Dietary Approaches to Stop Hypertension (DASH) diet and mental well-being: a systematic review

Abstract

Context

The Dietary Approaches to Stop Hypertension (DASH) diet is one of the most commonly prescribed diets for preventing and controlling hypertension. However, the relationship between the DASH diet and mental health and well-being has yet to be systematically understood.

Objective

To fill this gap, the present review systematically examined the current literature on the association between the DASH diet and mental health.

Data Sources

The Web of Science, PsycINFO, and PubMed databases were systematically searched to identify eligible publications up to May 2021. Interventional and observational studies published in English allowing for inferences about mental well-being were included.

Data Extraction

Two authors independently assessed the quality of reviewed studies using the NIH quality assessment tool and extracted qualitative data. Conflicts were adjudicated by the senior author.

Data Analysis

A total of 16 studies involving 48 824 participants were included in the final review: Ten were observational studies and 6 were randomized controlled trials. On average, the methodological quality of the studies was medium strength. Mixed results on psychological measures were reported, but in general, the DASH diet was associated with better mental well-being. Five observational studies supported a negative association between adherence to the DASH diet and depressive symptoms/depression. Four randomized controlled trials provided evidence of the beneficial effects of the DASH diet on mental health, including quality of life and emotional symptoms.

Conclusion

The DASH diet likely has positive effects on mental well-being, but the results were inconsistent across different studies, which might be likely due to differences in methods of assessments of the DASH diet and mental health outcomes. Well-powered randomized controlled trials with mental well-being as the primary outcome are needed in the future.

Systematic Review Registration

PROSPERO registration no. CRD42021267667.

DASH diet, dietary approaches to slow hypertension, dietary index, epidemiology, mental health psychological well-being, systematic review

INTRODUCTION

The Dietary Approaches to Stop Hypertension (DASH) diet is a dietary pattern promoted to prevent and control hypertension and has been shown to lower blood pressure substantially.¹ It focuses on 8 dietary components: a high intake of fruits, vegetables, low-fat dairy, nuts and legumes, and whole grains, as well as a low intake of sodium, sweetened beverages, and red and processed meats. In addition to lowering blood pressure, the DASH diet has been shown to have beneficial effects on cardiometabolic health, chronic kidney disease, and oxidative stress.^2–4 Given its wide range of benefits on health, the DASH diet is documented as 1 of 3 healthy diets recommended in the 2020–2025 US Dietary Guidelines.⁵ It is also one of the most prescribed diets for reducing blood pressure and diseases related to hypertension and is an important behavioral treatment, given that cardiovascular disease is a leading cause of premature morbidity and mortality across the world.^6,7

While substantial evidence has supported the case for a beneficial role of the DASH diet on physical health, limited studies have investigated the psychological consequences of the DASH diet. This is an important knowledge gap to fill because of the strong connection between diet and psychological health and well-being, as being evidenced by the rapidly emerging field of nutritional psychiatry, an approach aimed at preventing and reducing symptoms of mental impairments through diets.⁸ Moreover, both cross-sectional and longitudinal studies have found that people who eat more processed foods are more likely to develop mental illness, while people who consume more healthy diets with balanced nutrition are less likely to develop such illnesses.⁹ A recent systematic review presented further empirical evidence of consistently significant and positive effects of dietary interventions in reducing symptoms of depression and anxiety.¹⁰ Also, in a cross-sectional baseline analysis of a large-scale Spanish primary prevention trial, Predimed-Plus, researchers found a positive association between adherence to the Mediterranean diet and mental and emotional functioning.¹¹

An increasing number of studies suggest that the DASH diet may similarly have potential psychological benefits. For example, existing research on the DASH diet has suggested that the diet is associated with reduced depression, anxiety, and aggressive behavior symptoms, although some results were mixed.^12,13 To the best of our current knowledge, there has been no systematic review on this topic to date. Thus, intervention and observational studies were systematically reviewed in this study to understand the potential impact of the DASH diet on mental well-being. Mental well-being was defined as any outcome encompassing emotion and mood, quality of life, or mental disorder symptomatology. This systematic review was pre-registered in PROSPERO (CRD42021267667) with a pre-specified review protocol.

METHODS

Search strategy

The present study was performed according to the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) statement. A complete list of PRISMA checklist items with corresponding page numbers is given in Appendix S1 in the Supporting Information online. From May 2021 through July 2021, a comprehensive search for studies from the PubMed, PsycINFO, and Web of Science databases was conducted. The databases were searched systematically up to May 2021. Search terms included (“DASH diet” OR “dietary approaches to stop hypertension”) AND (“depression” OR “mental health” OR “psychology” OR “psychosocial” OR “emotion” OR “mood” OR “quality of life” OR “anxiety” OR “mental” OR “body image” OR “affect” OR “eating disorders” OR “binge eating” OR “anorexia” OR “bulimia” or “eating disorders not otherwise specified” OR “EDNOS”). The reference lists of included studies were also checked to avoid missing any relevant publications.

Inclusion criteria

The output of the initial search was screened by 2 authors independently (J.T. and C.W.) to identify potentially eligible studies. Titles and abstracts of all retrieved articles were reviewed for eligibility. Full texts of articles were reviewed if titles and abstracts were not sufficient to determine whether the articles were relevant. Only English articles published in peer-reviewed journals were included. Inclusionary studies met the following criteria: (1) had an intervention group of the DASH diet or a score measuring the adherence to the DASH diet, (2) involved measurement of at least 1 relevant mental well-being outcome, (3) were observational studies or interventional studies, and (4) were conducted in humans. Studies were not excluded based on publication year. The inclusion criteria are summarized within the PICOS framework in Table 1.

Table 1

PICOS criteria for inclusion of studies

Parameter	Criterion
Population	Human subjects
Intervention/exposure	DASH diet intervention/high adherence to the DASH diet
Control/comparator	Other dietary patterns/low adherence to the DASH diet
Outcome	Mental well-being measures
Study design	Peer-reviewed articles published in English, including intervention studies and observational studies

Data extraction

Two authors (J.T. and C.W.) independently extracted the following data from each eligible article: publication year, first author, study region, study design, sample size, participant characteristics (including age, biological sex/gender, disease history), assessment methods of psychological outcomes, measures of the adherence to the DASH diet, comparison groups and intervention groups of interventional studies, and study results. The extracted data were summarized and presented in tables.

