Article | Published:


Non-alcoholic beverage consumption and risk of depression: epidemiological evidence from observational studies

European Journal of Clinical Nutritionvolume 72pages15061516 (2018) | Download Citation



Recent epidemiological studies have examined associations between various types of non-alcoholic beverage consumption and risk of depression, but the associations were inconsistent. To provide a quantitative assessment of this association, we performed a systematic review and meta-analysis of observational studies.


We searched PubMed and Web of Science databases through February 2017 for eligible studies and examined the reference lists of the retrieved articles. A random-effects model was used to calculate pooled relative risks (RR) with 95% CIs after adjusting for important confounders.


We identified fifteen observational studies (9 cross-sectional studies; 6 prospective studies) of beverage consumption and depression, including 20,572 cases of depression among 347,691 participants. For coffee and tea consumption, the pooled RRs of depression for the high vs. low categories of consumption were 0.73 (95% CI 0.59–0.90) and 0.71 (95% CI 0.55–0.91), respectively. For soft drinks, however, the pooled RR for the high vs. low category of consumption was 1.36 (95% CI 1.24–1.50). The inverse association with coffee or tea consumption and the positive association with soft drink consumption for risk of depression did not vary by gender, country, high consumption category, and adjustment factors such as alcohol, smoking and physical activity.


Our findings suggest that high consumption of coffee and tea may reduce the risk of depression, while high consumption of soft drinks may increase the risk of depression. Further well-designed large prospective studies are needed to provide definitive evidence to address the effects of various types of beverages on risk of depression.


Depression, a common mental disorder, is the leading cause of disability, which is a major contributor to the overall global burden of disease. According to the World Health Organization, at a global level, more than 300 million people live with depression. In 2015, it was estimated that 4.4% of the global population suffered from depression, which was more common in women (5.1%) than men (3.6%) [1]. Several studies have suggested that depression is associated with unhealthy lifestyles including physical inactivity [2, 3], smoking [4], alcohol drinking [5], and poor diet pattern [6, 7]. Inflammation and oxidation contribute to the pathophysiology of depression [8]. Obesity also has been associated with depression, as a determinant of depression as well as an outcome of depression [9]. Persons with depression through unhealthy lifestyles develop more obesity over time, and obesity through its negative effects on self-image leads to the development of depression over time.

Coffee, tea and sweetened beverages are the most consumed non-alcoholic beverages in the world [10]. Coffee and tea are the common sources of caffeine [11, 12]. Caffeine has been associated to generate psychostimulant effects through modulating dopaminergic transmission, which is related to depression [13]. In addition to caffeine, coffee also contains significant amounts of substances that have anti-inflammatory and antioxidant effects [14, 15]. Tea also includes other substances such as l-theanine and polyphenols that have neuro-protective activities [16,17,18]. Sweetened beverage such as soft drinks includes sugary substances. Recent studies reported that sugary substances can have adverse effects on health (e.g., increased insulin resistance, diabetes, and obesity) [19, 20], which has been associated to risk of depression [21, 22]. Several epidemiological studies have been conducted to demonstrate the associations between coffee, tea or soft drink consumption and risk of depression [23,24,25,26,27,28,29,30,31,32,33,34,35,36,37], but the associations remain unclear.

Therefore, we performed a comprehensive meta-analysis of observational studies to compare the risk of depression in subjects who consumed higher coffee, tea or soft drinks with that in those with lower consumption. Because of high prevalence of coffee, tea and soft drink consumption, even small effects on depression in persons could have a big impact on public health.

Materials and methods

Literature search and selection

The databases of PubMed and Web of Science were searched for relevant studies published through February 2017, using the following terms: “(coffee, tea, sugar-sweetened beverage, sweetened beverage, soft drink or beverage)” combined with “(depression or depressive disorder)”. We also searched the reference lists and review articles manually to get additional articles. Only those that were published in English were included. The criteria for inclusion were as follows: (1) studies that presented original data from observational human studies; (2) the exposure was the consumption of coffee/tea/soft drink (including coke, pepsi, sprite, soda, pop, or other carbonated soft drinks); (3) the outcome was depression or depressive disorder; and (4) studies reported effect estimates along with confidence intervals. A flowchart of selection process of studies is shown in Fig. 1.

