other/Multiracial	1,959 (6.1)	126 (6.4)	1,833 (6.1)
Education				< 0.001
Less than high school	5,294 (14.9)	698 (25.6)	4,596 (14.1)
High school	5,317 (22.5)	490 (27.4)	4,827 (22.1)
Some college	6,935 (31.9)	601 (33.7)	6,334 (31.8)
College or above	5,617 (30.7)	215 (13.3)	5,402 (32.1)
Marital Status				< 0.001
Married/living with partner	14,116 (65.2)	935 (49.1)	13,181 (66.4)
Widowed/divorced/separated	5,050 (18.1)	670 (30.5)	4,380 (17.1)
Never married	3,997 (16.7)	399 (20.4)	3,598 (16.4)
Poverty Income Ratio (PIR)				< 0.001
Poor	4,647 (13.1)	749 (28.8)	3,898 (11.9)
Not poor	18,516 (86.9)	1,255 (71.2)	17,261 (88.1)
Drinking Status				< 0.001
Non-drinker	6,526 (23.0)	549 (23.4)	5,977 (23.0)
1–5 drinks/month	11,518 (50.1)	1,085 (54.9)	10,433 (49.7)
5–10 drinks/month	1,828 (9.4)	123 (7.5)	1,705 (9.6)
10 + drinks/month	3,291 (17.5)	247 (14.2)	3,044 (17.7)
Smoking Status				< 0.001
Never smoker	12,641 (54.7)	792 (37.7)	11,849 (56.1)
Former smoker	5,888 (25.7)	453 (22.0)	5,435 (26.0)
Current smoker	4,634 (19.6)	759 (40.3)	3,875 (18.0)
Sleeping Time (hours)				< 0.001
≤ 6	8,907 (35.5)	1,114 (54.2)	7,793 (34.0)
7–8	12,469 (57.1)	717 (37.4)	11,752 (58.7)
≥ 9	1,787 (7.3)	173 (8.4)	1,614 (7.2)
Physical Activities				< 0.001
Inactive	11,380 (42.8)	1,374 (64.3)	10,006 (41.1)
Moderate	6,418 (29.6)	397 (23.4)	6,021 (30.1)
Vigorous	5,365 (27.6)	233 (12.3)	5,132 (28.8)
Body Mass Index (BMI)				< 0.001
Underweight	335 (1.4)	32 (1.6)	303 (1.4)
Normal	6,187 (28.1)	442 (23.0)	5,745 (28.6)
Overweight	7,684 (33.3)	516 (26.6)	7,168 (33.8)
Obese	8,957 (37.1)	1,014 (48.8)	7,943 (36.2)
Diabetes				< 0.001
Yes	4,207 (14.0)	525 (20.7)	3,682 (13.4)
No	18,956 (86.0)	1,479 (79.3)	17,477 (86.6)
Hypertension				0.2
Yes	4,125 (14.9)	367 (16.1)	3,758 (14.8)
No	19,038 (85.1)	1,637 (83.9)	17,401 (85.2)
Perceived Health				< 0.001
Fair or poor	5,246 (16.5)	1,182 (51.9)	4,064 (13.7)
Good	9,252 (38.7)	621 (34.5)	8,631 (39.0)
Very good	6,474 (33.7)	162 (11.5)	6,312 (35.5)
Excellent	2,191 (11.1)	39 (2.1)	2,152 (11.8)
Dietary Inflammatory Index (DII)				< 0.001
Anti-inflammatory	5,884 (28.4)	284 (14.7)	5,600 (29.5)
Pro-inflammatory	17,279 (71.6)	1,720 (85.3)	15,559 (70.5)
Work Status				< 0.001
Working	12,485 (61.5)	663 (39.3)	11,822 (63.3)
Not working	10,678 (38.5)	1,341 (60.7)	9,337 (36.7)
Notes: PIR = poverty income ratio; BMI = body mass index; DII = dietary inflammatory index.
^a Depressed adults with a PHQ-9 score ≥ 10, while non depressed adults with a PHQ-9 score < 10. Mean ± standard deviation was used to describe continuous variables, and number (constituent ratio [%]) was used to describe categorical variables. The percentages, means, and standard deviations were derived by applying the full sample 2 year Mobile Examination Center (MEC) exam weight provided in the National Health and Nutrition Examination Survey (NHANES).