Quality assessment

Study Quality Assessment Tools developed by the National Heart, Lung, and Blood Institute from the National Institutes of Health (NIH) were used to assess the quality of the studies included in this systematic review.¹⁴ The NIH study quality assessment tools ask specific questions (eg, “Was a sample size justification, power description, or variance and effect estimates provided?” “Was loss to follow-up after baseline 20% or less?” etc.). The assessment tool that matched each study’s design (eg, controlled intervention studies, cross-sectional studies, etc.) was used. The NIH tools offer guidance on how to engage in critical appraisal to determine an overall quality rating. Two authors (J.T. and C.W.) independently assessed all eligible studies based on the checklist of questions related to study design and reporting and categorized them into “Poor,” “Fair,” or “Good” quality according to the NIH guidelines. Disagreements were resolved through discussion and consultation with the senior author.

RESULTS

Study characteristics

The numbers of studies screened, assessed for eligibility, and included in the review are provided in Figure 1. The search yielded a final sample of 16 studies. Of these, 10 (62.5%) were observational studies, and 6 (37.5%) were randomized controlled trials (RCTs). Among the 10 observational studies, 5 were cross-sectional, 4 were longitudinal, and 1 was both cross-sectional and longitudinal. The studies were published between 2005 and 2021. The sample size of studies ranged from 36 to 19 270. Six (37.5%) study samples were drawn from European countries (Netherlands, France, Spain, Greece), 5 (31.25%) from Iran, 3 (18.74%) from the United States, 1 (6.25%) from Israel, and 1 (6.25%) from Australia. In terms of the NIH study quality assessment, 6 observational studies were rated as good, and 4 were rated as fair. Details of the study quality assessment for all studies are listed in Table S1 in the Supporting Information online. Of the RCTs, 3 were good and 3 were fair. Details of the 16 included studies’ characteristics are listed in Tables 2^12,15–23 and 3^24–29.

Figure 1

Flow diagram of the literature search process.

Table 2

Included research studies and their characteristics (observational studies)

Reference	Country	Study design	Sample size	Sex (female/male)	Study group	NIH quality
Gianfredi et al (2021)15	Netherlands	Longitudinal	2646	1154/1177	Individuals aged 40–75 y with type 2 diabetes mellitus	Good
Khayyatzadeh et al (2018)16	Iran	Cross-sectional	535	All girls	Adolescent girls living in Iran	Good
Polanska et al (2021)12	Europe	Cross-sectional	11 870	Mother–child pairs	Data from the ALPHABET project	Fair
Ferranti et al (2013)17	USA	Longitudinal	640	429/211	Healthy university and academic health center employees aged over 18 y	Fair
Faghih et al (2020)18	Iran	Cross-sectional	240	208/32	Healthy university students with an average age of 21.5 y	Fair
Valipour et al (2017)19	Iran	Cross-sectional	3846	2134/1712	Healthy adults with an average age of 36 y, who have participated in the Epidemiology of Psycho-Alimentary Health and Nutrition project	Good
Elstgeest et al (2019)20	Amsterdam	Cross-sectional	1312	686/636	Individuals over 55 y who have participated in the Longitudinal Aging Amsterdam Study	Fair
		Longitudinal (short-term change)	1233	635/598
		Longitudinal (long-term change)	687	347/340
Saharkhiz et al (2021)21	Iran	Cross-sectional	181	All women	Healthy women aged 18–25 y	Good
Recchia et al (2020)22	France	Longitudinal	5632	3413/6895	Individuals aged 35–55 y who participated in the Whitehall II prospective cohort study	Good
Perez-Cornago et al (2017)23	Spain	Longitudinal	19 270	8304/5747	Healthy adults with an average age of 37.5 y	Good

Table 3

Included research studies and their characteristics (RCT)

Reference	Country/location	Sample size (intervention/control)	Sex (female/male)	Study group	NIH quality
Torres and Nowson (2012)24	Australia	111/95	All women	Individuals with a body mass index (BMI) from 18 kg/m2 to 35 kg/m2, a home systolic blood pressure (BP) of ≥116 mmHg or a home diastolic BP of ≥78 mmHg, and were experiencing menopausal symptoms, or had passed through menopause	Fair
Green et al (2014)25	USA	18/18	26/10	Individuals with an overweight and obese BMI (BMI of 25 kg/m2–44.9 kg/m2) and who had been taking at least 1 antipsychotic medication at any consistent dose for a minimum of 30 d at the time they were identified	Good
Kirpizidis et al (2005)26	Greece	99/102	100/101	Hypertensive outpatients	Fair
Ziv et al (2013)27	Israel	CALM-BP: 58 DASH diet + exercise: 55	56/57	Adults aged 22 y–75 y with mean systolic and diastolic BP of 120 mmHg–180 mmHg and 70 mmHg–100 mmHg, respectively, and who have been treated with at least 1 antihypertensive drug 3 mo before inclusion	Fair
Khoshbakht et al (2021)28	Iran	43/43	1/79	Children aged 6 y to 12 y recently diagnosed with attention-deficit/hyperactivity disorder	Good
Ma et al (2016)29	USA	46/44	67/23	Individuals with an age of 18 y–70 y, BMI 18.5 kg/m2–39.9 kg/m2, a low-quality diet, and uncontrolled persistent asthma; race/ethnicity breakdown (intervention: control): non-Hispanic white: 46:41 non-Hispanic black: 13:9 Asian/Pacific Islander: 28:34 Hispanic/Latino: 13:16	Good

Abbreviations: NIH, National Institutes of Health; RCT, randomized controlled trial.

Mental health outcomes

Table 4^{12,15–23,30–34} displays the dietary assessment tools and psychological measurements used by all the observational studies included in the systematic review. Table 5^12,15–23 presents the specific variables the observational studies examined, including covariates and their results. Table 6^24–29 describes the intervention and comparison groups of all the RCTs included in the systematic review. Lastly, Table 7^24–29 includes the dietary assessment tools and psychological measurements used by the RCTs, together with their results.