Fig. 1
Fig. 1

Flowchart of the selection of studies for inclusion in the meta-analysis

Data extraction

Data were extracted independently by 2 investigators (D.K. and Y.K) based on the guidelines of the Meta-Analysis of Observational Studies in Epidemiology (MOOSE) [38]. For a study that provided the results from both longitudinal and cross-sectional analyses in participants within the same cohort [31], those from longitudinal analyses were included. We extracted the following information from the studies: last name of first author; publication year; study design; country; sex; age; number of subjects and cases; category of beverages; adjustment factors; and odds ratios (ORs)/relative risks (RRs) and 95% CIs that had the greatest adjustment for confounders. For a study that provided β values and its 95% CIs of the association between coffee consumption and depression [37], the ORs were recalculated using the exponential function.

Statistical analysis

Using DerSimonian and Laird random-effects models [39], we estimated pooled RRs of depression. For statistical heterogeneity among the studies, we used the Cochran Q statistic [40], and I2 statistic [41]. To investigate the variations in risk estimates among studies, we conducted subgroup analyses by study characteristics, and tested the variations through the meta-regression analyses. We also conducted stratified analyses by median cut-offs for the highest category of coffee (~3 cups/d), tea (~1 cup/d), and soft drinks (~1 cup/d). In addition, we investigated whether the studies had controlled for important confounders such as alcohol, smoking, physical activity, and body mass index (BMI) that might influence the relationship between beverage consumption and depression. Furthermore, to evaluate the robustness of the pooled RR, we conducted a sensitivity analysis eliminating one study at a time. Finally, publication bias was evaluated using the Begg’s [42] and Egger’s tests [43]. We considered a two-tailed p < 0.05 to be statistically significant. All statistical analyses were performed by using Stata/SE version 14.0 software (Stata Corp LP, College Station, TX, USA).


Study characteristics

A total of fifteen studies with 347,691 participants and 20,572 cases were identified [23,24,25,26,27,28,29,30,31,32,33,34,35,36,37]. Table 1 shows the characteristics of individual studies. By study design, there were 9 cross-sectional (n = 6028 cases) [23,24,25, 30,31,32,33, 35,36,37] and 6 prospective studies (n = 14,544 cases) [26,27,28,29, 31, 34] that examined risk of depression in relation to beverage consumption. Of the fifteen studies, there were 8 studies for coffee [23, 24, 27, 28, 30, 34, 36, 37], 10 for tea [23, 24, 26, 29,30,31,32, 34, 36], and 5 for soft drinks [25, 33,34,35,36]. By country, 2 studies were conducted in Finland (n = 259 cases) [23, 27], 2 in the USA (n = 13,918 cases) [28, 34], 1 in Australia (n = 327 cases) [25], and 10 in Asia (n = 6068 cases) including 1 from Korea [36], 4 from China [26, 29, 32, 35], 3 from Japan [24, 30, 37], 1 from Iran [33] and 1 from Taiwan [31]. By gender, 12 studies [23,24,25, 29,30,31,32,33,34,35,36,37] included both men and women, and among the remaining 3 studies, one study included men only [27], and two studies included women only [26, 28]. Many studies controlled for important confounders such as alcohol drinking [23,24,25, 27, 30, 34,35,36], smoking [23,24,25, 27, 28, 30, 34,35,36], physical activity [24,25,26, 28,29,30, 32,33,34,35,36,37], and BMI [24, 25, 27, 28, 30, 33,34,35,36]. For dietary assessment, most of studies used food frequency questionnaires [23, 24, 26, 28,29,30,31,32,33,34,35,36,37], one study used a 4-day food record [27], and one study used a computer-assisted telephone interviewing system [25]. Depression was assessed in various ways as follows: self-reported physician diagnosis [25, 34, 36], self-reported physician diagnosis and antidepressant use [28], diagnosis by a physician with International Classification of Diseases (ICD) criteria [27], Global School-based student Health Survey questionnaires [33], the hospital anxiety and depression scale (HADS) [37], the Beck Depression Inventory (BDI) [23], the Geriatric Depression Scale (GDS) [24, 29, 32], the Center for Epidemiological Studies Depression Scale (CES-D) [26, 30, 31], and the Zung Self-Rating Depression Scale (SDS) [35].