[Insert Table1 here]

Association of DII, work status, work types, and shift work with depression

Table 2 demonstrates the results of binary logistic regression of the role of inflammatory diet and workforce participation (including work status, work types, and shift work) on depression in three different models. As shown in model 3, the adults with pro-inflammatory diet were more likely to suffer from depression than those with anti-inflammatory diet (OR = 1.30; 95%CI: 1.10–1.55), and working adults with lower depression risk than those who was not working (OR = 0.46; 95%CI: 0.40–0.53). For specific work types and shift work, compared with adults who was not working, eminently, depression risk of those who was working as private employee and working on rotating shift was significantly reduced by 55% (OR = 0.45; 95%CI: 0.38–0.52) and 57% (OR = 0.43; 95%CI: 0.28–0.65), respectively.

Table 2

The association of DII, work status, work types, and shift work with depression.^a
Independent Variables	OR (95CI%)
Independent Variables	Model 1	Model 2	Model 3
Dietary Inflammatory Index (DII)
Anti-inflammatory	Reference	Reference	Reference
Pro-inflammatory	2.42 (2.08, 2.81)***	1.73 (1.48, 2.03)***	1.30 (1.10, 1.55)**
Work Status
Not Working	Reference	Reference	Reference
Working	0.38 (0.33, 0.42)***	0.38 (0.33, 0.43)***	0.46 (0.40, 0.53)***
Work Types
Not working	Reference	Reference	Reference
Private employee	0.39 (0.34, 0.44)***	0.38 (0.33, 0.43)***	0.45 (0.38, 0.52)***
Government employee	0.35 (0.27, 0.46)***	0.40 (0.30, 0.53)***	0.50 (0.37, 0.68)***
Self-employed	0.34 (0.24, 0.48)***	0.39 (0.27, 0.56)***	0.52 (0.35, 0.77)**
Unpaid family business or farm	0.48 (0.12, 1.94)	0.40 (0.10, 1.56)	0.59 (0.16, 2.18)
Shift Work
Not working	Reference	Reference	Reference
Regular daytime	0.35 (0.31, 0.40)***	0.37 (0.32, 0.42)***	0.45 (0.38, 0.53)***
Evening/Night	0.57 (0.41, 0.79)***	0.51 (0.36, 0.71)***	0.58 (0.40, 0.83)**
Rotating shift	0.40 (0.27, 0.60)***	0.38 (0.25, 0.56)***	0.43 (0.28, 0.65)***
Another schedule	0.38 (0.27, 0.55)***	0.40 (0.28, 0.57)***	0.52 (0.36, 0.76)***

[Insert Table2 here]

Moderation effect of work status, work types, and shift work on the association between DII and depression

Table 3 shows the interaction and slope effects of workforce participation (including work status, work types, and shift work) between DII and depression in three different models. After adjusting for all covariates in model 3, the interaction effect of DII and workforce participation remained significant (P < 0.001). Especially, for specific work status and shift work, depressive tendency of adults strengthened with a increased level of DII when they in not working status (β = 0.12; 95%CI: 0.07–0.16), while this tendency weakened with a increased level of DII when they working on the regular daytime (β = -0.06; 95%CI: -0.10–0.01).