Table 4

Dietary assessment tool and psychological measurements (observational studies)

Reference	Dietary assessment tool	Psychological assessment methods
Gianfredi et al. (2021)15	FFQs, the Dutch Healthy Diet score 2015 (DHD-score), and DASH scores determined based on Fung et al (2008)32	Depression: 9-item Patient Health Questionnaire (PHQ-9) and Mini-International Neuropsychiatric Interview (MINI)
Khayyatzadeh et al. (2018)16	168-item FFQs and DASH scores determined based on Fung et al (2008)32	Depression: Persian version of the Beck Depression Inventory (BDI), and aggression: Persian version of Buss–Perry questionnaire
Polanska et al. (2021)12	FFQs and DASH scores determined based on Fung et al (2008)32	Emotional symptoms: Strength and Difficulties Questionnaire (SDQ)
Ferranti et al. (2013)17	2005 Block FFQ, DASH scores determined based on modified version of Folsom et al (2007)34; sweets intake determined by following Dixon et al (2007)30	Perceived stress: 14-item Cohen Perceived Stress Scale (PSS); depression: 21-item Beck Depression Inventory II (BDI-II)
Faghih et al. (2020)18	FFQs and DASH scores determined based on Fung et al (2008)32	Mental health: 12-item General Health Questionnaire (GHQ-12); depression, anxiety, and psychological distress: Depression Anxiety Stress Scale-21 (DASS-21)
Valipour et al. (2017)19	106-item dish-based semi-quantitative FFQ and DASH scores determined based on Fung et al (2008),32 but modified by considering total grain intake as a nonhealthy food	Anxiety and depression: The Iranian validated version of Hospital Anxiety and Depression Scale (HADS); psychological distress: Iranian validated version of General Health Questionnaire (GHQ) with 12-items
Elstgeest et al. (2019)20	Dutch version of the FFQ of (HELIUS) and DASH scores determined based on Fung et al (2008)32	Depression: Center for Epidemiologic Studies Depression Scale (CES-D)
Saharkhiz et al. (2021)21	FFQs and DASH scores determined based on Fung et al (2008)32	Depression, anxiety, and psychological distress: DASS-21, and Quality of life: Short Form health survey (SF-12)
Recchia et al. (2020)22	Semi-quantitative FFQ and DASH scores determined based on Fung et al (2008)32	Depression: Center for Epidemiologic Studies Depression Scale (CES-D)
Perez-Cornago et al. (2017)23	136-item semi-quantitative FFQ, DASH scores determined based on Dixon et al (2007),30 Fung et al (2008),32 Günther et al (2009),33 and Mellen et al (2008)31	Depression: self-reported physician-diagnosed depression, clinical diagnosis of depression, and use of antidepressants

Abbreviations: DASH, Dietary Approaches to Stop Hypertension; FFQ, Food-Frequency Questionnaire.

Table 5

Results of included studies (observational studies)