Table 1 Characteristics of observational studies included in the meta-analysis of the association between coffee, tea and soft drink consumption and risk of depression

Coffee consumption and depression

Among eight studies for coffee consumption, 3 studies had a prospective design and 5 studies had a cross-sectional design. The multivariable study-specific RRs along with a pooled RR are provided in Fig. 2. The pooled RR for high vs. low coffee intake was 0.73 (95% CI: 0.59–0.90) with some evidence of heterogeneity (P = 0.006, I2 = 65.0%). When the analysis was stratified by study design, however, the significant heterogeneity decreased. For cross-sectional studies, the pooled RR was 0.62 (95% CI: 0.42–0.92) (P = 0.06, I2 = 56.9%), and for prospective studies, the pooled RR was 0.85 (95% CI 0.70–1.03) (P = 0.09, I2 = 58.8%). A sensitivity analysis showed that the study by Lucas et al. [28] had the most influence on the pooled RR, resulting in 0.65 (95% CI 0.47–0.91). After excluding the study by Omagari et al. [37] including diabetic patients as study subjects that provided the RR that had the least adjustment in the model, the pooled RR was 0.76 (95% CI 0.62–0.92), which was very close to that from the main analysis. The subgroup analyses showed no significant difference in results by the subgroups (Table 2). By country, pooled RRs were 0.55 (95% CI 0.34–0.88) in Asia and 0.55 (95% CI 0.16–1.86) in Finland, while a pooled RR in the USA was 0.89 (95% CI 0.83–0.97) (P for difference in Asia vs. USA = 0.14). By gender, the pooled RRs in men and women were 0.61 (95% CI 0.37–1.01) and 0.83 (95% CI 0.72–0.97), respectively. Comparing the results using 3–4 cups/day with those using 1–2 cups/day as a cut-off for the highest category, the pooled RRs were similar (P for difference = 0.60). The pooled RR from studies that had controlled for important confounders was 0.70 (95% CI 0.50–0.98), which was similar to the overall RR.

Fig. 2
Fig. 2

Forest plot of the observational studies of depression for high vs. low coffee consumption

Table 2 Summary of pooled relative risks (RR) of depression for high vs. low beverage consumption by study characteristics

Tea consumption and depression

Among ten studies for tea consumption, 5 studies had a prospective design and 5 studies had a cross-sectional design. The pooled RR was 0.71 (95% CI 0.55–0.91) with significant heterogeneity (P < .001, I2 = 77.7%) (Fig. 3). By study design, the inverse association was stronger among cross-sectional studies (pooled RR = 0.63, 95% CI 0.53–0.74), while the association was attenuated among prospective studies (pooled RR = 0.90, 95% CI 0.65–1.25) (P for difference = 0.05). After stratifying by study design, the significant heterogeneity disappeared among the cross-sectional studies (P = 0.39, I2 = 3.8%), but it remained among prospective studies (P = 0.01, I2 = 68.7%). The sensitivity analysis showed the pooled RRs ranging from 0.67 (95% CI: 0.51–0.87) excluding Ruusunen et al. [27] to 0.74 (95% CI 0.57–0.96) excluding Hintikka et al. [23] When the study by Chen et al. [26] including breast cancer survivors as study subjects was eliminated, the pooled RR was 0.72 (95% CI 0.55–0.93). When we excluded the prospective study by Feng et al. [29] conducted in Asia that showed a strong inverse association, the heterogeneity observed among prospective studies decreased (P = 0.07, I2 = 57.4%). The inverse association was stronger in Asian studies (pooled RR = 0.66, 95% CI 0.55–0.79) than the study from the USA (P for difference = 0.09) (Table 2). For type of tea, the pooled RR of studies that provided effect estimates of green tea consumption was 0.67 (95% CI 0.57–0.80), and the other tea such as black tea or oolong tea showed a similar RR (pooled RR = 0.67, 95% CI 0.47–0.91). For gender, high consumption category and adjustment factors, there was no significant difference as well.