Table 3

The moderating effect of work status, work types, and shift work on the association between DII and depression.^a
Moderating Effect	Independent Variables	β (95CI%)			P-value ^b
Moderating Effect		Model 1	Model 2	Model 3	P-value ^b
Interaction	DII × Work Status				< 0.001
Slope analysis	Not Working	0.40 (0.35, 0.45)***	0.26 (0.21, 0.30)***	0.12 (0.07, 0.16)***
Slope analysis	Working	0.19 (0.15, 0.23)***	0.07 (0.03, 0.12)**	-0.04 (-0.08, − 0.00)*
Interaction	DII × Work Types				< 0.001
Slope analysis	Not working	0.40 (0.35, 0.45)***	0.26 (0.21, 0.30)***	0.12 (0.07, 0.16)***
	Private employee	0.22 (0.17, 0.27)***	0.09 (0.04, 0.14)***	-0.03 (-0.08, 0.02)
	Government employee	0.12 (0.01, 0.23)*	0.01 (-0.09, 0.12)	-0.09 (-0.19, 0.01)
	Self-employed	0.11 (-0.02, 0.24)	0.03 (-0.10, 0.16)	-0.06 (-0.18, 0.06)
	Unpaid family business or farm	0.18 (-0.68, 1.05)	0.09 (-0.75, 0.94)	0.01 (-0.78, 0.79)
Interaction	DII × Shift Work				< 0.001
Slope analysis	Not working	0.40 (0.35, 0.45)***	0.26 (0.21, 0.30)***	0.12 (0.08, 0.16)***
	Regular daytime	0.17 (0.12, 0.22)***	0.05 (0.00, 0.10)*	-0.06 (-0.10, -0.01)*
	Evening/Night	0.31 (0.16, 0.45)***	0.18 (0.04, 0.32)*	0.05 (-0.08, 0.18)
	Rotating shift	0.17 (0.02, 0.32)*	0.06 (-0.09, 0.20)	-0.03 (-0.16, 0.11)
	Another schedule	0.20 (0.06, 0.35)**	0.10 (-0.04, 0.24)	-0.05 (-0.18, 0.08)

[Insert Table3 here]

Association of DII and hours worked per week with depression

We further clarified the association of DII and hours worked per week with depression risk at different levels by employing RCS (Fig. 2). We observed a significant decreased depression risk as hours worked per week increased until reaching around 45 hours (p value for nonlinearity < 0.05), and the adults with pro-inflammatory diet were always suffering from higher depression risk than those with anti-inflammatory diet on equal hours worked peer week (Fig. 2A). Furthermore, depression risk increased with DII in a linear pattern across all working duration groups (p value for nonlinearity > 0.05), and protective effect of anti-inflammatory diet for anti-depression (OR < 1) was only found during the working period of 25 to 45 hours peer week (Fig. 2B).

[Insert Fig. 2 here]

Subgroup analysis

Subgroup analysis was performed to further evaluate the association between DII and depression, with stratification factors including gender (male, female), age (≤ 39 years, 40–64 years, ≥ 65 years), race (non-Hispanic White, non-Hispanic Black, Mexican American, other Hispanic, Other/Multiracial), education (less than high school, high school, some college, college or above), marital status (married/living with partner, widowed/divorced/separated, never married), PIR (poor, not poor), drinking status (non-drinker, 1–5 drinks/month, 5–10 drinks/month, 10 + drinks/month), smoking status (never smoker, former smoker, current smoker), sleeping time (≤ 6 hours, 7–8 hours, ≥ 9 hours), physical activities (inactive, moderate, vigorous), BMI (underweight, normal, overweight, obese), diabetes (yes, no), hypertension (yes, no), and self-reported perceived health (fair or poor, good, very good, excellent). No correlation with the p for interaction meeting statistical significance was detected on all stratification factors (P for interaction > 0.05) (Fig. 3).

[Insert Fig. 3 here]

Sensitivity analysis

As shown in Supplementary eTable1, eTable2, eFig.1, and eFig.2, when analyzing the data without multiple imputation, or including all adult participants who aged 20 years or above, the results were reasonably consistent with our primary models. We also conducted additional PSM methods, and focusing on demographics, health behaviors (drinking and smoking) and health status (diabetes and hypertension), to mitigate the potential confounders between DII and depression. The PSM results also indicated the interaction effect of DII and workforce participation remained significant (P < 0.001), especially depressive tendency of adults strengthened with a increased level of DII when they in not working status (β = 0.13; 95%CI: 0.07–0.19), while weakened with a increased level of DII when they working on the regular daytime (β = -0.06; 95%CI: -0.13-0.01) (Supplementary eTable3).