Reference	Covariates adjusted	Association	Results
Gianfredi et al (2021)15	Age, sex, level of education, diabetes status, hypertension, total cholesterol, and high-density lipoprotein (HDL) cholesterol, history of cardiovascular diseases (CVDs), waist circumference, and partner status	Depression risk and DASH score	No association (HR: .95; 95% CI: .83–1.07)
		Depression risk and DASH diet adherence levels (tertials)	Medium adherence: no association (OR: 1.06; 95% CI: .81–1.39) High adherence: no association (OR: .83; 95% CI: .60–1.15)
Khayyatzadeh et al (2018)16	Age, energy intake, mother’s job status, passive smoker, menstruation, parent death, parent divorce, physical activity, body mass index (BMI), socioeconomic level, and education stage	Depression risk and DASH diet adherence levels (quartiles)	Q2: No association (OR: .79; 95% CI: .46–1.33) Q3: No association (OR: .68; 95% CI: .37–1.2) Q4: Negative association (OR: .47; 95% CI: .23–.92)
		Aggression and DASH diet adherence levels (quartiles)	Q2: No association (OR: .75; 95% CI: .42–1.31) Q3: No association (OR: .79; 95% CI: .46–1.33) Q4: No association (OR: .87; 95% CI: .45–1.59)
Polanska et al (2021)12	Parental birthplace/ethnic background, parental age, educational level, household income, household status during pregnancy, smoking, alcohol consumption during pregnancy, maternal passive smoking status, parity, maternal height, pre-pregnancy BMI and energy intake	Offspring depression risk and anxiety symptoms and maternal DASH score	Negative association (OR: .97; 95% CI: .95–.99)
		Offspring aggressive behavior symptoms and maternal DASH score	Negative association (OR: .97; 95% CI: .94–.99)
		Offspring ADHD symptoms and maternal DASH score	Negative association (OR: .97; 95% CI: .95–.98)
Ferranti et al (2013)17	No mention of covariates adjustment	Perceived stress and DASH score	No association (test statistics not provided)
		Depressive symptoms and DASH score	No association (test statistics not provided)
Faghih et al (2020)18	No mention of covariates adjustment	DASS-21 total score (depression, anxiety and stress) and DASH score	Negative association (P < .001)
		DASS-21 depression score and DASH score	Negative association (P < .001)
		DASS-21 anxiety score and DASH score	Negative association (P < .001)
		DASS-21 stress score and DASH score	Negative association (P < .001)
		GHQ-12 score and DASH score	Negative association (P < .001)
Valipour et al (2017)19	Age, sex, energy intake, marital status, socioeconomic status, smoking, physical activity, chronic disease, antidepressant use, supplement use, pregnant or lactating women, frequency of spice consumption, and BMI	HADS depression risk and DASH diet adherence levels (tertiles)	Whole population Moderate adherence: negative association (OR: .73; 95% CI: .59–.90) High adherence: no association (OR: .99; 95% CI: .79–1.25)
		HADS anxiety score and DASH diet adherence levels (tertiles)	Whole population Moderate adherence: no association (OR: .78; 95% CI: .60–1.02) High adherence: no association (OR: .77; 95% CI: .57–1.05) BMI < 25 population High adherence: negative association (OR: .61; 95% CI: .37–.98) BMI ≥ 25 population Moderate adherence: negative association (OR: .63; 95% CI: .42–.95)
		GHQ distress score and DASH diet adherence levels (tertiles)	Whole population Moderate adherence: no association (OR: 1.13; 95% CI: .91–1.41) High adherence: no association (OR: .98; 95% CI: .76–1.26)
Elstgeest et al (2019)20	Age, sex, cohort, education level, marital status, physical activity, smoking, and number of chronic diseases	Cross-sectional depression risk and DASH score	Men: no association (beta: –.40; 95% CI: –1.53 to .73) Women: positive association (beta: .86; 95% CI: .07–1.66)
		Short-term changes in depression risk and DASH score	Emerging depressive symptoms: no association (beta: –.33; 95% CI: –1.38 to .72) Remitted depressive symptoms: no association (beta: –.06; 95% CI: –1.19 to 1.07) Chronic/recurrent depressive symptoms: no association (beta: –.43: 95% CI: –1.52 to .67)
	Age, sex, cohort, education level, marital status, physical activity, smoking, number of chronic diseases, and current depressive symptoms	Long-term history of depression risk and DASH score	Men: no association (beta: –1.12; 95% CI: –2.63 to .40) Women: no association (beta: .86; 95% CI: .27–1.99)
Saharkhiz et al (2021)21	Age, BMI, and energy intake	DASS-21 depression risk and DASH diet adherence levels (tertiles)	Medium adherence: no association (OR: .90; 95% CI: .44–1.86) High adherence: no association (OR: 1.6; 95% CI: .78–3.42)
		DASS-21 anxiety score and DASH diet adherence levels (tertiles)	Medium adherence: no association (OR: .53; 95% CI: .26–1.12) High adherence: no association (OR: 1.6; 95% CI: .78–3.42)
		DASS-21 stress score and DASH diet adherence levels (tertiles)	Medium adherence: negative association (OR: .32; 95% CI: .14–.71) High adherence: negative association (OR: .70; 95% CI: .34–1.47) (P = .067)
Recchia et al (2020)22	Sex, age, ethnicity, total energy intakes, occupational grade, marital status, smoking behavior, physical activity, alcohol consumption, type 2 diabetes, cardiovascular disease, hypertension, HDL-cholesterol, body mass index, and cognitive impairment	Recurrent depression risk and DASH score	No association (OR: .91, 95% CI: .83–1.00)
Perez-Cornago et al (2017)23	Age, sex, smoking, physical activity, energy intake, living alone, unemployment, marital status, baseline hypertension, weight change, and personality traits	Diagnosis of depression and DASH diet adherence levels (Dixon's DASH index: Quartiles)	Q2: No association (HR: 1.14; 95% CI: .89–1.46) Q3: No association (HR: 1.21; 95% CI: .90–1.63) Q4: Positive association (HR: 1.44; 95% CI: 1.05–1.99)
		Diagnosis of depression and DASH diet adherence levels (Mellen’s DASH index: Quintiles)	Q2: Negative association (HR: .68; 95% CI: .51–.91) Q3: Negative association (HR: .55; 95% CI: .37–.83) Q4: No association (HR: .93; 95% CI: .70–1.24) Q5: No association (HR: 1.02; 95% CI: .76–1.37)
		Diagnosis of depression and DASH diet adherence levels (Fung’s DASH index: Quintiles)	Q2: Negative association (HR: .62; 95% CI: .46–.84) Q3: No association (HR: .79; 95% CI: .58–1.09) Q4: Negative association (HR: .68; 95% CI: .51–.92) Q5: No association (HR: 1.09; 95% CI: .81–1.45)
		Diagnosis of depression and DASH diet adherence levels (Günther’s DASH index: Quintiles)	Q2: No association (HR: .84; 95% CI: .62–1.13) Q3: Negative association (HR: .72; 95% CI: .52–.99) Q4: No association (HR: .87; 95% CI: .64–1.18) Q5: No association (HR: 1.22; 95% CI: .90–1.65)
		Diagnosis of depression and DASH diet adherence levels (low vs high)	Fung’s DASH index: negative association (HR: .76; 95% CI: .61–.94) Dixon’s DASH index: no association (HR: 1.21; 95% CI: .97–1.51) Mellen’s DASH index: no association (HR: .80; 95% CI: .64–1.00) Günther’s DASH index: no association (HR: .88; 95% CI: .69–1.11)
		Diagnosis of depression and antidepressant use and DASH diet adherence levels (low vs high)	Fung’s DASH index: negative association (HR: .63; 95% CI: .41–.95) Dixon’s DASH index: no association (HR: 1.68; 95% CI: .45–1.04) Mellen’s DASH index: no association (HR: 1.47; 95% CI: .95–2.30) Günther’s DASH index: no association (HR: 1.01; 95% CI: .63–1.62)

Significant results are indicated by bold type. Abbreviations: B, unstandardized regression coefficient; CI, confidence interval; DASH, Dietary Approaches to Stop Hypertension; DASS, Depression Anxiety Stress Scale; GHQ-12, 12-item General Health Questionnaire; HADS, Hospital Anxiety and Depression Scale; HR, hazard ratio; OR, odds ratio.

Table 6

Intervention and comparison groups of included studies (RCTs)

Reference	Comparison group	Intervention group	Intervention duration
Torres and Nowson (2012)24	Healthy diet (HD): based on general dietary guidelines to decrease fat, increase complex carbohydrates, and restrict red meat consumption	Moderate-sodium DASH-type diet (VD): based on the low-sodium DASH diet, but different in including 6 servings per week of lean red meat (DASH diet restricts to 3.5 servings per week)	14 weeks
Green et al (2014)25	Usual care control condition	DASH diet: provided education and encouragement to make lifestyle changes	12 weeks
Kirpizidis et al (2005)26	Candesartan group: treated with candesartan tablet	DASH diet + Candesartan: treated with candesartan tablet and were given a weekly DASH eating plan menu informing the patient about the food group to be followed	16 weeks
Ziv et al (2013)27	DASH + exercise: including group walk, grocery shopping and healthful home cooking education, and dietary plan	Comprehensive Approach to Lower Measured Blood Pressure (CALM-BP): including group walk, grocery shopping and healthful home cooking education, dietary plan, and relaxation and stress management intervention	16 weeks
Khoshbakht et al (2021)28	Usual diet of Iranian children: designed for each participant according to their estimated energy requirement by investigators	DASH diet: designed for each participant according to their estimated energy requirement by investigators	12 weeks
Ma et al (2015)29	No intervention	A dietician-delivered behavioral intervention that promoted the DASH eating pattern, had individual and group sessions, and was grounded in social cognitive theory	6 months

Abbreviations: DASH, Dietary Approaches to Stop Hypertension; RCT, randomized clinical trial.

