The effect of MIND diet on sleep status, mental health, and serum level of BDNF in overweight/obese diabetic women with insomnia: a randomized controlled trial

Abstract

Insomnia is common in type 2 diabetes mellitus (T2DM) and affects mental health and quality of life. The present study aimed to examine the efficacy of MIND (Mediterranean-DASH Intervention for Neurodegenerative Delay) diet on the anthropometric measurements, sleep quality, depression, anxiety, and serum levels of cortisol and brain derived neurotrophic factor (BDNF) in type 2 diabetic women with insomnia. This randomized controlled trial (RCT) involved 44 type 2 diabetic women with insomnia, aged 30 to 65 years, who were randomly assigned to be under the MIND low-calorie diet (n = 22) or a low-calorie diet (LCD) as the control group (n = 22) for 12 weeks. The above-mentioned variables were assessed at the beginning and the end of intervention. Following the MIND diet for 12 weeks accompanied by the significant decrease of waist circumference and significant improvement of sleep quality, depression, and anxiety compared to the control group. In addition, the MIND diet vs. LCD group exhibited a significant reduction in the cortisol levels and a significant increase in BDNF. This study provides promising evidence of the effectiveness of the MIND diet in improving the sleep quality, mental health, and some related biochemical parameters in diabetic women with insomnia.

Trial registration: IRCT20181111041611N8.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-91389-y.

Introduction

Type 2 diabetes mellitus (T2DM), is a most common endocrine disorder and a global public health problem in the 21st century¹. The International Diabetes Federation (IDF) estimates that 783 million people will suffer from diabetes by 2045². T2DM is associated with different physical and psychological complications³. Sleep disorders like insomnia, obstructive sleep apnea and restless legs syndrome⁴, depression and anxiety are common psychological disorders in T2DM³ caused by some direct adverse effects on the central nervous system (CNS) or other mechanisms like increased oxidative stress and inflammatory responses^5–8.

Insomnia is usually characterized by inability to fall or stay asleep, or waking up earlier than planned despite having the sufficient time to rest⁹. The prevalence of insomnia is significantly higher in T2DM. The reported rate is 39%, four times higher than in the general population¹⁰. Increased activity of the sympathetic nervous system due to dysautonomia can lead to the occurrence of microarousals in these patients, which can ultimately impair the quality of sleep^11–13. Diabetics are often under increased psychological pressure to manage their disease, which can lead to depression, a major contributor poor sleep^3,14. On the other hand, insomnia can increase the risk of depression through some mechanisms. Insomnia is often associated with increased activity of the hypothalamic-pituitary-adrenal (HPA) axis and increased levels of cortisol and inflammatory cytokines^15–18. Cortisol and the HPA axis are negatively related to brain-derived neurotrophic factor (BDNF), the main neurotrophin of the CNS, which may play a major role in depression¹⁹.

Benzodiazepines are typically considered as the most popular option for the treatment of insomnia, depression, and anxiety²⁰. Benzodiazepines increase the level of gamma-aminobutyric acid (GABA) in the CNS²¹. However, long-term use of these drugs is associated with different side effects, including addictive potential²², dementia²¹, balance disorders²³ and cognitive decline²⁴. Therefore, it appears necessary to use alternative methods^25–27 that not only protect the CNS but also effectively treat the insomnia.

Numerous studies have examined the effects of dietary interventions on improving the sleep quality. In a recent study, it was found that adherence to the DASH (Dietary Approaches to Stop Hypertension) diet resulted in a significant reduction in sleep quality, depression and stress scores after a 12-week period²⁸. In addition, previous studies revealed a negative association between adherence to the Mediterranean diet and insomnia symptoms^29,30.

