Table II.
Results (for studies investigating association of diet quality with mental health parameters)
| Hypothesis outcome | Effect sizea | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author, year | Diet quality tool | Depression tool | Dietary assessment | Model | Adjustment | Result | OR, HR or RR, β coefficients, or other statistics | 1 | 2 | 3 | Small | Medium | Large |
| Mental health parameter: depression | |||||||||||||
| Açik and Cakiroglu, 2019 [44] | DII | ZSRDS | 3-day food records | Multivariate logistic regression analysis | Age, smoking, alcohol, PA level, anthropometric measurements | Poor diet quality was positively associated with depression scores | OR = 2.90 (95% CI 1.51–5.98) |
X | X | ||||
| Jeffers et al., 2019 [45] | General estimating equations of dietary quality | PANAS | EMA | Generalized estimating equations | Each food item was examined as a predictor in separate models and each of the negative and positive effect was used as separate dependent variables | There was a positive association between fruits and positive affect (i). There was a positive association between sugary foods and negative affect (ii) | (i) Estimate = 1.37 (SE 0.49, P < 0.005) (ii) Estimate = 0.06 (SE 0.03, P < 0.02) |
X | X | ||||
| Faghih et al., 2020 [46] | DASH | DASS-21 | Semi-quantitative FFQ | Pearson’s correlation coefficients | Socio-economic, lifestyle and anthropometric characteristics | There was a negative correlation between diet quality and depression | Pearson’s coefficient = −0.434 (P < 0.001) | X | X | ||||
| Ramón- Arbués et al., 2019 [64] |
HEI | DASS-21 | N/A | Pearson’s correlation coefficients | Age, sex, study area, habitual residence, relationship status, height, weight, perceived economic situation, smoking, alcohol consumption, PA and sedentary lifestyle | There was no significant association between HEI and depression | N/A | X | |||||
| Attlee et al., 2022 [65] | E-DII | DASS-21 | 24-h dietary recall | Logistic regression analysis | Body habitus measures (BMI and WC), nutrient intakes and specific food groups, smoking status, PA categories | No significant association | N/A | X | |||||
| Lee et al., 2022 [63] | N/A | DASS-21 | FFQ | Linear regression | Age, gender, ethnicity, relationship status, employment, income, living arrangements, number of children, education | The likelihood of more severe depression increased with higher consumption of grain (cereal) food (i) and lower consumption of dairy products (ii) | (i) β = 1.61, 95% CI, 0.22, 3.01 (ii) β = −3.38, 95% CI, −5.39, −1.38 |
X | X | ||||
| Stanton et al., 2021 [62] | N/A | DASS-21 | Previously validated Australian FFQ | Multivariate regression analysis | Gender, age, enrolment, ethnicity, relationship status, living arrangement, work, health conditions | Intake of snack foods was associated with higher depression scores | β = 8.66, P < 0.05 | X | X | ||||
| Abramson 2017 [119] | HEI | BDI | FFQ (5 days) |
Spearman and partial correlations | Age, gender | There was no significant association between HEI and depression | N/A | X | |||||
| Quehl et al., 2017 [48] | HEI | CES-D | 3-day food records | Linear regression | Age | Diet quality was negatively associated with depression scores | β= −0.016 (95% CI −0.029 to −0.003, P = 0.017) |
X | X | ||||
| Sakai et al. 2017 [47] | DQS | CES-D | Diet history questionnaire | Multivariate analysis | BMI, current smoking, medication use, self-reported level of stress, dietary reporting status, PA, energy intake and living alone |
Diet quality was negatively associated with depression | OR for depression in highest versus lowest quintiles of diet quality was 0.65 (95 % CI 0.50–0.84, P = 0.0005) |
X | X | ||||