Table 7

Results of included studies (RCTs)

Reference	Dietary data collection	Psychological assessment methods	Results
Torres and Nowson (2012)24	Food group diaries	An abbreviated 37-item version of the Profile of Mood States (POMS)	Significant improvement in mood and anger in the DASH-type diet compared with the comparison group (P < .05)
Green et al (2014)25	Food and exercise diaries	Eating Habits Confidence and Exercise Confidence questionnaires and Wisconsin Quality of Life Index Medication Adherence items	Net improvements in the intervention group shown in body image and self-esteem (P = .001)
Kirpizidis et al (2005)26	Notebook to keep track of what has been eaten	Subjective Symptoms Assessment Profile (SSAP) questionnaire	Significant improvement for the emotional and mental component scores with the combination of candesartan and DASH diet (P < .03)
Ziv et al (2013)27	Adherence questionnaires at week 4 and week 10	Life quality Short Form 36 (SF36) questionnaires	The Comprehensive Approach to Lower Measured Blood Pressure group increased in physical, mental, and total scores of the SF36 life quality (a net score improvement of 11.5 [95% CI: 5–17, P < .001], 10.1 [95% CI: 3.5–16.6, P = .003], and 10.9 [95% CI: 4.6–17.1, P = .001], respectively), but there were no significant improvements in any of these parameters in the DASH group
Khoshbakht et al (2021)28	3 d dietary records	Abbreviated 10-item Conner’s scale (ACS), 18-item Swanson, Nolan and Pelham (SNAP-IV) scale, and Strengths and difficulties questionnaire (SDQ)	Emotion and attention-deficit/hyperactivity disorder symptoms significantly improved in the intervention group (P < .05)
Ma et al (2016)29	Multiple-pass 24-h diet recalls and DASH based on Mellen et al (2008)31	15-item Mini Asthma-specific Quality of Life Questionnaire	Significantly higher percentages of intervention participants showed significant improvements in quality of life (overall: the difference between the 2 groups was .4; 95% CI 0–.8)

Abbreviations: CI = confidence interval. DASH, Dietary Approaches to Stop Hypertension; RCT, randomized clinical trial.

Clinically diagnosed depression/self-reported depression risk and depressive symptoms

Clinical diagnosis of depression

No RCTs examined clinically diagnosed depression incidence. One observational study investigated the association between adherence to the DASH diet and the incidence of clinically diagnosed depression by analyzing longitudinal data of healthy adults in Spain and fully adjusting for age, sex, socioeconomic factors (eg, educational level, marital status, and household income), health behaviors (eg, smoking status, physical activity, and alcohol consumption), total energy intake, weight change, and personality traits of competitiveness and dependency.²³ They investigated the association using 4 previously published DASH diet indices—Dixon’s, Mellen’s, Fung’s, and Günther’s indices.^30–33 They also examined the association using 2 different definitions of depression: 1 less conservative (self-reported physician-diagnosed depression) and 1 more conservative (diagnosed depression and antidepressant use). Interestingly, a weak positive association between Dixon’s DASH index and incident depression was found when comparing the fourth quartile with the first quartile of the DASH score. Moreover, Perez-Cornago et al²³ revealed a non-linear, U-shaped relationship between diagnosed depression and Mellen’s, Fung’s, and Günther’s DASH diet indices. Medium adherence to the DASH diet measured by these indexes is associated with significantly lower hazard ratios of developing depression. However, when adherence to the DASH diet was divided into 2 groups (low adherence versus high adherence), only Fung’s DASH index displayed an inverse relationship with incident depression for both less conservative and more conservative definitions.

Self-reported depression risk

Similarly, no RCTs examined depression risk determined by self-reported questionnaires. Seven observational studies investigated associations between self-reported depression risk and adherence to the DASH diet.^{12,15,16,19–22} The findings were mixed. Three studies found a significant association.^12,16,19 Khayyatzadeh et al¹⁶ conducted a cross-sectional study examining the association between adherence to the DASH diet and depression risk among girls aged 12–18 in Iran. They found that a high adherence to the DASH diet (fourth quartile of the DASH score) was associated with lower depression risk compared with a low adherence to the diet (first quartile of the DASH score) after fully adjusting for age, socioeconomic variables, health behaviors, parental status, and energy intake. Valipour et al¹⁹ also explored the association between adherence to the DASH diet and the odds of depression by analyzing cross-sectional data of healthy adults in Iran. Their analyses showed that, instead of a high adherence to the DASH diet (third tertial of the DASH score), a moderate (second tertial of the DASH score) adherence was inversely associated with depression risk, when taking demographic characteristics (eg, age and sex), socioeconomic factors, health behaviors, health status variables (eg, gestational diabetes, hypertension, psychological distress, and depression), and medication use into account. Finally, Polanska et al¹² examined cross-sectional data of mother–child pairs from 4 European cohorts (United Kingdom, France, Netherlands, and Poland). They found that a greater maternal adherence to the DASH diet was associated with a lower depression risk of offspring after fully adjusting for potential confounding variables, including parental demographic variables, parental socioeconomic factors, health behaviors, health status before and during pregnancy, and energy intake during pregnancy.

The other 4 studies did not find an inverse association between the DASH score and depression risk in the crude and fully adjusted models.^15,20–22 Recchia et al²² conducted analyses on a London sample of men and women aged 35 years–55 years from a longitudinal study. They found no inverse association between long-term adherence to the DASH diet and recurrent depression risk, fully controlling for demographic variables, total energy intake, health behaviors, and health status variables. Gianfredi et al¹⁵ obtained the same null result when examining longitudinal data of individuals aged 40 years –75 years with type 2 diabetes mellitus in the Netherlands, with the associations fully adjusted for sociodemographic variables (eg, age, sex, and educational levels), health status variables, and health behaviors. Saharkhiz et al²¹ examined the association between the DASH diet and depression among girls aged 18–25 years in Iran. Their cross-sectional study also found a null relationship after adjusting for age, BMI, and energy intake. Interestingly, Elstgeest et al²⁰ found mixed results by looking into both cross-sectional and longitudinal data of Dutch individuals older than age 55 years and fully adjusted for sociodemographic variables, health behaviors, and the number of chronic diseases. Their cross-sectional data revealed a positive association between DASH diet adherence and depression risk among women, whereas their longitudinal data revealed no significant association between these 2 variables.