Fig. 3
Fig. 3

Forest plot of the observational studies of depression for high vs. low tea consumption

Soft drink consumption and depression

Among five studies for soft drinks, 4 studies had a cross-sectional design and one study had a prospective design. The pooled RR was 1.36 (95% CI 1.24–1.50) with no significant heterogeneity (P = 0.33, I2 = 13.0%) (Fig. 4). When the analysis was limited to cross-sectional studies, the positive association was slightly stronger (pooled RR = 1.44, 95% CI 1.25–1.67). Based on the sensitivity analysis, the pooled RRs ranged from 1.34 (95% CI 1.24–1.45) excluding Yu et al. [35] to 1.44 (95% CI 1.25–1.67) excluding Guo et al. [34]. By country, a study from Australia showed a higher RR (RR = 1.63, 95% CI 1.03–2.58) than a RR of 1.30 (95% CI 1.17–1.44) from the USA or a pooled RR of 1.44 (95% CI 1.10–1.88) from Asia, but there was no significant difference by the regions (Table 2). Out of five studies, 4 studies had adjusted for important confounders and the pooled RR of the studies was 1.39 (95% CI 1.13–1.71), which was very similar to the pooled estimate of all studies included. No significant difference in results by high consumption category was found.

Fig. 4
Fig. 4

Forest plot of the observational studies of depression for high vs. low soft drink consumption

Publication bias

For the analyses of coffee and tea consumption, the Egger test suggested some evidence of publication bias (coffee: P = 0.01; tea: P = 0.03), but the evidence of bias was not shown in the Begg’s test (coffee: P = 0.06; tea: P = 0.59). The difference in the results from the tests may be due to a greater statistical power of the Egger regression method [43]. On the other hand, for soft drink consumption, there was no evidence of publication bias in both the Begg’s (P = 0.46) and Egger’s tests (P = 0.56)


To evaluate the associations of beverage consumption with risk of depression quantitatively, we conducted a comprehensive analysis of fifteen observational studies, including 20,572 cases among 347,691 participants. The results of the meta-analysis indicated that high coffee or tea consumption is inversely associated with depression, while high soft drink consumption is associated with increased risk of depression. People consuming high coffee (≥3cups/day, median) or high tea (≥1cup/day, median) had 27 and 30% lower risk of depression, respectively, compared to those consuming low/no coffee or tea. People consuming high soft drinks (≥1cup/day, median) had 1.36 times greater risk of depression than those consuming low/no soft drink. The decreased risk of depression by coffee or tea consumption and increased risk of depression by soft drinks did not vary by gender, country, high intake category or adjustment factors, substantially.

The studies related to soft drinks had no significant heterogeneity, and some evidence of heterogeneity observed in studies for coffee consumption decreased after stratified by study design. The observed heterogeneity among prospective studies of tea consumption seemed to be explained by the prospective study by Feng et al. [29] that showed the strongest inverse association of tea consumption with depression (RR = 0.30, 95% CI 0.11–0.83). The study [29] was conducted in China where people consumed relatively high amount of tea, and showed the highest level of tea consumption category (≥6 cups/day) among the studies. After excluding the Chinese study, the heterogeneity observed among the prospective studies decreased. Although no significant difference in results was found by country, the associations of coffee or tea consumption with risk of depression tended to be stronger in Asian studies compared to those from the USA. Out of fifteen observational studies in our study, two studies used patients such as breast cancer survivors [26] and diabetic patients [37] as study subjects, which showed a non-significant and significant inverse association for tea and coffee consumption, respectively [37]. The results from the sensitivity analyses by excluding the two studies remained similar to those from the main analysis.

A few meta-analyses about coffee and tea consumption and risk of depression have been published [44,45,46], which suggested that coffee or tea may decrease the risk of depression. A study of tea and depression combined all of the duplicate data from the same cohort study [44] and a study of coffee and depression did not conduct subgroup analyses by study design, although the study provided design-specific estimates for caffeine consumption [45]. The study of coffee/tea consumption and depression also performed a dose-response analysis by combing the data including cross-sectional studies [46]. Since we had limited data for prospective studies and it is not appropriate to conduct a dose-response meta-analysis by including cross-sectional data, we did not conduct the dose-response analysis. As a way of evaluating the influence of different levels of beverage consumption as a cut-off for the highest category, a stratified meta-analysis by the median cut-offs was conducted. For soft drink consumption and depression, the association has never been investigated in a systematic manner, previously. We included five studies in the meta-analysis of soft drink consumption and depression, and the 4 out of 5 studies suggested increased risks of depression among subjects with high consumption of soft drinks, and one study conducted on Korean adults showed a non-significant association [36].