| Hamazaki et al., 2015 [50] | N/A | CES-D | Customary intake frequency | Multivariate logistic analysis | Age, gender, academic performance, friendships, financial matters, smoking status, consumption of alcohol, PA | Fish intake was negatively associated with depression | OR= 0.65, (95% CI 0.46–0.92) of highest versus lowest category of fish consumption | X | X | ||||
| Liu et al., 2007 [51] | N/A | CES-D | FFQ | Stepwise logistic regression | Gender, grade, city, perceived weight, smoking level and alcohol use | Risk of depression was increased with low fruit frequency and decreased with low ready to eat food, low snack food frequency and low fast food frequency. |
OR for depression was 1.62 (P < 0.0001) for low fruit frequency, frequency, 0.70 |
X | X | ||||
| BMI was not significantly associated with depression scores | (P < 0.0001) for low ready to eat food frequency, 0.73 (P < 0.05) for low snack food and 0.40 (P < 0.05) for low fast food frequency | ||||||||||||
| Peltzer and Pengpid, 2017a [52] | N/A | CES-D | FFQ | ANCOVA, descriptive statistics | Age, sex, subjective socio-economic status, country, BMI and PA | Fruit consumption was negatively associated with depression. Unhealthy dietary behaviours were positively associated with depression | Depression score was 13.28 for no fast food versus 13.70 for highest fast food consumption | X | |||||
| Peltzer and Pengpid, 2017b [53] | N/A | CES-D | FFQ | Stepwise multiple linear regression | Fruit and vegetable consumption, socio-demographic and health-related factors | Depression decreased with any increase in fruit and vegetable consumption | Strongest decrease in depression was with six servings of fruit and vegetables, b = −1.04 (P < 0.001) |
X | X | ||||
| Smith-Marek et al., 2016 [54] | N/A | CES-D | Three items taken from the Family Transitions Project survey | Path analysis | Trauma, diet and exercise | A healthier diet was positively associated with lower depression scores |
b = 2.57 (P < 0.001) |
X | X | ||||
| Breiholz, 2010 [120] | N/A | CES-D | FFQ | Independent samples t-test | Gender | There was no association between high consumption of fruits/vegetables and depression | N/A | ||||||
| El Ansari et al 2014 [55] | N/A | BDI | FFQ (12 items) |
Regression analyses | University, sex | Unhealthy food was positively correlated with depression scores (i) Fruit/vegetable intake was negatively correlated with depression scores (ii) | (i) Coefficient = 0.072 for female, 0.158 for male. (ii) Coefficient = −0.081 for female, −0.115 for male |
X | X | ||||
| Mikolajczyk et al., 2009 [59] | N/A | BDI | FFQ | Multivariable linear regression analysis | Gender and country | In females only, poor diet quality was positively associated with depression | Estimates for change in BDI per unit of food group frequency scale was −1.69 (P = 0.002), −1.62, (P = 0.003), −1.47 (P = 0.003) for less frequent consumption of fruits, vegetables and meat respectively | X | X | ||||
| Oleszko et al., 2019 [56] | N/A | BDI | FFQ (for 30 days before study) | Non-parametric Tau Kendall’s test | N/A | Diet quality was negatively associated with depression | Tau Kendall’s = −0.09 (P < 0.01) | X | X | ||||
| Rossa-Rocor et al., 2021 [49] | DSQ | PHQ-9 | One item dietary preference | Multivariate regression analysis | Age, gender, ethnicity, PA, sleep, weight satisfaction, stress, stressful life events, social support | The junk food component was positively associated with depression | β = 0.26, P < 0.001 | X | X | ||||
| Romijn, 2020 [57] | N/A | PHQ-9 | FFQ | Pearson’s correlation coefficients | Gender, ethnicity, year of study, eating disorder | Diet quality was negatively associated with depression | Pearson’s coefficient = −0.38 (P < 0.001) | X | X | ||||
| Rossa- Roccor, 2019 [60] |
N/A | PHQ-9 | Posteriori self-reported diet | Multiple linear regression | Social support, PA, stress, body image and stressful life events | The processed food diet pattern was positively associated with depression scores (z-score β = 0.21, P ≤ .001). | z-score β= 0.21 (P ≤ 0.001) | X | X | ||||