Healthy dietary patterns like Mediterranean and DASH diets, are associated with numerous health benefits, such as reducing the risk of cardiovascular disease, obesity, T2DM, and certain cancers^31,32. The MIND (Mediterranean-DASH Intervention for Neurodegenerative Delay) dietary pattern recently introduced by Morris et al. is a combination of the DASH and Mediterranean dietary patterns³³. This dietary approach was initially investigated for its association with Alzheimer’s disease and cognitive function^33,34. The MIND dietary pattern is characterized by specific features, with an emphasis on the consumption of berries and green leafy vegetables rich in polyphenolic compounds^33,35,36. Phenolic compounds, such as flavonoids, phenolic acids and tannins play a critical role in the promoting cognition and brain health³⁷. Results of a meta-analysis showed the effectiveness of polyphenols in enhancing neurotrophic factors³⁷. Phenolic compounds have anti-inflammatory effects and protect against the oxidative stress³⁸. On the other hand, polyphenols have a positive effect on BDNF and cortisol levels through different mechanisms^39,40. MIND dietary pattern also includes food items for brain health such as: vegetables, nuts, legumes, whole grains, seafood, chicken, olive oil as the main cooking oil and moderate wine. Moreover, the MIND diet restrict the consumption of foods that are unhealthy for the brain, such as meat, butter and margarine, cheese, sweets and fried foods, and fast food³³. Modern artificial intelligence (AI) tools, such as ChatGPT, can serve as valuable resources to obtain information about diets such as the MIND diet. However, users must engage with these tools thoughtfully and critically to ensure they are making informed decisions based on accurate data. This approach helps to mitigate the risks associated with dietary misinformation and achieve better health outcomes⁴¹. There is evidence that greater adherence to the MIND diet decreased odds of experiencing poor sleep quality and related sleep outcomes⁴². In addition, previous studies have shown reverse associations between the MIND diet and depression, stress and anxiety^43–46. Albeit, some studies have failed to demonstrate this association^42,47. It should be mentioned that, most of the studies in this area are observational, and to the best of our knowledge, no clinical trial has yet been conducted to investigate the impact of the MIND diet on sleep patterns, depression and anxiety.

Given the high prevalence of insomnia in T2DM and its association with mental health, the present study aimed to investigate the effects of MIND diet on the anthropometric measurements, sleep quality, depression, anxiety, and serum levels of cortisol and BDNF in type 2 diabetic women with insomnia.

Materials and methods

Participants and sample size

This open-label, randomized controlled trial (RCT) involved 44 patients with type 2 diabetes and insomnia in Sanandaj, Iran. Participants were recruited voluntarily through a city-wide announcement. The sample size for each group was calculated to be 18, based on the mean and standard deviation (SD) of sleep quality from Daneshzad et al.’study²⁸, aiming for a 5% probability of type 1 error (α = 0.05) and 95% study power.  To account for a potential 20%  dropoout rate, the sample size was increased to 22 per group.

Inclusion and exclusion criteria

Eligible type 2 diabetic women aged 30–65 years with an Insomnia Severity Index (ISI) of more than 7 and a Body Mass Index (BMI) of 25 to 35 kg/m² were voluntarily participated in the study. Exclusion criteria were pregnancy and breastfeeding, smoking and alcohol consumption, insulin use, sleep apnea, shift work, use of weight loss diets and medication in the last three months, clinically diagnosed endocrine, metabolic, renal, neurological or psychiatric disorders and use of sedatives, analgesics and hypnotics. An initial screening regarding the inclusion and exclusion criteria was conducted by telephone.

Procedure

Subjects were randomly assigned in a 1:1 ratio to the MIND low-calorie diet (MLCD) or a low-calorie diet (LCD) using a random sequence generated by the random block method, which creates 11 blocks through the website https://www.sealedenvelope.com. The study period was 12 weeks with a clause follow-up.

Due to the nature of the intervention, this study was conducted in an open-label form. Subjects were assigned to study groups by a person not directly connected to the participants. The two groups had no contact after the fixed schedule. To minimize bias, the data collector was unaware of the assignment of participants to the study groups.