The mechanisms of the inverse association of coffee and tea consumption and the positive association of soft drink consumption for risk of depression are unclear, but some biological explanations have been suggested. The inverse associations of coffee and tea with depression might be mediated by the caffeine content in the beverages. Caffeine stimulates the central nervous system as a nonspecific adenosine receptor antagonist and enhances dopaminergic neurotransmission [13, 47], which may counteract the depressed status. Besides caffeine, coffee also includes significant amounts of substances of polyphenolic compounds such as chlorogenic or caffeic acids that have anti-inflammatory and antioxidant effects [14, 15]. Increased inflammation and oxidative stress dysregulation have been related to pathogenesis of depression [48, 49]. Tea is also a rich source of polyphenols, catechins. The epigallocatechin gallate, the most potent polyphenol in tea, has an anti-oxidative property [16]. Other than the polyphenols in tea, theanine can increase the brain serotonin or dopamine, which have neuroprotective effects [17, 18]. It was reported that the administration of theanine improved the behavioral depression resulting from chronic psychosocial stress in mice [50], and was shown to have an antidepressant effect in humans as well [17]. For the association of soft drink consumption with depression, the sugar content in soft drinks may be a possible explanation. A large amount of sugar substances in soft drinks could result in having a high glycemic load (GL) [51]. Although it was reported that the consumption of sugar relieved some psychological symptoms immediately [52], it seems not applicable in the long term. The consumption of high-GL diets was related to high levels of C-reactive protein, a marker of low-grade inflammation [53], while the consumption of low-GL diets was associated with low levels of pro-inflammatory cytokines [51]. A study showed that hyperglycemia from chronic cola consumption induced a noticeable increase in triglycerides and oxidative stress possibly related to major depression [54]. Although the positive association of soft drink consumption and depression might be mostly due to high consumption of sugary substances, we cannot exclude the possibility that components in soft drinks other than sugar may influence the depression as well.

Our study has a few strengths. This is the most comprehensive study including the most recent data to assess the associations of various types of beverage and risk of depression, systematically. A relatively large number of cases and study participants were included in the meta-analyses, and thus we could perform subgroup analyses by several study characteristics. The results of our sensitivity analyses supported that the main findings for the consumption of individual beverages and depression were robust.

The current study also has some potential limitations. First, our study was based on observational studies and thus there was a possibility of unknown confounding that may have influenced the effect estimates from original studies as well as pooled estimates. Most of the studies had controlled for potential confounders and the stratified analyses by the adjustment factors showed no significant difference. Nonetheless, it is possible that the unmeasured confounders (i.e., psychosocial factors) may have affected, in part, the inverse associations of coffee and tea consumption and depression. Second, although the subgroup analysis by gender was conducted for coffee and tea, the analysis by gender for soft drinks was not conducted due to lack of data. Out of 5 studies for soft drinks, only one cohort study provided sex-specific estimates [34]. The RR of 4 cans of soft drink or more per day vs. none was 1.30 (95% CI 1.37–1.44), and showed similar results in men (RR = 1.32, 95% CI 1.13–1.53) and women (RR = 1.25, 95% CI 1.07–1.45). Third, the cut-offs for the highest categories varied among the studies, but there was no significant difference in pooled RRs stratified by the median cut-offs. Fourth, due to the limited data which had a prospective design, we could not conduct a dose-response meta-analysis. For coffee and tea, the studies that used higher cut-offs for the highest category tended to show slightly stronger inverse associations than those with lower cut-offs, although there was no significant difference by the highest category. Fifth, the association of beverage consumption with depression may vary by study characteristics. Although we found no significant difference in the effect estimates by age, gender, and country through the meta-regression analyses, there might be some interactions between beverage consumption and lifestyle factors such as smoking, alcohol drinking, medication, and dietary factors. Since data on lifestyle-specific RRs were not available from individual studies, we could not assess the interactions in the current study. Finally, no publication bias was indicated for soft drinks, but not for coffee and tea based on more powerful Egger’s test.