Self-reported depressive symptoms

In terms of self-reported depressive symptoms, 1 RCT²⁴ investigated the effect of the DASH diet on depressed mood by randomly assigning postmenopausal women aged 49 years and older to either the DASH-type vitality diet group or a healthy diet group for 14 weeks. In the DASH-type vitality diet group, participants adhered to a low-sodium, high-potassium, and high-magnesium diet with 6 servings per week of lean red meat (2.5 servings more than the original DASH diet). In the healthy diet group, participants adhered to a diet that was higher in whole-grain bread/cereals and lower in fruits and vegetables than the low-sodium DASH diet. Torres and Nowson²⁴ found no significant difference in depressed mood between the 2 groups after the intervention. Participants in both groups had improved scores on depressed mood.

Two observational studies examined the association between the 2 variables.^17,18 Faghih et al¹⁸ conducted a cross-sectional study among healthy university students in Iran. They obtained a negative correlation coefficient for the association between the DASH diet and depressive symptoms assessed by the Depression Anxiety Stress Scale-21 (DASS-21) without any indication of controlling for covariates. However, the study by Ferranti et al¹⁷ did not find the same result: they found a null relationship after analyzing longitudinal data of healthy adult university and academic health center employees in the United States, with no indication of controlling for covariates.

Emotion and mood

Two RCTs found the DASH diet had a positive impact on emotion and mood.^24,28 One RCT examined the effect of the DASH diet on children aged 6 years–12 years with attention-deficit/hyperactivity disorder (ADHD) in Iran.²⁸ The children were randomly assigned to receive a DASH or control diet for 3 months. The control diet mirrored the Iranian children’s usual diet, had lower amounts of fruits and vegetables compared with the DASH diet, and allowed refined grains, full-fat dairy, meats, and simple sugars that were avoided in the DASH diet. With emotional symptoms defined as “unhappy” and “downhearted” compared with their counterparts in the control diet group, their study revealed that the children in the DASH diet group experienced significant improvement in their emotional symptoms after adjusting for sociodemographic variables, parental education and occupation, energy intake, and baseline values. In addition, Torres and Nowson²⁴ compared the mood of postmenopausal women randomly assigned to adhere to a DASH-type vitality diet or to a healthy diet. They found decreased anger and enhanced mood measured by the Profile of Mood States among women in the vitality diet group compared with those in the healthy diet group.²⁴

Three cross-sectional studies investigated the association between the DASH diet and anxiety.^18,19,21 While using the same questionnaire to assess anxiety levels, DASS-21, Faghih et al¹⁸ found a negative correlation between the DASH diet and anxiety levels without adjustment for covariates, but Saharkhiz et al²¹ obtained a null result after adjusting for sociodemographic variables and energy intake. Valipour et al¹⁹ demonstrated that the association between DASH diet adherence and anxiety varied based on participants’ BMI in their subgroup analyses, controlling for demographic characteristics, health behaviors, health status variables, and socioeconomic factors. Specifically, greater adherence to the DASH-style diet was associated with lower odds of anxiety in normal-weight individuals. In individuals with overweight BMI, moderate but not high adherence to the DASH-style diet was inversely associated with anxiety levels.

The evidence for an association between adherence to the DASH diet and psychological distress or stress was also mixed. Two cross-sectional studies found an inverse association between adherence to the DASH diet and psychological distress.^18,21 However, 1 cross-sectional study and 1 longitudinal study failed to find any significant association.^17,19 It should be noted that the studies of both Valipour et al¹⁹ and Faghih et al¹⁸ were cross-sectional, focused on the Iranian population, and both of them used the General Health Questionnaire.³⁵ The only differences were that Faghih et al¹⁸ recruited university students as participants and did not mention controlling for covariates, while Valipour et al¹⁹ analyzed data from general adults and fully adjusted for demographic characteristics, health behaviors, health status variables, and socioeconomic factors.

Quality of life

Studies that investigated the association between the DASH diet and quality of life found conflicting results. Two RCTs found that the DASH diet enhanced scores on the Subjective Symptoms Assessment Profile.^26,29 Specifically, Ma et al²⁹ examined the effects of the DASH diet and the control diet on the quality of life among adults with objectively confirmed uncontrolled asthma and a low-quality diet in the United States. Their study demonstrated that after receiving the 6-month intervention, participants in the DASH diet group significantly improved their quality of life compared with their counterparts in the control group, who maintained their usual diet in the crude model, and after adjusting for asthma long-term controller or/and short-acting β-agonist medication use. In addition, the study of Kirpizidis et al²⁶ compared the effects of candesartan, a type of blood pressure–lowering medicine, when used alone (control intervention) or in combination with the DASH diet on the quality of life with a focus on adults with mild to moderate hypertension. They found the addition of the DASH diet significantly improved participants’ emotional and mental components of the Subjective Symptoms Assessment Profile compared with the control intervention after 16 weeks. However, the RCT of Ziv et al²⁷ failed to find any significant improvement in life quality in the DASH diet group using the life quality Short Form 36 questionnaire. In their study, adults treated with antihypertensive drugs were randomly assigned to either the Comprehensive Approach to Lowering Measured Blood Pressure (CALM-BP) intervention or the DASH intervention for 16 weeks. Those in the CALM-BP group followed the rice-based diet, which contained a lower percentage of protein and fat and a higher percentage of potassium compared with the DASH diet. Additionally, while both interventions included group walks, grocery shopping, and healthful home cooking education, the participants in the CALM-BP group received stress management sessions, but those in the DASH group did not. Participants’ quality of life between the 2 groups was not compared.

Aggressive behavior

Two observational studies investigated the association between the DASH diet and aggressive behavior using a cross-sectional design.^12,16 A null relationship was found among girls aged 12 years–18 years in Iran in the crude model, and after adjustments for demographic characteristics, energy intake, health behaviors, and socioeconomic factors.¹⁶ However, a negative association between offspring aggressive behavior symptoms and maternal DASH score was observed by Polanska et al¹² when examining European mother–child pairs and fully adjusting for demographic characteristics, socioeconomic factors, health behaviors, health status variables, and energy intake during pregnancy.

