Depression, a common mental disorder, is characterized by persistent sadness and anhedonia in the context of previously rewarding or enjoyable activities (1). It is estimated that 5% of adults suffer from depression worldwide (2). The effectiveness of current prevention and treatment methods for depression, however, is limited and frequently associated with side effects. Interestingly, plant-based dietary patterns have been reported to greatly influence mental health, suggesting that diet has great use as an important adjunct therapy in the clinical management of depression (3).

Adherence to healthy diets is important to reduce the risk of illness throughout life. The Mediterranean diet (MED), which is rich in fruits and vegetables, whole grains, legumes, nuts, fish, and olive oil, is widely considered to beneficially impact health and longevity (4, 5). Not only has MED been incorporated into the national dietary guidelines set forth by the United States Department of Agriculture (USDA) but it has also attracted attention as a lifestyle modification with significant potential to reduce risks of menopausal metabolic syndrome, polycystic ovary syndrome, type 2 diabetes and cognitive decline (6–9). Furthermore, several studies reported that MED is associated with a decreased risk of depression (10–12).

However, there are several unaddressed issues in previous studies exploring MED and depression. Firstly, the alternate Mediterranean Diet (aMED) score, an important indicator for assessing adherence to MED, is a continuous variable that warrants further study to confirm its non-linear relationship with depression. Secondly, the etiology of depression is complicated and includes social, genetic and psychological factors (13, 14). Previous studies reported that the manifestation and severity of depression are not only affected by factors such as gender, educational level, marital status, and income but also depend on other comorbid conditions including stroke, thyroid problems and Parkinson's disease (PD) (15–20). Importantly, a single nucleotide variation in the oxoglutarate dehydrogenase-like gene rs2293239 (p.Asn725Ser) was identified as a major genetic predictor of familial depression (21). Thus, it is necessary to further analyze the influence of different factors on the relationship between MED and depression. Finally, prior reports warrant more detailed clarification of dietary habit influences on mental health in order to provide a foundation for effective lifestyle intervention guidelines (22).

In this study, we hypothesized aMED score and depressive symptom may have a non-linear relationship. We investigated associations between aMED score and depressive symptom via multivariate logistic regression, weighted generalized additive model (GAM) and two-step linear regression models, analyzing data from a large sample compiled in the National Health and Nutrition Examination Survey (NHANES) 2005–2018. Then, we identified the most important aMED score factor that contributed to depressive symptom prevalence utilizing random forest. We also examined whether the relationship between aMED score and depressive symptom differs by age, race, sex, socioeconomic variables, lifestyle- and health-related variables, and chronic medical conditions, utilizing subgroup analyses.

Over the past several decades, interest concerning the prevention and treatment of diseases via dietary modification has shifted from studying the effects of single nutrients and foods to studying dietary patterns (31). A meta-analysis of randomized controlled trials suggested that dietary interventions hold promise for effectively reducing symptoms of depression (22). Adherence to MED in particular appears to reduce manifestation of chronic illnesses and premature mortality (32). Although evidence of benefits concerning a lower prevalence of depression in the setting of proper MED adherence is scarce, some studies have reported such findings (10–12). Our results similarly suggest a negative association between aMED score and the prevalence of depressive symptom. On the basis of multivariate logistic regression analysis, we applied GAM to further evaluate the non-linear association between aMED score and the prevalence of depressive symptom. Importantly, GAM can not only address non-parametric smoothing but also fit a regression spline to data, which is obviously advantageous in analysis of non-linear associations (33). Here, we noted a negative association between aMED score and the prevalence of depressive symptom at an aMED score of 3 points or greater. These findings suggest that individuals with depressive symptoms should ideally strive for aMED scores of >3, thus increasing the likelihood that MED alleviates clinical depression.

The beneficial effects of the MED can be mainly attributed to its numerous components rich in anti-oxidants and possessing anti-inflammatory properties (34). Positive effects in the setting of depression are likely mainly due to the activity of unsaturated fatty acids and polyphenols (2018). Moreover, our findings suggest that high-fiber food, including vegetables and fruit, plays a greater role than foods such as fish, meat, alcohol, legumes, dairy and cereals, with respect to the negative association between aMED score and depressive symptom. These results demonstrate MED is a dietary pattern based on plant-derived foods (32). Importantly, vegetables and fruits are rich in phytochemicals, phenolic compounds and flavonoids. Most of these compounds possess significant anti-oxidant properties which likely improve depressive symptoms via inhibition of oxidative stress and alleviation of chronic inflammation. For example, lutein, a dietary phytochemical capable of crossing the blood-brain barrier, has been demonstrated to exert an antidepressant-like effect and alleviate neurochemical imbalances (35). Curcumin, a polyphenol extracted from the rhizome of the turmeric plant, has been demonstrated to exert antidepressant effects in both animal and human trials (36, 37). Similarly, S-equol, a metabolite of dietary soy isoflavones, was reported to alleviate depressive-like behavior in mice by inhibiting neuroinflammation and enhancing synaptic plasticity (38). Moreover, omega-3 polyunsaturated fatty acids may reduce inflammation and benefit mood by fostering the production of the anti-inflammatory prostaglandin E3 (as opposed to the pro-inflammatory prostaglandin E2) (39). Although vegetable and fruit scores were noted to exert a greater effect on the association between aMED score and the prevalence of depressive symptom, other ingredients also contribute to the relationship. As such, the anti-inflammatory and mood-related benefits of MED adherence result due to a synergistic action of nutrients composing this dietary pattern.