| Jaalouk et al., 2019 [121] | N/A | PHQ-9 | 73-item FFQ | Multivariable linear regression analyses | Age, sex, income, PA, BMI, family history of mental illness, alcohol consumption, stressful life events, worrying about loss of control over how much they eat, use of antidepressants | There was no association of identified dietary patterns (traditional Lebanese, Western fast food, dairy, Lebanese fast food, fruits) with depression scores | N/A | X | |||||
| Tran et al., 2017 [61] | N/A | Clinical screening | Dietary questionnaire | Multivariate logistic regression models | Age, gender, blood pressure, heart rate, BMI, presence of depressive disorder, anxiety disorder and panic attack disorder | Poor diet quality was associated with increased risk for depression | OR 1.49 (P < 0.0001) | X | X | ||||
| Wattick et al., 2018 [58] | N/A | Centre for Disease Control and Prevention’s Healthy Days Measure | Dietary questionnaire | Logistic regression | Gender, housing and food security | Fruit and vegetable intake were negatively associated with depression in males | OR 0.68 (95% CI 0.50–0.89) | X | X | ||||
| Mental health parameter: anxiety | |||||||||||||
| Author, year | Diet quality tool | Anxiety tool | Dietary assessment | Model | Adjustment | Result | OR, HR or RR, β coefficients, or other statistics | Hypothesis outcome | Effect sizea | ||||
| 1 | 2 | 3 | Small | Medium | Large | ||||||||
| Faghih et al., 2020 [46] | DASH | DASS-21 | Semi-quantitative FFQ | Pearson’s correlation coefficients | Socio-economic, lifestyle, anthropometric characteristics | Diet quality was negatively associated with anxiety scores | Pearson’s correlation coefficient = −0.325 (P < 0.001) | X | X | ||||
| Ramón- Arbués et al., 2019 [64] |
HEI | DASS-21 | N/A | Pearson’s correlation coefficients | Age, sex, study area, habitual residence, relationship status, height, weight, perceived economic situation, smoking, alcohol consumption, PA and sedentary lifestyle | Diet quality was negatively associated with anxiety scores | Pearson’s correlation coefficient = −0.10 (P < 0.01) | X | X | ||||
| Attlee et al., 2022 [65] | E-DII | DASS-21 | 24-h dietary recall | Logistic regression analysis | Body habitus measures (BMI and WC), nutrient intakes and specific food groups, smoking status, PA categories | Each point increase in the E-DII score was associated with symptoms of anxiety | OR = 1.35; 95% CI: 1.07–1.69; P = 0.01 | X | X | ||||
| Lee et al., 2022 [63] | N/A | DASS-21 | FFQ | Linear regression | Age, gender, ethnicity, relationship status, employment, income, living arrangements, number of children, education | The likelihood of more severe anxiety increased with higher consumption junk food | β = 0.62, 95% CI: 0.01, 1.22 | X | X | ||||
| Rossa-Rocor et al., 2021 [49] | DSQ | GAD-7 | One item dietary preference | Multivariate regression analysis | Age, gender, ethnicity, PA, sleep, weight satisfaction, stress, stressful life events, social support | The junk food component was positively associated with anxiety | β = 0.18, P = 0.001 | X | X | ||||
| Romijn, 2020 [57] | N/A | GAD-7 | FFQ | Pearson’s correlation coefficients | Gender, ethnicity, year of study, eating disorder | Diet quality was negatively correlated with anxiety scores | Pearson’s correlation coefficient = −0.31 (P < 0.001) | X | X | ||||
| Rossa-Roccor, 2019 [60] | N/A | GAD-7 | Posteriori self-reported dietary patterns | Multiple linear regression | Social support, PA, stress, body image and stressful life events | The processed food diet pattern was positively associated with anxiety | β = 0.14 (P ≤ 0.001) | X | X | ||||
| Wattick et al., 2018 [58] | N/A | Centre for Disease Control and Prevention Healthy Days Measure | DSQ | Logistic regression | Gender, housing and food security | Higher added sugars intake was positively associated with anxiety in females | OR = 1.18 (95% CI 1.05–1.32) | X | X | ||||