In our nationally representative cross-sectional study of 23,163 US adults, we observed that high DII and not working were risk factors for depression. Workforce participation moderated the significant correlation between DII and depression, especially with the 0.12 and − 0.06 moderating effect of not working and regular daytime working on the association of DII with depressive tendency. Furthermore, positive effect of anti-inflammatory diet for anti-depression was found only during the working period of 25 to 45 hours peer week. Subgroup analysis and interaction evaluations demonstrated that the association between DII and depression risk remained consistent across all stratification factors. These association and moderating effect persisted even when data of non-multiple imputation and all adult participants included were used to test. These findings suggest that modest regular daytime working would weaken the positive association of pro-inflammatory diet with increased depression risk, while not working would strengthen this association.

Although the association between DII and depression has received increasing attention, but results and conclusions are mixed [16–18, 28]. One recent study that enrolled a total of 30,627 participants showed a J-shaped association between DII and depression [17], while another research presented a linear association between DII and depression in patients with chronic diseases and comorbidity [18]. Our results supported the association between DII and depression was linear among US adults aged 20 years or above, as for J-shaped association, may be due to the moderating effect of workforce participation on the association between DII and depression were ignored. In addition, some evidence that the population with anti-inflammatory diet (DII < 0) was less likely to suffer from depression than those with pro-inflammatory diet (DII ≥ 0) [28], and we further found that this protective response was active only during the modest working period (25 to 45 hours peer week). Accordingly, more research is need to confirm the association between DII and depression among adults.

Depression was confirmed as a highly prevalent and prominent public health problem in workforce [8], and major depressive symptom was an independent risk factor for all-cause mortality [39]. The relationship between workforce participation and depression has been discussed repeatedly, but remains inconclusive. Previous studies have shown that self-employed was a protective factor against depression [19], night shift work was associated with an increased risk of depression [20], our results agreed with these arguments significantly. Besides, a study of 5958 adults aged between 18 and 65 years indicated that a positive association of long working hours (more than 40 hours per week) with increased depression risk (OR = 1.74; 95%CI: 1.43–2.12) [40], yet another research argued that burdens of depression attributable to exposure to long working hours (41 hours or above per week) appeared insufficient evidence, examinations for the association between long working hours and risk of depression were needed to address the limitations of the current evidence [41]. Further, we presented a significant decreased depression risk as working hours per week increased until reaching around 45 hours peer week (p value for nonlinearity < 0.05) in this study.

Successive studies have continuously shown that depression risk distributes disproportionately in population with different demographic, socioeconomic, diet, and health-related characteristics. For example, black or brown/mixed Brazilians were more likely to suffer from untreated depression, employment was one of the main contributing factors to these racial inequalities in depression [42]. Male and female workers at various depression risk when with under-employed or over-employed hours worked [43]. A recent research demonstrated that higher DII was associated with depression risk increased in participants with chronic diseases and comorbidity, especially among those aged less than 60 years and men [18]. Moreover, some evidence that DII played a mediating role between shift work and depressive symptoms [9]. Our results showed that workforce participation moderated the association between DII and depressive tendency. Correspondingly, more research is needed on demographic, socioeconomic, diet, and health-related disparities in the relationship between workforce participation, DII, and depression, especially for aging, ethnicity/race, gender, and chronic diseases (e.g., diabetes, hypertension).

The notable strength of this study lies in its status as the first and latest cross-section exploration for the moderating effect of workforce participation on the association between DII and depression among adults based on the NHANES data, encompassing a substantial and representative sample size. Nonetheless, several limitations are needed to be acknowledged in our study. Firstly, self-reported responses derived from designated cross-sectional survey may be at a risk of interference with recall bias, and it does not support us to make causal claims for the relationship between DII and depression, workforce participation and depression, or workforce participation and DII. Secondly, multiple potential confounding factors could affect workforce participation, DII, and depression. Despite incorporating numerous most likely relevant covariates into all our models, it remains challenging to fully eliminate the influences of other potential confounding variables. Thirdly, the external validity of the study is limited due to the fact that the NHANES database only uses the PHQ-9 as depressive symptom assessment tool. However, despite these limitations, we firmly believe that this study demonstrates a significant moderating effect of workforce participation on the association between DII and depression in US adults.

According to this cross-sectional study, regular daytime working could weaken the positive association of pro-inflammatory diet with increased depression risk, not working could strengthen this association, and protection of anti-inflammatory diet against depression was effective just in the working duration of 25 to 45 hours peer week. Furthermore, further extensive prospective studies are needed to better understand the causality between workforce participation, DII, and depression among adults.