In conclusion, our study showed the quantitative evidence that high coffee and tea consumption may lower the risk of depression, while high soft drink consumption may increase the risk of depression. Indeed, as this meta-analysis was based on the observational studies, it is hard to conclude the causality of this association. Moreover, the number of prospective studies for this subject was still insufficient. Further well-designed large prospective studies, especially for soft drinks, are warranted to address the effects of various types of beverages on risk of depression.


  1. 1.

    THE GLOBAL BURDEN OF DISEASE 2004 UPDATE. Accessed on 10 Mar 2017.

  2. 2.

    Krogh J, Nordentoft M, Sterne JA, Lawlor DA. The effect of exercise in clinically depressed adults: systematic review and meta-analysis of randomized controlled trials. J Clin Psychiatry. 2011;72:529–38.

  3. 3.

    Zhai L, Zhang Y, Zhang D. Sedentary behaviour and the risk of depression: a meta-analysis. Br J Sports Med. 2015;49:705–9.

  4. 4.

    Breslau N, Peterson EL, Schultz LR, Chilcoat HD, Andreski P. Major depression and stages of smoking. A longitudinal investigation. Arch Gen Psychiatry. 1998;55:161–6.

  5. 5.

    Wang PW, Lin HC, Yeh YC, Liu TL, Yen CF. The relation of substance use with different levels of depressive symptoms and the moderating effect of sex and age in Taiwanese adolescents. Compr Psychiatry. 2012;53:1013–20.

  6. 6.

    Li Y, Lv MR, Wei YJ, Sun L, Zhang JX, Zhang HG, et al. Dietary patterns and depression risk: a meta-analysis. Psychiatry Res. 2017;253:373–82.

  7. 7.

    Lopez-Garcia E, Schulze MB, Fung TT, Meigs JB, Rifai N, Manson JE, et al. Major dietary patterns are related to plasma concentrations of markers of inflammation and endothelial dysfunction. Am J Clin Nutr. 2004;80:1029–35.

  8. 8.

    Behr GA, Moreira JC, Frey BN. Preclinical and clinical evidence of antioxidant effects of antidepressant agents: implications for the pathophysiology of major depressive disorder. Oxid Med Cell Longev. 2012;2012:1–13.

  9. 9.

    Luppino FS, de Wit LM, Bouvy PF, Stijnen T, Cuijpers P, Penninx BW, et al. Overweight, obesity, and depression: a systematic review and meta-analysis of longitudinal studies. Arch Gen Psychiatry. 2010;67:220–9.

  10. 10.

    Popkin BM, Armstrong LE, Bray GM, Caballero B, Frei B, Willett WC. A new proposed guidance system for beverage consumption in the United States. Am J Clin Nutr. 2006;83:529–42.

  11. 11.

    Frary CD, Johnson RK, Wang MQ. Food sources and intakes of caffeine in the diets of persons in the United States. J Am Diet Assoc. 2005;105:110–3.

  12. 12.

    Lara DR. Caffeine, mental health, and psychiatric disorders. J Alzheimers Dis. 2010;20:S239–248.

  13. 13.

    Kaasinen V, Aalto S, Nagren K, Rinne JO. Dopaminergic effects of caffeine in the human striatum and thalamus. Neuroreport. 2014;15:281–5.

  14. 14.

    Dorea JG, Da Costa TH. Is coffee a functional food? Br J Nutr. 2005;93:773–82.

  15. 15.

    Dos Santos MD, Almeida MC, Lopes NP, de Souza GE. Evaluation of the antiinflammatory, analgesic and antipyretic activities of the natural polyphenol chlorogenic acid. Biol Pharm Bull. 2006;29:2236–40.

  16. 16.

    Zhu WL, Shi HS, Wei YM, Wang SJ, Sun CY, Ding ZB, et al. Green tea polyphenols produce antidepressant-like effects in adult mice. Pharmacol Res. 2012;65:74–80.

  17. 17.

    Kimura K, Ozeki M, Juneja LR, Ohira H. L-Theanine reduces psychological and physiological stress responses. Biol Psychol. 2007;74:39–45.

  18. 18.