| Tran et al., 2017 [61] | N/A | Clinical screening | Questionnaire about dietary behaviour | Multi variate logistic regression models | Age, gender, blood pressure, heart rate, BMI, presence/absence of depressive disorder, anxiety disorder and panic attack disorder | There was no association between bad dietary behaviour and anxiety. | N/A | X | |||||
| Mental health parameter: stress | |||||||||||||
| Author, year | Diet quality tool | Stress tool | Dietary assessment | Model | Adjustment | Result | OR, HR or RR, β coefficients, or other statistics | Hypothesis outcome | Effect sizea | ||||
| 1 | 2 | 3 | Small | Medium | Large | ||||||||
| Faghih et al., 2020 [46] | DASH | DASS-21 | Semi-quantitative FFQ | Pearson’s correlation coefficients | Socio-economic, lifestyle, anthropometric characteristics | Diet quality was negatively correlated with stress score | Pearson’s coefficient = −0.408 (P < 0.001) | X | X | ||||
| Saharkhiz et al., 2021 [68] | DASH score | DASS-21 | FFQ | Multinomial logistic regression | Age, BMI, energy intake | Adherence to DASH style-pattern was associated with a lower stress score | OR = 0.32; 95% CI: 0.14–0.71, P = 0.009; second tertile with first DASH tertile | X | X | ||||
| Ramón- Arbués et al., 2019 [64] |
HEI | DASS-21 | N/A | Pearson’s correlation coefficients | Age, sex, study area, habitual residence, relationship status, height, weight, perceived economic situation, smoking, alcohol consumption, PA and sedentary lifestyle | Diet quality was negatively correlated with stress score | Pearson’s coefficient = −0.07 (P < 0.05) | X | X | ||||
| Attlee et al., 2022 [65] | E-DII | DASS-21 | 24 h dietary recall | Logistic regression analysis | Body habitus measures (BMI and WC), nutrient intakes and specific food groups, smoking status, PA categories | Each point increase in the E-DII score was associated with symptoms of stress. | OR = 1.41; 95% CI: 1.12–1.77; p = 0.003 | X | X | ||||
| Stanton et al., 2021 [62] | N/A | DASS-21 | Previously validated Australian FFQ | Multivariate regression analysis | Gender, age, enrolment, ethnicity, relationship status, living arrangement, work, health conditions | Intake of snack foods was associated with higher stress scores | β = 3.92, P = 0.055 | X | X | ||||
| Lee et al., 2022 [63] | N/A | DASS-21 | FFQ | Linear regression | Age, gender, ethnicity, relationship status, employment, income, living arrangements, number of children, education | The likelihood of more severe stress increased with lower consumption of dairy products | β = −1.94, 95% CI, −3.65, −1.23 | X | X | ||||
| Fabian et al., 2013 [122] | Dietary guideline adherence index | 27-item stress questionnaire | FFQ | Pearson’s chi-squared test | Age, gender, household income, school, BMI | Dietary patterns were not associated with stress levels | N/A | X | |||||
| Alfreeh et al., 2020 [67] | E-DII | PSS-10 | FFQ (Saudi) | Multiple linear regression analyses | Age, marital status, education level, course, income, financial status, sleep, PA, previous weight reduction diet | Pro-inflammatory diets were associated with increased stress. | A higher E-DII score per 1 SD (1.8) was associated with 2.4-times higher PSS score. 95% CI: 1.8, 3.1 Pearson’s partial correlation coefficient of the relationship between E-DII scores and PSS scores was (r) 0.46 |
X | X | ||||
| El Ansari et al., 2015a [66] | Dietary guideline adherence index | PSS | 12-item FFQ | Spearman rank coefficients | Age, sex, living situation, economic situation, moderate PA and BMI | Diet quality was negatively correlated to stress | Males: r = −0.21, P < 0.001 Females: r = −0.13, P < 0.001 Normal weight: r = −0.13, P < 0.001 Overweight: r = −0.21, P = 0.002 |
X | X | ||||
