Depressive symptoms are common, heterogeneous conditions involving physical and psychological symptoms, and significantly produce health decrements (1, 2). Among non-fatal disease burden, depression will become the leading cause by 2030 according to WHO (3), whose prevalence is in excess of 300 million people (4), and it is estimated that major depressive disorder (MDD) has a lifetime prevalence (the proportion of one person who experienced the disorder at certain time during lifespan) of 16.6% (2). Combined with one of the symptoms being depressed mood, anxiety or absence of enjoyment, motivation, and interest, MDD is on the rise globally and now becoming the most concerned public health issues nowadays (5). In the United States, annual estimated highest expenditures to treat MDD could be up to $238.3 billion (6, 7), hence, depression not only brings mental disorders to human beings, but also may cause a large consumption of medical resources (8). Given the considerable influence of depression in general population, there is a pressing need to identify preventive or counter measures to deal with depression.

There is abundant evidence that the antidepressant effects could emerge with non-medical intervention represented by leisure-time physical activity (LPA) in teenagers' (9–11), middle-aged (12) and elder (13–15) groups. A recent meta-analysis, which included 49 unique prospective cohort studies and 266,939 participants, found that people with high levels of physical activity had lower odds of developing depression compared with low levels (adjusted OR = 0.83, 95% CI: 0.79 ~ 0.88; I² = 0.00) (16). Simultaneously, other meta-analyses have demonstrated that benefits of physical activity for depression not only beyond risk reduction (17–19), but also can be regarded as a useful addition to pharmacotherapy and psychotherapy (16, 20), and also can reduce the risk of cardiovascular disease in people with depression (21). One cohort study which included 107,901 Korean adults have also indicated that when an appropriate level of physical activity is maintained, it can result in a positive effect by decreasing incident depression (22). In our previous research, we also concluded several biological and psychosocial mechanisms through which physical exercise exerted antidepressant effects, which further supplemented to explain the antidepressant effect of LPA (23, 24).

Nutrients have been established as a relevant factor in the development of depression. A large body of systematic review pointed out that pro-inflammatory diet was linked with increased risk of depression (25–27). Plentiful meta-analyses have also been put forward on the influence of diet on depression incidence, including but not limited to diet quality or diet pattern (27–30), and most of these phenomena might be explained by the effect of diverse dietary components. For instance, fiber, omega-3 fatty acids or saturated fats included in food can affect the incidence or strength of depression by inducing neuro-inflammation and stimulating changes in the oxidative stress, neurotransmitters and hypothalamic-pituitary-adrenal (HPA) axis (31, 32). The level of Interleukin 6 (IL-6), C-reactive protein (CRP) or other pro-inflammatory biomarkers can be enhanced by pro-inflammatory diets, and further raise the risk for the development of depression (32). However, people do not consume nutrients or food in isolation. In the past few years, scholars have increasingly emphasized the importance of diet as a whole (33, 34). Using the dietary index to quantify the inflammatory potential of the entire diet may increase the robustness and effectiveness of detecting the relationship between diseases (33). Recent study has demonstrated that dietary inflammatory index (DII) is inversely associated with depression (7). One meta-analysis showed that a 1-unit increment in DII was associated with an increased risk of 6% for depressive symptoms (35). Previous research also indicated that increased risk of depression was associated with the Western diet, characterized by consumption of high red and processed meat and refined carbohydrates, and decreased risk of depression with the Mediterranean diet, represented by high amount of fruit, vegetables, and whole grains (36, 37). In special group with chronic disease or comorbidity, there were higher prevalence of depression in groups with higher DII scores (38, 39).

Further studies have tentatively explored the joint effect of physical activity and diet on depression. Two recent randomized controlled trials indicated that a combined diet and exercise intervention can significantly led to positive alteration of depression, anxiety and quality of life in patients with cancer, no matter short or long term impact (40, 41). Another integrative review also showed that psychosocial well-being interacted with diet as well as with physical activity in women with Gestational Diabetes Mellitus (GDM) (42). Owing to depression also prime larger cytokine responses to stressors and pathogens that do not appear to habituate, negative health behaviors (e.g., poor diet, a sedentary lifestyle) may act as mediating pathways that lead to further, unrestrained inflammation and depression (43). However, there are still no investigations that take into account interactions between lifestyle factors (i.e., physical activity and dietary inflammatory preference) on depression by mediation analysis. A mediator is a third variable affecting the direction and/or strength of the relationship between a predictor and outcome variable (44). This mediator can thus act as a vulnerability or protective factor. Identifying mediator allows us to target the mediator in prevention and to identify at-risk populations.