    Nagasawa K, Aoki H, Yasuda E, Nagai K, Shimohama S, Fujimoto S. Possible involvement of group I mGluRs in neuroprotective effect of theanine. Biochem Biophys Res Commun. 2004;320:116–22.

  19. 19.

    Whitehouse CR, Boullata J, McCauley LA. The potential toxicity of artificial sweeteners. AAOHN J. 2008;56:251–9.

  20. 20.

    Humphries P, Pretorius E, Naude H. Direct and indirect cellular effects of aspartame on the brain. Eur J Clin Nutr. 2008;62:451–62.

  21. 21.

    Timonen M, Laakso M, Jokelainen J, Laakso M, Härkönen P, Koskela P, et al. Insulin resistance and depression: cross sectional study. BMJ. 2005;330:17–18.

  22. 22.

    Preiss K, Brennan L, Clarke D. A systematic review of variables associated with the relationship between obesity and depression. Obes Rev. 2013;14:906–18.

  23. 23.

    Hintikka J, Tolmunen T, Honkalampi K, Haatainen K, Koivumaa-Honkanen H, Tanskanen A, et al. Daily tea drinking is associated with a low level of depressive symptoms in the Finnish general population. Eur J Epidemiol. 2005;20:359–63.

  24. 24.

    Niu K, Hozawa A, Kuriyama S, Ebihara S, Guo H, Nakaya N, et al. Green tea consumption is associated with depressive symptoms in the elderly. Am J Clin Nutr. 2009;90:1615–22.

  25. 25.

    Shi Z, Taylor AW, Wittert G, Goldney R, Gill TK. Soft drink consumption and mental health problems among adults in Australia. Public Health Nutr. 2010;13:1073–9.

  26. 26.

    Chen X, Lu W, Zheng Y, Gu K, Chen Z, Zheng W, et al. Exercise, tea consumption, and depression among breast cancer survivors. J Clin Oncol. 2010;28:991–8.

  27. 27.

    Ruusunen A, Lehto SM, Tolmunen T, Mursu J, Kaplan GA, Voutilainen S. Coffee, tea and caffeine intake and the risk of severe depression in middle-aged Finnish men: the Kuopio Ischaemic Heart Disease Risk Factor Study. Public Health Nutr. 2010;13:1215–20.

  28. 28.

    Lucas M, Mirzaei F, Pan A, Okereke OI, Willett WC, O’Reilly EJ, et al. Coffee, caffeine, and risk of depression among women. Arch Intern Med. 2011;171:1571–8.

  29. 29.

    Feng L, Li J, Kua EH, Lee TS, Yap KB, John Rush A, et al. Association between tea consumption and depressive symptoms in older Chinese adults. J Am Geriatr Soc. 2012;60:2358–60.

  30. 30.

    Pham NM, Nanri A, Kurotani K, Kuwahara K, Kume A, Sato M, et al. Green tea and coffee consumption is inversely associated with depressive symptoms in a Japanese working population. Public Health Nutr. 2014;17:625–33.

  31. 31.

    Tsai AC, Chi SH, Wang JY. Cross-sectional and longitudinal associations of lifestyle factors with depressive symptoms in ≥ 53-year old Taiwanese—results of an 8-year cohort study. Prev Med. 2013;57:92–97.

  32. 32.

    Feng L, Yan Z, Sun B, Cai C, Jiang H, Kua EH, et al. Tea consumption and depressive symptoms in older people in rural China. J Am Geriatr Soc. 2013;61:1943–7.

  33. 33.

    Zahedi H, Kelishadi R, Heshmat R, Motlagh ME, Ranjbar SH, Ardalan G, et al. Association between junk food consumption and mental health in a national sample of Iranian children and adolescents: the CASPIAN-IV study. Nutrition. 2014;30:1391–7.

  34. 34.

    Guo X, Park Y, Freedman ND, Sinha R, Hollenbeck AR, Blair A, et al. Sweetened beverages, coffee, and tea and depression risk among older US adults. PLoS One. 2014;9:e94715.

  35. 35.

    Yu B, He H, Zhang Q, Wu H, Du H, Liu L, et al. Soft drink consumption is associated with depressive symptoms among adults in China. J Affect Disord. 2015;172:422–7.

  36. 36.