| El Ansari et al., 2014 [55] | N/A | PSS | 12-item FFQ | Regression analyses | University, sex | Unhealthy foods were positively correlated with stress for females (i). Fruits and vegetables were negatively correlated with stress (ii) |
(i) Coefficient = 0.051 (ii) Coefficient = −0.067 for female, −0.092 for male |
X | X | ||||
| Liu et al., 2007 [51] | N/A | PSS | FFQ | Stepwise logistic regression | Gender, grade, city, perceived weight, smoking level and alcohol use | Low fruit frequency was positively correlated with stress (i). Low ready to eat food frequency (ii) and low snack food frequency (iii) were negatively correlated with stress. There was no association between BMI and stress scores |
(i). OR = 1.53 (P < 0.01) (ii) OR = 0.69 (P < 0.01) (iii) OR = 0.75 (P < 0.05) |
X | X | ||||
| Mikolajczyk et al., 2009 [59] | N/A | PSS | 12-item FFQ | Multivariable linear regression analysis | Gender and country | In females only, consumption of sweets was positively associated with stress (i). In females only, consumption of fruits (ii) and vegetables (iii) was negatively associated with stress |
(i) Estimate = 0.54 (P = 0.04) (ii) Estimate = −1.17 (P < 0.001) (iii) Estimate = −0.82 (P = 0.003) |
X | X | ||||
| Lockhart, 2017 [123] | N/A | 5-item emotional distress scale | FFQ | Multiple linear regression | Exercise and rest | No correlation between consumption of fruits and vegetables and emotional distress | N/A | X | |||||
| Mental health parameter: general mental well-being | |||||||||||||
| Author, year | Diet quality tool | Mental well-being tool | Dietary assessment | Model | Adjustment | Result | OR, HR or RR, β coefficients or other statistics | Hypothesis outcome | Effect sizea | ||||
| 1 | 2 | 3 | Small | Medium | Large | ||||||||
| Aceijas et al., 2017 [39] | REAP-S | SWEMWBS | N/A | Multivariate analysis | Gender, lack of help-seeking behaviour in case of distress, negative attitudes towards nutrition-related activities, financial difficulties | Low diet quality almost doubled the risk of low mental well-being | OR = 1.7 (95% CI 1.0-2.7, P = 0 0.04). | X | X | ||||
| Lo Moro et al., 2021 [71] | MEDAS | WEMWBS | N/A | Linear regression analysis | Age, gender | The mental well-being and adherence to MD were positively associated | AdjB 0.676, 95% CI 0.277–1.075, P = 0.001 | X | X | ||||
| El Ansari et al., 2015b [69] | Dietary guideline adherence index | Assessment of self-reported health complaints (22 items) | 12-item FFQ | Multi nomial logistic regression model | Age group, living situation, economic situation, PA, BMI | There was a negative correlation between diet quality and psychological health complaints | Beta coefficient = 0.06 | X | X | ||||
| Hendy, 2012 [70] | Scores for total calories, carbohydrate percentage of calories, grams saturated fat and milligrams of sodium | PANAS | Anonymous 7-day record of foods | Multiple regression analyses | Restrained eating scores and gender | Consumption of calories (i), saturated fat (ii) and sodium (iii) was significantly associated with increased negative affect. There was no association for carbohydrate consumption | (i) b = 0.45 (ii) b = 0.43 (iii) b = 0.45 |
X | X | ||||
| Lopez- Olivares, 2020 [72] |
PRE-DIMED Questionnaire | PANAS | N/A | Multiple regression models | Age, sex, PA, general state of health | A strict adherence to the MD was positively associated with positive emotional state. There was no association with negative emotional state | Coefficient = 0.018 (P = 0.009) |
X | X | ||||
| Faghih et al., 2020 [46] | DASH | GHQ-12 | Validated 168-item semi-quantitative FFQ | Pearson’s correlation coefficient | Socio-economic, lifestyle, anthropometric characteristics | Diet quality was positively correlated with mental health well-being | Pearson’s correlation coefficient = −0.431, (P < 0.001) | X | X | ||||
| Mochi- masu et al., 2016 [73] |