Diets

The energy requirement for each participant was calculated based on the World Health Organization (WHO) equations⁴⁸. From the calculated amount, 250–500 kilocalories were then subtracted to lose 0.25–0.5 kg per week⁴⁹. In both groups, the diet was designed so that carbohydrates made up 55–55%, proteins 15–20% and fats 30% of the total calories, which were distributed over six meals and snacks. In the control group, only calorie restriction and the consumption of healthy foods were prescribed, and the diet did not follow a specific dietary pattern. The diet designed for the intervention group was based on the MIND dietary pattern developed by Morris et al. (Supplemental Table 1)³⁴ with some modification due to the religious beliefs as the consumption of wine was removed from this dietary pattern. For participants in the MIND diet group, meals and snacks of each day were designed weekly according to the diet’s guidelines. If the person was unable to prepare every menu provided in the diet, the meal was changed according to the dietitian’s opinion. In both groups, calorie intake was restricted so that both groups could benefit from weight loss in accordance with ethical principles. To encourage adherence to the diet, participants’ diets were assessed weekly by telephone along with a monthly face-to-face meeting. The approximate time of consultation with a dietitian was 90 min for each person in each session, and thereafter the participants communicated with the dietitian whenever they needed an advice. All subjects were instructed to maintain their current level of physical activity throughout the study.

The MIND diet score

The subjects were instructed to precisely record their food intake for three days (two days during the week and one day on the weekend) at baseline and 12th weeks. Dietary data were then analyzed using Nutritionist IV software modified for Iranian foods (First Databank, San Bruno, CA, USA, version 3.5.2.). The values are considered on the basis of servings per 1000 kcal, and the MIND diet score was calculated from 14 points, as wine consumption was not taken into account³³. The MIND diet score was also considered to determine the level of dietary adherence. The scores above 80% indicated adherence to the diet.

Primary endpoint

Assessment of sleep quality

Sleep quality was assessed using the 19-question Pittsburg Sleep Quality Index (PSQI), which examines 7 different aspects of sleep. The validity and reliability of PSQI were established in the Persian version by Farrahi et al.⁵⁰. This index included sleep quality, time to fall asleep, sleep duration, sleep efficiency, sleep disturbances, sleep medication use, and daytime dysfunction. The scores ranged from 0 to 21, and the higher scores indicating a poorer sleep quality. This index was examined at the beginning and end of the study.

Secondary endpoint

Anthropometric measurements, the assessment of mental health, and serum levels of BDNF and cortisol

The participants’ weight, height, waist and hip circumference were measured at the beginning and end of the study using the calibrated equipment. The BMI was determined by dividing the weight in kilograms by the square of height in meters. The waist circumference was measured in a standing position at the narrowest point between the lowest rib and the iliac crest, whereas the hip circumference was measured as the greatest gluteal circumference. Waist-to-hip ratio (WHR) and waist-to-height ratio (WHtR) were also calculated⁵¹.

Depression, anxiety and stress were assessed using the Depression Anxiety Stress Scale-21 (DASS-21) questionnaire at the beginning and end of the study. The validity and reliability of questionnaire were confirmed by Sahebi et al.⁵². Each of the subscales comprised 7 questions, and each question was scored from 0 to 3. High scores on each subscale indicated higher levels of depression, anxiety, and stress.

After 10–12 h of fasting, 8 cc of blood samples was taken from all subjects at the beginning and end of intervention to measure serum levels of BDNF (ZellBio-Kit; Germany) and cortisol (Roche-Kit; Germany) using the ELISA and ECLIA method, respectively. All experiments were performed in accordance with relevant named guidelines and regulations.

Statistical analysis

The data was analyzed using SPSS software version 21 and STATA software version 1. Results were expressed as mean ± SD or frequency (percentage). The Kolmogorov-Smirnov test was first used to check the normality of the variables. Possible differences at baseline among treatment groups were assessed by independent sample t test for normally distributed parameters and the Mann–Whitney test for nonparametric data. At the end of intervention, the linear regression model was used to determine the relationship between the independent and dependent variables and to adjust for confounding variables. Moreover, to assess within-group differences of biochemical parameters, paired-sample t test and the nonparametric Wilcoxon signed-rank test were employed. The qualitative data was assessed by the chi-square test. The significance level was set at p < 0.05.