    Park RJ, Moon JD. Coffee and depression in Korea: the fifth Korean National Health and Nutrition Examination Survey. Eur J Clin Nutr. 2015;69:501–4.

  37. 37.

    Omagari K, Sakaki M, Tsujimoto Y, Shiogama Y, Iwanaga A, Ishimoto M, et al. Coffee consumption is inversely associated with depressive status in Japanese patients with type 2 diabetes. J Clin Biochem Nutr. 2014;55:135–42.

  38. 38.

    Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA. 2000;283:2008–12.

  39. 39.

    DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177–88.

  40. 40.

    Cochran WG. The combination of estimates from different experiments. Biometrics. 1954;10:101–29.

  41. 41.

    Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327:557–60.

  42. 42.

    Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994;50:1088–101.

  43. 43.

    Sterne JAC, Gavaghan D, Egger M. Publication and related bias in meta-analysis: power of statistical tests and prevalence in the literature. J Clin Epidemiol. 2000;53:1119–29.

  44. 44.

    Dong X, Yang C, Cao S, Gan Y, Sun H, Gong Y, et al. Tea consumption and the risk of depression: a meta-analysis of observational studies. Aust N Z J Psychiatry. 2015;49:334–45.

  45. 45.

    Wang L, Shen X, Wu Y, Zhang D. Coffee and caffeine consumption and depression: a meta-analysis of observational studies. Aust N Z J Psychiatry. 2016;50:228–42.

  46. 46.

    Grosso G, Micek A, Castellano S, Pajak A, Galvano F. Coffee, tea, caffeine and risk of depression: a systematic review and dose-response meta-analysis of observational studies. Mol Nutr Food Res. 2016;60:223–34.

  47. 47.

    Kaasinen V, Aalto S, Nagren K, Rinne JO. Dopaminergic effects of caffeine in the human striatum and thalamus. Neuroreport. 2004;15:281–5.

  48. 48.

    McNally L, Bhagwagar Z, Hannestad J. Inflammation, glutamate, and glia in depression: a literature review. CNS Spectr. 2008;13:501–10.

  49. 49.

    Liu T, Zhong S, Liao X, Chen J, He T, Lai S, et al. A meta-analysis of oxidative stress markers in depression. PLoS One. 2015;10:e0138904.

  50. 50.

    Unno K, Fujitani K, Takamori N, Takabayashi F, Maeda K, Miyazaki H, et al. Theanine intake improves the shortened lifespan, cognitive dysfunction and behavioural depression that are induced by chronic psychosocial stress in mice. Free Radic Res. 2011;45:966–74.

  51. 51.

    Levitan EB, Cook NR, Stampfer MJ, Ridker PM, Rexrode KM, Buring JE, et al. Dietary glycemic index, dietary glycemic load, blood lipids, and C-reactive protein. Metabolism. 2008;57:437–43.

  52. 52.

    Murakami K, Ssaki S, Takahashi Y, Uenishi K, Watanabe T, Kohri T, et al. Dietary glycemic index is associated with decreased premenstrual symptoms in young Japanese women. Nutrition. 2008;24:554–61.

  53. 53.

    Liu S, Manson JE, Buring JE, Stampfer MJ, Willett WC, Ridker PM. Relation between a diet with a high glycemic load and plasma concentrations of high-sensitivity C-reactive protein in middle-aged women. Am J Clin Nutr. 2002;75:492–8.

  54. 54.

    Bilici M, Efe H, Koroglu MA, Uydu HA, Bekaroğlu M, Değer O. Antioxidative enzyme activities and lipid peroxidation in major depression: alterations by antidepressant treatments. J Affect Disord. 2001;64:43–51.

Download references


This work was supported by the Basic Science Research Program through the National Research Foundation of Korea, funded by the Ministry of Science, ICT and Future Planning (NRF-2015R1A1A1A05001362). Funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Author information


  1. Department of Food and Nutrition, Kyung Hee University, Seoul, South Korea

    • Dami Kang
    • , Youngyo Kim
    •  & Youjin Je


  1. Search for Dami Kang in:

  2. Search for Youngyo Kim in:

  3. Search for Youjin Je in:

Conflict of interest

The authors declare that they have no conflict of interest.

Corresponding author

Correspondence to Youjin Je.

About this article

Publication history






Further reading