Due to the complex etiology of depression, there is currently no universal and effective treatment. In order to make the case for clinical application of MED in effective prevention and management of depression, we used subgroup analyses to explore the association between aMED score and the prevalence of depressive symptom in different populations. Results revealed that aMED score has a significantly negative relationship with depressive symptom in most populations, except among patients earning a high annual income or suffering diabetes, borderline blood glucose levels or PD. Previous study reported that high annual income earners generally have higher aMED scores (5). In addition, most American adults do not adhere to dietary patterns resembling MED. This phenomenon may explain the lack of a statistically significant association between aMED score and depressive symptom among high annual income earners. Furthermore, cognitive impairment is one key risk factor of depression (40). Cognitive impairment is also known to be a main complication of diabetes and PD (41, 42). This may similarly explain the lack of a statistically significant association between aMED score and depressive symptom in patients suffering diabetes or PD. Interestingly, the protective effect of MED on depressive symptoms was noted to gradually decrease as BMI increased. The prevalence of depression is known to be increased in the setting of obesity, a condition that promotes inflammation as well as resistance to insulin and leptin (43, 44). Thus, a combination of MED adherence and proper weight control is guaranteed to improve the effectiveness of MED adherence in regard to alleviating depression. No or little association of MED score with depression among younger women, however, has also been reported due to genetic factors (45). We found that age and sex do not influence the relationship between aMED score and depressive symptom. Therefore, further research is needed to confirm the aforementioned discrepancies.

There were some limitations in our study. First, due to the nature of cross-sectional design of present study, residual confounding from unmeasured confounders should be a concern. Second, dietary intake was assessed by asking subjects to recall what was eaten over the prior 2 days and thus may not accurately reflect the typical dietary patterns of subjects. Third, due to missing data, some psychiatric comorbidities were not considered as covariates in our analyses. For example, eating disorders and attention deficit hyperactivity disorder were not considered in this study (46, 47). Moreover, the state of mental health cannot be ruled out in regard to exerting a significant influence on the association between aMED score and depression as cognitive function and prescription medications were not assessed in this study. In addition, possible misclassification of dietary behavior due to memory bias may have occurred during data collection. Finally, although the PHQ-9 is a validated screening tool designed to assess depressive symptoms, it is not intended for use in the diagnosis of clinical depression. Furthermore, as diagnostic delay of depression in clinical practice is common, our findings might be caused by reversion causation; we were unable to consider this phenomenon when performing analyses due to a lack of relevant data (48). Future studies are warranted to clarify the impact of diagnostic delay on the relationship between MED and depression.

In this study, the relationship between aMED score and depressive symptom was demonstrated to be non-linear. Our findings confirm that MED alleviates depressive symptom at aMED scores of 3 or greater. However, the association between aMED score and depressive symptom in populations earning a high annual income, as well as suffering diabetes, borderline blood glucose levels or PD, requires further evaluation.

The data presented in the study are deposited in the the NHANES website, accession link: https://www.cdc.gov/nchs/nhanes/index.htm.

The studies involving human participants were reviewed and approved by the Research Ethics Committee of the National Public Health Institute. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin (https://www.cdc.gov/nchs/nhanes/irba98.htm accessed on December 9, 2021).

Conceptualization: WX and MZ. Methodology, validation, and writing—original draft preparation: YF. Software and visualization: ZD. Formal analysis, data curation, and writing—review and editing: LZ. Investigation: ML. Resources: ZH. Supervision and funding acquisition: MZ. Project administration: WX. All authors have read and agreed to the published version of the manuscript.

The present study was supported by the Traditional Chinese Medicine Bureau of Guangdong Province (20211345), the Natural Science Foundation of Guangdong Province (2022A151510450), the Shenzhen Science and Technology Innovation Committee Subject (JCYJ20210324123614040), Bao'an TCM Development Foundation (2020KJCX-KTYJ-130 and 2020KJCX-KTYJ-133), and Sanming Project of Medicine in Shenzhen (SZZYSM202106009).

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.