Ethics approval

This randomized clinical trial was approved by the Research Ethics Board of Kermanshah University of Medical Sciences (IR.KUMS.REC.1402.170) and was approved by the Iranian Registry of Clinical Trials (IRCT20181111041611N8) on 07/08/2023. In addition, in accordance with the Declaration of Helsinki Law, all subjects read and signed a written informed consent form before participating in the study.

Results

Participant characteristics

This study was conducted in October 2023 and lasted 6 months. All 44 participants who met the inclusion criteria completed the study, as shown in Fig. 1. Participants in the study reported no adverse effects from the intervention. The baseline characteristics of the participants are shown in Table 1. At the beginning of study, there was no significant difference between groups in the aforementioned parameters (p > 0.05).

Fig. 1.

Flowchart illustrating the CONSORT methodology used in this study. LCD, low-calorie diet; MLCD, MIND low-calorie diet.

Table 1.

Baseline characteristics of study participants.

Variables	MIND diet group (n = 22) Mean (SD) or n (%)	Control group (n = 22) Mean (SD) or n (%)
Age; years	55.36 (7.44)	54.68 (7.68)
Marital status
Married	18 (81.8)	20 (90.9)
Single	4 (18.2)	2 (9.1)
Education
Under diploma	17 (77.3)	19 (86.4)
College’s degree	5 (22.7)	3 (13.6)
Weight (kg)	78.00 (9.29)	79.05 (10.47)
BMI (kg/m2)	30.94 (3.52)	31.31 (3.79)
Waist circumference (cm)	101.76 (8.87)	104.62 (5.97)
Drugs use
Blood glucose lowering agents	22 (100)	22 (100)
Lipid lowering agents	16 (73.0)	14 (63.6)
Antihypertensive agents	17 (77.0)	16 (73.0)
Duration of diabetes; years	6.83 (4.93)	7.44 (5.48)
FBS (mg/dl)	146.77 (41.97)	143.32 (41.67)
HbA1C (%)	7.64 (0.93)	7.60 (1.24)
Insulin (µU/ml)	15.21 (4.76)	16.09 (12.44)
ISI (score)	13.45 (4.08)	14.36 (4.63)

BMI, body mass index; FBS, fasting blood sugar; HbA₁C, hemoglobin A₁C; ISI, insomnia severity index.

Changes in food intake

Table 2 shows the changes of calorie intake and other nutrients. As expected, the amount of calories consumed decreased in both groups and the protein intake increased (p < 0.001). In addition, the amount of saturated fatty acids intake decreased (p = 0.012) and the amount of fiber intake increased (p = 0.007) significantly in the MIND group compared to the control group. These results were maintained statistically significant even after adjustment for the baseline values (p < 0.001).

Table 2.

The differences in nutrient intake of the MIND diet group compared to the control group during the study.