NHANES: National Health and Nutrition Examination Survey; DII: Dietary Inflammatory Index; PHQ: Patient Health Questionnaire; PIR: Poverty Income Ratio; BMI; Body Mass Index; RCS: Restricted Cubic Spline.

Ethics approval and consent to participate

Data for this study was gathered from the National Health and Nutrition Examination Survey (NHANES) 2005-2018. The survey was carried out according to the guidelines of the Declaration of Helsinki, and approved by the Research Ethics Review Board of the National Center for Health Statistics. All participants provided informed consent before enrollment.

Consent for publication

All authors have read and approved the final version of the manuscript being submitted. We declared that we have not published, posted, or submitted any related papers from the same study, in whole or in part.

Availability of data and materials

Data for this study was gathered from the National Health and Nutrition Examination Survey (NHANES) 2005-2018. All NHANES data for this study are publicly available and can be visited at: https://wwwn.cdc.gov/nchs/nhanes (accessed on 31 May 2024).

Competing interests

The authors declare no competing interests.

Funding

This work was supported by the National Natural Science Foundation of China (72374110), and the Postgraduate Research & Practice Innovation Program of Jiangsu Province (KYCX23_1900).

Authors' contributions

ZYS, XC and DFQ contributed to the study design. ZYS, HRT and HLG preformed the data analysis. ZYS and HRT wrote the manuscript. ZYS, XC and DFQ revised the manuscript. All authors have read and approved the final manuscript.

Acknowledgements

We gratefully acknowledge the National Health and Nutrition Examination Survey (NHANES). We also thank Jiada James Zhan of Nutrition & Health Sciences Doctoral Program, Emory University, USA, for his helpful suggestions on the calculation of dietary inflammation index.