Considering that evidence about the interactive effect of DII and physical activity on depression is limited and inconsistent, and it only concentrates on the small-scale and specific population. We aimed to use the National Health and Nutrition Examination Survey (NHANES) to assess associations between LPA and depressive symptoms under different levels of DII. Our hypothesis is that inadequate LPA and higher DII scores may promote the development of depression, while anti-inflammatory diet (lower DII scores) and increased physical activity are beneficial to alleviate depression.

On the basis of the analysis of 11,078 samples in the NHANES 2007–2018 cycle, this current study for the first time reported the relationship between LPA and depressive symptoms under different levels of DII and verified the mediation effect of DII in this association. We firstly found a negative relationship between the baseline LPA status and DII level which indicated that groups with higher DII scores tended to participate in less physical activities. Subsequently, we detected a negative association between LPA and depressive symptoms while this finding statistically held only in the anti–inflammatory diet group (corresponding to a lower DII scores). Furthermore, our mediation analysis showed that the association between LPA and depression was partly mediated by DII.

Our findings on the relationships between LPA, DII and depressive symptoms supported and extended those from previous publications. As a behavioral phenotype, being physically inactive or keeping more sedentary time has been proved to be connected with low to moderate grade systemic inflammation (56, 57). One cross–sectional study carried among European teenagers assessed that a healthy diet might attenuate the positive sedentary time and inflammation relation (58). Importantly, the prevalence of these physical inactive behavior increased not only through adolescence, and once installed in this phase, there was a strong tendency it remained into adulthood (59). In adults across a wide age spectrum from 20–85 years, one previous study further reported the association between physical activity and dietary inflammatory index on mortality risk in U.S. adults (60). In this research, the author suggested that the combination of consuming a more anti–inflammatory diet and having adequate levels of physical activity can reduce the risk of mortality. Compared with the former research, our research was an improvement which further detected that groups with higher DII scores tended to be less physically active. Our findings, if demonstrated in other large–scale prospective studies, may provide information that with the increment of DII, people not only did not tend to do more LPA to rescue, but also become more physically inactive.

When it comes to depression, prolonged physical inactivity were important risk factors. Behavioral intervention such as increasing moderate–to–vigorous exercise has been detected to exhibit approximate antidepressant effects (17, 61), which was consistent with our findings that doing more LPA was correlated with lower emergence of depression. From the mechanism, 'myokines' (cytokines produced or released by muscle) played an important role, evidence showed that muscle–induced peripheral factors can induce direct crosstalk between muscle and brain function, including the effects on depressive symptoms (62–64). From the dose–effect perspective, one meta–analytic evidence from prospective cohort studies proved that doing moderate–to–vigorous physical activity around 150 min per week would have a protective effect against prevalence of depression (16). However, these evidence might be inconsistent in different dietary preference groups. One research showed that obesity is a risk factor connected with depression and could to some tent be prevented by accomplishing more LPA and low energy–dense diet (65). In accordance with the previous finding, our study also detected that in groups with anti–inflammatory diet, LPA can significantly reduce the prevalence of depression. It seemed that increasing LPA, and keeping anti–inflammatory dietary habits can subsequently prevent obesity, which might be helpful in struggling with psychosocial stress including depression.