Self-esteem and body image

An RCT investigated the impact of a 12-week DASH dietary intervention on body image and self-esteem among adults who had overweight and obese BMI and who had been taking at least 1 antipsychotic medication in the United States.²⁵ The researchers found that the DASH intervention improved the body image and increased the self-esteem of participants in the DASH diet group compared with the control group, who maintained their usual diet.

ADHD symptoms

One RCT examined the effect of the DASH diet on children aged 6 years–12 years with ADHD in Iran.²⁸ As mentioned earlier, children participating in the study were randomly assigned to receive the DASH diet or the control diet (which was similar to Iranian children’s usual diets) for 3 months. Khoshbakht et al²⁸ revealed that children receiving the DASH-style diet significantly improved their ADHD symptoms as assessed by the abbreviated Conner’s scale and Strengths and Difficulties Questionnaire, compared with children receiving the control diet, after adjusting for sociodemographic variables and energy intake. The cross-sectional study conducted by Polanska et al¹² also found that ADHD symptoms were less likely to be observed among children whose mothers had a higher DASH score, after fully adjusting for demographic characteristics, socioeconomic factors, health behaviors, health status variables, and energy intake during pregnancy.

DISCUSSION

The DASH diet is widely prescribed, but the psychological impacts of the diet are not fully understood. This systematic review of currently available studies revealed that there is some evidence showing that the DASH diet can promote better psychological outcomes such as mood, emotional status, incidence of depression, and depressive symptoms.

As RCTs are the gold standard for studying causal relationships, their results should be emphasized when evaluating the impact of the DASH diet. Among the 6 RCTs included in this systematic review, 5 studies found a positive impact of the DASH diet on mental health.^{24–26,28,29} Torres and Nowson²⁴ (2012) found that the DASH diet significantly improved mood among postmenopausal women. This result was consistent with their previous study showing an overall positive effect of a dietary pattern low in sodium and high in potassium on mood.³⁶ However, the authors acknowledged that other factors might have contributed to the mood improvements, such as regular contact with motivating research staff. The perception of success in meeting dietary targets could have resulted in participants feeling empowered and happy. In addition, Green et al²⁵ revealed a net improvement in body image and self-esteem among overweight and obese adults who had been taking at least 1 antipsychotic medication in the United States. Kirpizidis et al²⁶ showed that the DASH diet significantly improved the emotional and mental component scores in hypertension outpatients who were taking candesartan tablets. The ADHD symptoms of Iranian children aged 6 years–12 years with ADHD were shown to be reduced by the DASH diet.²⁸ The DASH diet also significantly improved the quality of life among people with uncontrolled persistent asthma.²⁹

However, it is noteworthy that most studies focused on different populations and aspects of mental health, so it was difficult to evaluate the consistency of most results across RCTs, and it is likewise difficult to evaluate generalizability to other patient populations. Quality of life was the only measure that was investigated by more than 1 RCT, but the findings were mixed. Contrary to the study of Ma et al²⁹, which revealed an improvement in the Asthma Control Questionnaire (ACQ-15) quality of life among people with uncontrolled persistent asthma, that of Ziv et al²⁷ failed to find any improvement in the 36-Item Short Form Health Survey (SF-36) life quality among adults who took antihypertensive medication. Such a difference might be due to the variations between the 2 study populations or the surveys used to evaluate the quality of life. More studies with harmonized assessments are needed to reach a more definite conclusion.

The quality ratings of studies included in this systematic review were evaluated based on the NIH study quality assessment tool.¹⁴ Some studies were defined as having a low risk of bias, and thus more confidence could be placed in their results than those of fair quality studies. Taking the strength of each study into account, the DASH diet’s positive effects on body image and self-esteem,²⁵ emotion and ADHD symptoms,²⁸ and quality of life²⁹ were supported with greater confidence.

Compared with RCTs, there was more evidence regarding depression risk, depressive symptoms, stress, and anxiety from observational studies. All observational studies included in this systematic review investigated the association between the DASH diet and depression/depressive symptoms.^12,15–23 Most observational studies investigated depression risk as assessed by self-reported questionnaires. One longitudinal study and 3 cross-sectional studies found a negative association with the DASH diet,^12,16,19,23 while 2 longitudinal studies^15,22 and 1 cross-sectional study²¹ found a null result. Interestingly, 1 study looked into both cross-sectional and longitudinal data and found an unexpected positive association among women from the cross-sectional data and no association from the longitudinal data in the fully adjusted model.²⁰ The rest of the observational studies examined self-reported depressive symptoms; 1 cross-sectional study¹⁸ supported the association, and 1 longitudinal study¹⁷ did not. This inconsistency remained after stratifying the studies based on the quality of the studies. Such inconsistency might have resulted from different measures of DASH adherence, different scales used to measure depressive symptoms, and different categorizations of the adherence levels. Some measures of the DASH diet consisted of fewer components compared with other indices, and their score ranges thus varied. Six different questionnaires/scales were implemented to assess depression/depressive symptoms across the studies included in the current review. Finally, the adherence levels of the DASH diet ranged from 2 levels (low adherence vs high adherence) to 5 levels (quintiles). Notably, among studies that found a significant inverse association between the DASH diet and depression/depressive symptoms, 2 out of 3 studies with multiple adherence levels of the DASH diet observed that a moderate adherence to the DASH diet instead of a high adherence was associated with a lower depression risk, suggesting a nonlinear relationship.

Observational studies for anxiety and stress provided some evidence supporting a negative association between the DASH diet and anxiety/perceived stress.^18,19,21 As the DASH diet is considered a type of healthy diet, this association is consistent with previous studies showing that healthy diets were linked to lower odds of anxiety.³⁷ However, some studies obtained different results. The cross-sectional study of Saharkhiz et al²¹ found no association between the DASH diet and anxiety after fully controlling for covariates. The longitudinal study of Ferranti et al¹⁷ and the cross-sectional study of Valipour et al¹⁹ failed to find any association between the DASH diet and stress. The inconsistency in the study by Ferranti et al¹⁷ might be due to its use of a different DASH index (developed by Folsom et al³⁴) compared with the index used by other studies (developed by Fung et al³²). Similarly, a potential explanation for the inconsistency observed in Valipour et al¹⁹ is its modification of the DASH index.