Boswell EB, Stoudemire A, Adams RJ. Major depression in the primary care setting. Am J Med. 1996;101(6):S3–9.
Herwig U, Opialla S, Cattapan K, Wetter TC, Jäncke L, Brühl AB. Emotion introspection and regulation in depression. Psychiatry Research: Neuroimaging. 2018;277:7–13.
Swetlitz N. Depression's Problem With Men. AMA J Ethics. 2021;23(7):E586–9.
World Health Organization. (2023). Depression, 31 March. https://www.who.int/news-room/fact sheets/detail/depression. (Accessed 4 July 2024).
World Health Organization. (2024). Institute of Health Metrics and Evaluation. Global Health Data Exchange (GHDx). https://vizhub.healthdata.org/gbd-results/. (Accessed 4 July 2024).
Daly M, Sutin AR, Robinson E. Depression reported by US adults in 2017–2018 and March and April 2020. J AFFECT DISORDERS. 2021;278:131–5.
Shen W, Su Y, Guo T, Ding N, Chai X. The relationship between depression based on patient health questionaire-9 and cardiovascular mortality in patients with hypertension. J AFFECT DISORDERS. 2024;345:78–84.
McIntyre RS, Liauw S, Taylor VH. Depression in the workforce: the intermediary effect of medical comorbidity. J AFFECT DISORDERS. 2011;128(Suppl 1):S29–36.
Wirth MD, Shivappa N, Burch JB, Hurley TG, Hébert JR. The Dietary Inflammatory Index, shift work, and depression: Results from NHANES. HEALTH PSYCHOL. 2017;36(8):760–9.
Marx W, Lane M, Hockey M, Aslam H, Berk M, Walder K, Borsini A, Firth J, Pariante CM, Berding K, et al. Diet and depression: exploring the biological mechanisms of action. MOL PSYCHIATR. 2021;26(1):134–50.
Zhao D, Long X, Wang J. Association of dietary vitamin C intake with depression in adults: A cross-sectional study of NHANES from 2005 to 2020. J AFFECT DISORDERS. 2024;358:113–20.
Berk M, Williams LJ, Jacka FN, O'Neil A, Pasco JA, Moylan S, Allen NB, Stuart AL, Hayley AC, Byrne ML, et al. So depression is an inflammatory disease, but where does the inflammation come from? BMC MED. 2013;11:200.
Han Q, Yu J. Inflammation: a mechanism of depression? NEUROSCI BULL. 2014;30(3):515–23.
Varaee H, Mirzaei M, Salehi-Abargouei A, Ahmadi N, Hosseinzadeh M. Evaluation of lifestyle and dietary inflammatory score and their relationship with the odds of depression, stress, and anxiety in adults living in Yazd, Iran; based on YaHS and TAMYZ cohort study. J AFFECT DISORDERS. 2024;347:486–91.
Shivappa N, Steck SE, Hurley TG, Hussey JR, Hébert JR. Designing and developing a literature-derived, population-based dietary inflammatory index. PUBLIC HEALTH NUTR. 2014;17(8):1689–96.
Shakya PR, Melaku YA, Shivappa N, Hébert JR, Adams RJ, Page AJ, Gill TK. Dietary inflammatory index (DII®) and the risk of depression symptoms in adults. CLIN NUTR. 2021;40(5):3631–42.
Zhao L, Sun Y, Liu Y, Yan Z, Peng W. A J-shaped association between Dietary Inflammatory Index (DII) and depression: A cross-sectional study from NHANES 2007–2018. J AFFECT DISORDERS. 2023;323:257–63.
Jiang C, Yin H, Liu A, Liu Q, Ma H, Geng Q. Dietary inflammatory index and depression risk in patients with chronic diseases and comorbidity. J AFFECT DISORDERS. 2022;301:307–14.
Patel PC, Reid SW, Wolfe MT. Self-employment, depression, and older individuals: A cross-country study. J AFFECT DISORDERS. 2020;265:175–84.
Chen Y, Yang H, Zhang Y, Zhou L, Lin J, Wang Y. Night shift work, genetic risk, and the risk of depression: A prospective cohort study. J AFFECT DISORDERS. 2024;354:735–42.
Floria I, Kontele I, Grammatikopoulou MG, Sergentanis TN, Vassilakou T. Quality of Life of Hemodialysis Patients in Greece: Associations with Socio-Economic, Anthropometric and Nutritional Factors. Int J Environ Res Public Health. 2022;19(22):15389.
Li Y, Peng X, Wang X, Lin R, Liu X, Meng F, Liu X, Li L, Bai R, Wen S et al. Association of shift work and dietary inflammatory potential with all-cause death among us hypertensive population: national health and nutrition examination study, 2005–2010. BMC Public Health 2023, 23(1).
Zhang L, Yin J, Sun H, Dong W, Liu Z, Yang J, Liu Y. The relationship between body roundness index and depression: A cross-sectional study using data from the National Health and Nutrition Examination Survey (NHANES) 2011–2018. J AFFECT DISORDERS. 2024;361:17–23.
Kroenke K, Spitzer RL, Williams JBW. The PHQ-9: Validity of a brief depression severity measure. J GEN INTERN MED. 2001;16(9):606–13.
Levis B, Benedetti A, Thombs BD. Accuracy of Patient Health Questionnaire-9 (PHQ-9) for screening to detect major depression: individual participant data meta-analysis. BMJ-BRIT MED J 2019:l1476.