With respect to the joint effect of LPA and dietary inflammation on depression, little is known about the distinct biological mechanisms, while inflammatory biomarkers seemed to play an important role. A growing number of publications have reported that exercise can induce considerable physiological adaptions in the immune system, cytokines (i.e., IL−6, IL−8, IL−10, and TNF–α), and stress hormones (66–68), which was beneficial to achieve similar effects with antidepressants. On the other hand, pro–inflammatory dietary behavior might involve in the modulation of pathways including oxidative stress, epigenetics and mitochondrial dysfunction (69). Recent evidence also put into the fact that higher dietary inflammation led to the leaky gut and bacterial translocation (70, 71), which may eventually promote depressive symptoms. Hence it is of great significance to examine the risk–benefit relationship between LPA and DII on depression. Our results suggested that diet inflammation, indicated by DII score, could mediated the negative association between LPA and depression, which provide convincing evidence on the hypothesized complex relationships. One latest evidence published in Nature detected that exercise–inducible metabolite can control food intake and influence systemic energy balance, which might partially support the mediation effects of dietary from the mechanism perspective (72). The combination of performing adequate level of physical activity and consuming a more anti–inflammatory diet might have positive effects on depressive symptoms, and there is necessity of clarifying the in–depth mechanisms underlying these relationships.

Yet unfortunately, no consensus has been reached that whether LPA was effective in preventing depression in individuals with higher dietary inflammation. Our statistical analysis showed that although there was a negative association between LPA and depression in the higher DII scores groups, this relationship was not significant. Given the abundant evidence supporting correlations between dietary inflammation (73) and physical exercise (74) with inflammatory biomarkers as well as the evidence for the role of inflammation in the development of depression, it led us to propose that the dose–effect relationship served as an important factor that affect the association between these lifestyle behaviors (i.e., diet and exercise patterns) with depression conditions. However, such provocative hypothesis was not entirely validated and future researches could assess these relationships in a dosage–dependent manner.

There were multiple strengths in this current study. Notably, the present study not only extended existing evidence but further highlighted the association between LPA and depression under different levels of DII, thereby further strengthening the possible application of LPA and DII in large population to prevent depression. Another key strength of our study was that it included a large sample size representing the general US population. Further, whereas previous studies have looked at the relationship of physical activity or dietary condition and depression, this current study took an innovative approach which to our knowledge was the first to examine the mediating effect of DII on the association between LPA and depression. This is important because it directed further studies and dietary interventions to consider relationship between physical exercise and depressive symptoms.

Our study had several limitations. Due to nature of this cross–sectional study design, we were temporally not able to deduce the causal inference. Secondly, depression data and LPA assessment in our study were based on self–reported questionnaire. However, PHQ9 and PAQ were regarded as “gold standard” criteria for large population–based studies, which has been applied in several previous publications (75, 76). In addition, DII score calculated by 24–h dietary recall may not only cause imprecise estimates due to inability in measuring day–to–day variability in dietary intake, but also characterize population or group intake rather than a measure of individual intake (77). Nutritional supplements should be considered as another impact factor that influence DII status. Furthermore, only BMI was reported as an indication for body composition in our analysis, and bioimpedance was considered as a more accurate and objective option. Recently, the comorbidity of obesity and depression received new attention (78–80), and the in–depth mechanisms regarding to inflammation and the gut microbiota are worthy of further exploration. To fully explore the comprehensive links between LPA, DII and depression, a longitudinal study with accurate and complete assessment of exercise and dietary behavior was needed.

In summary, our findings Indicated that (i) groups with higher DII scores tended to participate in less physical activities; (ii) performing more LPA was associated with a lower risk of depression in the anti–inflammatory diet groups; and (iii) DII was an important factor to mediate the association between LPA and depression. Our results could provide guidance for the development of prevention strategies with increased emphasis on the overall effect of physical activity status and diets inflammation for the prevalence of depression. Ultimately, findings of our study aligned with the idea that a comprehensive lifestyle management including conducting regular exercise and consuming a more anti–inflammatory diet would be beneficial to effectively reduce depression symptoms in the general population. These findings warranted further longitudinal and experimental research.

Publicly available datasets were analyzed in this study. This data can be found here: https://www.cdc.gov/nchs/nhanes/ (accessed date: 10 May 2022).

The studies involving human participants were reviewed and approved by National Center for Health Statistics (NCHS) and the NCHS Institutional Review Board (IRB). The patients/participants provided their written informed consent to participate in this study.

Conceptualization and formal analysis: YY, YC, and JY. Methodology, software, and writing—original draft preparation: YY and YC. Investigation: ZZ, KZ, JZ, and SJ. Writing—review and editing: YY, YC, JY, ZZ, and KZ. Supervision: JZ and SJ. All authors have read and agreed to the published version of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.