When study strength was taken into consideration, that there is an association between stress/distress and adherence to the DASH diet cannot be concluded with confidence because the evidence was rather mixed. Specifically, 2 cross-sectional studies found a negative association, while 1 cross-sectional and 1 longitudinal did not.^17–19,21 The negative association between depression and adherence to the DASH diet can be reasonably inferred because of the consistent findings across studies.^12,16,19,23 However, it is noted that all observational studies, by definition, cannot infer causality, unlike RCTs.

Strengths and limitations

This review has several strengths. A protocol registered in the PROSPERO database prior to launching the systematic review was developed and followed. Both observational and interventional studies were included in the review to ensure comprehensiveness of the review. However, the results of this review must be interpreted in the context of certain limitations. Owing to the small number of RCTs, a meta-analysis was not conducted, and thus it was not possible to provide an estimation of aggregated effect sizes. In addition, most studies included in this systematic review were observational studies, which cannot be used to demonstrate causality. Thus, further studies are required to determine the robustness of these findings and provide a more rigorous understanding of the effect of the DASH diet on mental health. Another limitation is that related outcomes such as post-traumatic stress disorder or stress or severe mental disorders like schizophrenia were not specifically searched for, and may result in missing highly relevant studies due to the discrepancy between the operationalization of mental well-being and the relatively narrower search strategy. Therefore, the scope of this systematic review is likely limited. Future systematic reviews on the DASH diet and psychological well-being should consider including those outcomes in their search to draw a more complete picture of the existing literature on the topic.

Future directions

Given the limited studies available in the literature on the DASH diet and psychological well-being, future research on the topic is needed. Depression and depressive symptoms were the most investigated mental health outcome associated with the DASH diet, but the current evidence was found to be inconsistent. To draw a clearer conclusion about the relationship between adherence to the DASH diet and the incidence of depression, more studies are needed to examine the relationship between different levels of DASH diet adherence and depressive symptoms. However, adherence is a difficult variable to assign randomly, and therefore future work may have to rely on quasi-experimental designs. Future studies should also consider employing more consistent measures of depressive symptoms and adherence to the DASH diet. In addition, more studies, especially RCTs, should consider examining these outcomes among individuals with mental health concerns. Only 1 RCT²⁸ in the current review targeted a population with mental disorders, representing a gap in the literature. Moreover, the longest intervention duration of current RCTs on this topic was 6 months. Future researchers could design RCTs with a longer follow-up time to better assess the long-term effect of the DASH diet on various mental health outcomes.

For future researchers to conduct higher quality studies that explore the effect of dietary adherence or interventions on mental well-being, some challenges should be considered. It is difficult to blind participants from behavioral interventions such as adhering to a diet. Future studies should use more rigorous active control groups such that participants in the control group receive a comparable diet or treatment, instead of having a no-treatment or usual care control group. Doing so will facilitate blinding, as participants could be blinded to the goals of the particular diet that they have been assigned to. At the very least, to avoid bias, researchers accessing the outcomes should be blinded to the treatment group assignment. For the observational studies, the presence of confounding variables has been a challenge in that it can distort the estimation of the association of interest, leading to an incorrect conclusion. Future studies should consider applying more sophisticated causal inference methods (such as directed acyclic graphs, propensity score matching, weighting, and mediation analysis) to better map out potential confounding variables and balance the covariates. Another challenge is obtaining sufficiently large sample sizes that can detect the difference in outcomes with relatively high statistical power. Because RCTs on diets require participants to adhere to the dietary guidelines or to receive dietary interventions, the number of participants enrolled in the study is often limited, and the dropout rate could be high. Therefore, future studies should conduct a priori sample size analyses predicting a conservatively high dropout rate. How to keep participants engaged throughout the study period should be a consideration for researchers while they are developing study protocols. Additionally, studies should strive to recruit participants who are representative of the study population of interest to achieve greater generalizability. Lastly, while most studies used Fung’s DASH index, there are still some studies that either adjusted Fung’s DASH index for the cultural context or used a different index.^17,19 As seen in the results of Perez-Cornago et al,²³ the association between the DASH diet and mental health outcomes can vary depending on the use of the various DASH indices. Therefore, future studies should be more cautious when deciding which DASH index to use and when interpreting the results. Future studies should also consider further investigating the differences between these DASH diet indices in order to standardize the use of the measure.

Implications

Because of the limited literature, the mechanisms through which the DASH diet influences psychological health remain unknown. However, a growing body of evidence indicates that oxidative stress and inflammation might play significant roles in common mental health problems such as depression, anxiety, and psychological stress. A cross-sectional observational study from 8 countries showed that sugar intake was related to depression due to increased oxidative stress,³⁸ which aligns with the low-glycemic index in the DASH diet. Other characteristics of the DASH pattern, such as its high vitamin K and mineral content, might also be responsible for observed associations between the DASH diet and mental well-being. Higher dietary intakes of vitamin K and minerals were reported to be independently associated with a decreased risk of depression, anxiety, and psychological stress.^39–41 Indeed, several RCTs in adults have shown that adherence to the DASH diet significantly improved the biomarkers of oxidative stress and inflammation.^42,43 As there is no consensus on the mechanism of how the DASH diet affects mental health, more mechanistic studies are needed. First, however, more controlled studies are necessary to determine whether the DASH diet has a causal effect on psychological well-being.

CONCLUSION

Literature on the relationship between the DASH diet and mental health outcomes is scarce, especially randomized controlled trials. Current evidence supports the view that the DASH diet is likely to have positive effects on mental well-being. However, some results were inconsistent, possibly resulting from differences in methods of assessments of the DASH diet and mental health outcomes. With future studies of sufficient methodological quality confirming the positive influence of the DASH diet on mental health, health practitioners will be empowered to design more comprehensive health strategies. A combination of psychological and nutritional perspectives is likely to promote better overall well-being.

Acknowledgments

Author contributions. A.J.T. oversaw the completion of this project. A.J.T. and J.T. contributed to the conceptualization and the design of the systematic review. J.T. and C.W. collected and screened articles for review, conducted study quality assessment, constructed the result tables, and critically interpreted the data. All authors contributed to the writing of the main text and participated in the revisions of the article.

Funding. This systematic review received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. A.J.T.’s time is supported by NIH grants R01DK128575 and R01HL158555.

Declaration of interest. The authors have no conflicts of interest to declare.

Supporting Information

The following Supporting Information is available through the online version of this article at the publisher’s website.

Table S1 Study quality assessment results⁴⁴

Appendix S1 PRISMA checklist⁴⁵