Zhan JJ, Hodge RA, Dunlop AL, Lee MM, Bui L, Liang D, Ferranti EP. Dietaryindex: A User-Friendly and Versatile R Package for Standardizing Dietary Pattern Analysis in Epidemiological and Clinical Studies. bioRxiv 2023.
Sun M, Wang L, Guo Y, Yan S, Li J, Wang X, Li X, Li B. The Association Among Inflammatory Diet, Glycohemoglobin, and Cognitive Function Impairment in the Elderly: Based on the NHANES 2011–2014. J ALZHEIMERS DIS. 2022;87(4):1713–23.
Wang F, Sun M, Wang X, Wu Z, Guo R, Yang Y, Wang Y, Liu Y, Dong Y, Wang S, et al. The mediating role of dietary inflammatory index on the association between eating breakfast and depression: Based on NHANES 2007–2018. J AFFECT DISORDERS. 2024;348:1–7.
Chang MJ, Vidafar P, Birk JL, Shechter A. The relationship of shift work disorder with symptoms of depression, anxiety, and stress. J Affect Disorders Rep. 2024;15:100713.
Guo Y, Qian D. Workforce Participation and Mortality Risk Among Chinese Older Adults: A Nationwide Population-Based Prospective Study. J GERONTOL B-PSYCHOL. 2023;78(11):1947–56.
Tsai AJ. Disparities in osteoporosis by race/ethnicity, education, work status, immigrant status, and economic status in the United States. EUR J INTERN MED. 2019;64:85–9.
Christensen K, Gleason CE, Mares JA. Dietary carotenoids and cognitive function among US adults, NHANES 2011–2014. NUTR NEUROSCI. 2020;23(7):554–62.
Li M, Yu X, Zhang W, Yin J, Zhang L, Luo G, Liu Y, Yang J. The association between weight-adjusted-waist index and depression: Results from NHANES 2005–2018. J AFFECT DISORDERS. 2024;347:299–305.
Kemp JM, Taylor VH, Kanagasabai T. Access to healthcare and depression severity in vulnerable groups the US: NHANES 2013–2018. J AFFECT DISORDERS. 2024;352:473–8.
Ruan Z, Xie X, Yu H, Liu R, Jing W, Lu T. Association between dietary inflammation and erectile dysfunction among US adults: A cross-sectional analysis of the National Health and Nutrition Examination Survey 2001–2004. FRONT NUTR. 2022;9:930272.
Mansournia MA, Nazemipour M, Adams RJ. Recommendations for accurate reporting in medical research statistics. Lancet (British edition). 2024;403(10427):611–2.
Wang L, Wang Y, Luo Y, Li Y, Li J. The mediating and moderating effect of health-promoting lifestyle on frailty and depressive symptoms for Chinese community-dwelling older adults: A cross-sectional study. J AFFECT DISORDERS. 2024;361:91–6.
Lu J, Zhang C, Xue Y, Mao D, Zheng X, Wu S, Wang X. Moderating effect of social support on depression and health promoting lifestyle for Chinese empty nesters: A cross-sectional study. J AFFECT DISORDERS. 2019;256:495–508.
Lee S, Lee JP, Lee J, Park JY, Kim EY. Association between depressive symptoms and the risk of all-cause and cardiovascular mortality among US adults. Prog Neuropsychopharmacol Biol Psychiatry. 2023;125:110755.
Yan T, Ding X, Xie T, Lan T, Niu D, Li J, Guo D, Wang M, Wang M. Physical activity (PA) influences the risk of depression associated with long working hours. J AFFECT DISORDERS. 2023;321:227–33.
Rugulies R, Sørensen K, Di Tecco C, Bonafede M, Rondinone BM, Ahn S, Ando E, Ayuso-Mateos JL, Cabello M, Descatha A, et al. The effect of exposure to long working hours on depression: A systematic review and meta-analysis from the WHO/ILO Joint Estimates of the Work-related Burden of Disease and Injury. ENVIRON INT. 2021;155:106629.
Mrejen M, Hone T, Rocha R. Socioeconomic and racial/ethnic inequalities in depression prevalence and the treatment gap in Brazil: A decomposition analysis. SSM - Popul Health. 2022;20:101266.
Kim S, Jeong W, Jang S, Park E, Park S. Is Work Hour Mismatch Associated with Depression? SAF HEALTH WORK-KR. 2021;12(1):96–101.

No competing interests reported.

SupplementaryMaterialBMCPublicHealth20240824.docx

Download PDF

Editorial decision: Revision requested
27 Aug, 2024
Editor assigned by journal
26 Aug, 2024
Submission checks completed at journal
26 Aug, 2024
First submitted to journal
24 Aug, 2024

You are reading this latest preprint version

The moderating role of workforce participation on the association between inflammatory diet and depression among US adults: findings from NHANES 2005-2018

Status:

Version 1

Abstract

Figures

Introduction

Methods

Study population

Assessment of depression (outcome)

Assessment of DII (exposure)

Assessment of workforce participation (moderator)

Assessment of covariates

Statistical analysis

Results

Participants' characteristics

Association of DII, work status, work types, and shift work with depression

Association of DII and hours worked per week with depression

Subgroup analysis

Sensitivity analysis

Discussion

Conclusions

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1