N/A | GHQ-12 | FFQ | Multiple regression analysis | BMI, PAL, energy and sucrose | Confectionaries intake was negatively associated with mental well-being and was the determining factor for the GHQ12 scores |
b = 0.160, (P = 0.042) |
X | X | ||||
| Knowlden et al., 2016 [74] | N/A | K-6 | FFQ (24 h) |
Pearson’s correlation and Cronbach alphas | Optimism, self-esteem and social support | Frequent fruit consumption (i) and infrequent consumption of sugar-sweetened beverages (ii) was associated with low levels of mental distress. No associations with BMI |
(i) H2 = 7.268 (P = 0.026) (ii) H2 = 18.15 (P < 0.001) |
X | X | ||||
| Lesani et al., 2016 [75] | N/A | Oxford Happiness Questionnaire | FFQ | ANCOVA | BMI, marital status, socio-economic status, PA, experience of stress in the last 6 months and having a defined disease | Amount of fruit and vegetable consumption was positively associated with mental well-being | P < 0.045 for 1 versus 3 servings per day | X | |||||
| Peltzer and Pengpid, 2017a [52] | N/A | SHS | FFQ | ANCOVA | Age, sex, subjective socio-economic status, country, BMI and PA | Diet quality was positively associated with happiness and high life satisfaction | SHS score was 2.87 for no fruit consumption versus 3.03 for consuming three fruits per day | X | |||||
| Piqueras et al., 2011 [76] | N/A | SHS | FFQ | Multi variate binary logistic regression | Gender, age, perceived stress and health behaviours | Intake of fruits and vegetables intake was positively associated with happiness | Adjusted OR = 1.34 (P = 0.000) |
X | X | ||||
| Schnettler et al., 2015 [77] | N/A | SWLS | SWFL and FFQ |
Dunnett’s T3 multiple comparisons test | Sex, age, residence, socio-economic factors | Students with healthful eating habits had higher levels of life satisfaction and satisfaction with food-related life | The group ‘satisfied with their life and their food-related life’ had a higher percentage of fruit (41.7%) and vegetable (57.6%) consumption daily | X | |||||
| Rossa- Rocor 2019 [60] |
N/A | QOL Single item |
Posteriori self-reported dietary patterns | Multiple linear regression | Social support, PA, stress, body image and stressful life events | There was no association between diet preference categories and mental well-being | N/A | X | |||||
| Mental health parameter: academic stress | |||||||||||||
| Author, year | Dietary score | Academic stress tool | Dietary assessment | Model | Adjustment | Result | OR, HR or RR, β coefficients or other statistics | Hypothesis outcome | Effect sizea | ||||
| 1 | 2 | 3 | Small | Medium | Large | ||||||||
| Chacon-Cuberos et al. 2019 [78] | KIDMED | Validated Scale of Academic Stress | N/A | Regression model | Sex, BMI | MD adherence decreased stress in ‘Communication of own idea’ for high versus low MD adherence | F = 2.801 (P = 0.045) |
X | X | ||||
| Mental health parameter: self-concept | |||||||||||||
| Author, year | Dietary score | Self-concept tool | Dietary assessment | Model | Adjustment | Result | OR, HR or RR, β coefficients or other statistics | Hypothesis outcome | Effect sizea | ||||
| 1 | 2 | 3 | Small | Medium | Large | ||||||||
| Chacon-Cuberos et al., 2018 [79] | KIDMED | AF-5 | N/A | Structural Equation Model, Pearson Chi-square test) | Task and Ego Climate, Tobacco consumption, adherence to MD, PA, alcohol consumption, VO2MAX, Self-Concept, gender | MD was positively associated with self-concept |
b = 0.08, (P < 0.05 for male) b = 0.17, (P < 0.01) for female) |
X | X | ||||
| Zurita- Ortega et al., 2018 [80] |
KIDMED | AF-5 | N/A | Chi-square analysis and ANOVA | MD, PA, gender, religious belief, university campus and place of residence | Adherence to MD was positively associated with academic self-concept and physical self-concept. There were no associations for social, emotional and family self-concept |
Academic self-concept (P = 0.001) and physical self-concept (P = 0.005) were more positive with high MD adherence (M = 3.67 and M = 3.39 respectively) compared with medium adherence (M = 3.45 and M = 3.16 respectively) | X | |||||
| Mental health parameter: psychological resilience | |||||||||||||
| Author, year | Dietary score | Psychological resilience tool | Dietary assessment | Model | Adjustment | Result | OR, HR or RR, β coefficients or other statistics | Hypothesis outcome | Effect sizea | ||||
| 1 | 2 | 3 | Small | Medium | Large | ||||||||
| Lutz et al., 2017 [81] | HEI | CDRS | The Block FFQ | Logistic regression | Race, ethnicity, education, smoking, age, BMI, sex and military branch | Higher diet quality was associated with an increased likelihood of a participant being in the high-resilience group | OR 1.02 (95% CI 1.01–1.04) |
X | X | ||||
| Mental health parameter: PTSD | |||||||||||||
| Author, year | Dietary score | PTSD tool | Dietary assessment | Model | Adjustment | Result | OR, HR or RR, β coefficients or other statistics | Hypothesis outcome | Effect sizea | ||||
| 1 | 2 | 3 | Small | Medium | Large | ||||||||
| Peltzer and Pengpid, 2017a [52] | N/A | B7ISQ | Food frequency questionnaire (FFQ) | ANCOVA | Age, sex, subjective socio-economic status, country, BMI and PA | Fruit consumption were negatively associated with traumatic stress symptoms | B7ISQ scores were 19.25 for consumption of 4 or more fruits versus 19.91 for no fruit consumption |
X | |||||
| Smith- Marek et al., 2016 [54] |
N/A | PCL-5 | Three items taken from the Family Transitions Project survey | Path analysis | Trauma, diet and exercise | A healthier diet was significantly associated with lower post-traumatic stress scores | b = 1.60 (P < 0.01) |
X | X | ||||
Note: Studies ordered according to diet quality tool used; if no diet quality tool used, studies were ordered according to depression tool.
Dietary measures: Diet inflammatory score (DII), Dietary Approaches to Stop Hypertension score (DASH), Energy-adjusted Dietary Inflammatory Index (E-DII), Healthy eating index (HEI), Diet quality score (DQS), Food frequency questionnaire (FFQ), Ecological Momentary Assessment (EMA), Dietary Screener Questionnaire (DSQ), Rapid Eating and Activity Assessment for Patients-Short Version (REAP-S), PREvención con DIeta MEDiterránea questionnaire (PREDIMED), Satisfaction with Food-related Life Scale (SWFL), Test of Adherence to Mediterranean Diet (KIDMED), Mediterranean diet (MD), Physical Activity (PA), Body Mass Index (BMI).
Mental health scores: Zung Self-Rating Depression Scale (ZSRDS), Positive and Negative Affect Scale (PANAS), Depression, anxiety and stress scale (DASS-21), Beck Depression Inventory (BDI), Centre for Epidemiologic Studies Depression Scale (CES-D), Patient health questionnaire (PHQ-9), Cohen’s Perceived Stress Scale (PSS), General anxiety disorder 7 (GAD-7), Warwick–Edinburgh Mental Wellbeing Scale short version (SWEMWBS), Positive and Negative Affect Scale (PANAS), 12-item general health questionnaire (GHQ-12), Kessler-6 Psychological Distress Scale (K-6), Subjective happiness scale (SHS), Satisfaction with Life Scale (SWLS), Connor-Davidson Resilience Scale (CDRS), Breslau’s 7-item screening questionnaire (B7ISQ), Post-traumatic stress Checklist (PCL-5), Five-Factor Self-Concept Questionnaire (AF-5).
Statistics: Odds Ratio (OR), Hazard Ratio (HR), Relative Risk (RR), Confidence Interval (CI), Standard Error (SE), Analysis of covariance (ANCOVA), Between group differences (H2), Mean (M), Regression coefficient (F).
Not applicable (N/A).
Hypothesis: Good diet quality will have a beneficial effect on mental health parameters, and/or bad diet quality will have a detrimental effect on mental health parameters.
Hypothesis outcomes:
(i) Hypothesis accepted.
(ii) Hypothesis rejected—good diet quality had an adverse effect on mental health.
(iii) Hypothesis rejected—no association between diet quality and mental health.
a
If applicable.