Variables	MIND diet group (n = 22) Mean (SD)	Control group (n = 22) Mean (SD)	Mean difference (95% CI)	P-value
Energy (Kcal/d)
Baseline	1959.91 (363.25)	1957.58 (494.66)	− 2.33 (− 266.38, 261.72)	0.588**
End of trial	1602.30 (169.54)	1535.80 (220.42)		0.228†
P-value	< 0.001*	< 0.001*
Carbohydrate (%kcal)
Baseline	57.57 (6.49)	56.88 (4.66)	− 0.70 (− 4.13, 2.74)	0.892**
End of trial	56.04 (1.85)	55.10 (3.44)		0.294†
P-value	0.291*	0.114*
Protein (%kcal)
Baseline	15.41 (1.67)	15.45 (2.05)	0.04 (− 1.10, 1.17)	0.710**
End of trial	18.50 (1.10)	18.32 (1.73)		0.648†
P-value	< 0.001*	< 0.001*
Fat (%kcal)
Baseline	29.68 (6.21)	30.09 (3.04)	0.41 (− 2.60, 3.42)	0.895**
End of trial	29.32 (2.02)	29.51 (2.80)		0.794†
P-value	0.806*	0.519*
Saturated fatty acid (%kcal)
Baseline	9.16 (2.32)	10.08 (1.97)	0.92 (− 0.39, 2.23)	0.012**
End of trial	6.80 (0.90)	9.48 (1.61)		< 0.001†
P-value	< 0.001*	0.224*
Monounsaturated fatty acid (%kcal)
Baseline	10.10 (2.37)	9.95 (1.00)	− 0.16 (− 1.28, 0.97)	0.656**
End of trial	10.42 (1.22)	9.95 (1.07)		0.171†
P-value	0.611*	0.994*
Fiber (gr/1000 kcal)
Baseline	16.54 (5.09)	17.93 (5.88)	1.39 (− 1.96, 4.74)	0.007**
End of trial	22.69 (2.88)	19.52 (2.89)		< 0.001†
P-value	< 0.001*	0.207*

*Obtained from paired sample t-test. **Obtained from independent sample t-test. ^†Obtained from linear regression after adjusted for baseline variables.

According to Table 3, as expected, the intake of green leafy vegetables, berries, nuts, olive oil, fish, beans and whole grains increased significantly and the intake of cheese decreased significantly in the MIND group compared to the control group after adjusting for baseline values (p < 0.05). There was no significant difference in changes of other vegetables, poultry, red meat, butter, fast fried foods and sweets intake between groups (p > 0.05). At the end of intervention as expected, the MIND diet group had an average score significantly higher than control group as 12.82 ± 0.59 vs. 8.39 ± 1.24.

Table 3.

The differences in dietary intake of the MIND diet group compared to the control group during the study.

Variables	MIND diet group (n = 22) Mean (SD)	Control group (n = 22) Mean (SD)	Mean difference (95% CI)	P-value
Green leafy vegetables (serving/week)a
Baseline	2.26 (1.35)	2.15 (2.18)	− 0.11 (− 1.21, 0.99)	< 0.001**
Follow up	7.07 (2.37)	3.22 (1.16)		< 0.001†
P-value	< 0.001*	0.023*
Other vegetables (serving/week)b
Baseline	9.06 (4.24)	8.49 (5.44)	− 0.58 (− 3.54, 2.39)	0.152**
Follow up	12.51 (3.87)	9.61 (6.56)		0.081†
P-value	0.007*	0.318*
Berries (serving/week)c
Baseline	0.46 (0.64)	0.30 (0.52)	− 0.16 (− 0.52, 0.19)	< 0.001**
Follow up	6.95 (4.12)	0.57 (0.88)		< 0.001†
P-value	< 0.001*	0.072*
Nuts (serving/week)
Baseline	2.62 (2.23)	2.41 (2.27)	− 0.21 (− 1.58, 1.16)	< 0.001**
Follow up	8.32 (3.06)	3.49 (2.97)		< 0.001†
P-value	< 0.001*	0.004*
Olive oil (primary oil)
Baseline	0	0		< 0.001**
Follow up	1	0		< 0.001†
P-value	< 0.001*	1*
Butter, margarine (table spoon/day)
Baseline	0.20 (0.16)	0.32 (0.33)	0.11 (− 0.05, 0.27)	0.665**
Follow up	0.04 (0.08)	0.13 (0.20)		0.257†
P-value	< 0.001*	0.003*
Cheese (servings/week)
Baseline	1.49 (0.94)	1.66 (1.16)	0.17 (− 0.47, 0.82)	< 0.001**
Follow up	0.42 (0.30)	2.20 (1.44)		< 0.001†
P-value	< 0.001*	0.043*
Whole grains (serving/day)
Baseline	2.15 (1.67)	2.08 (1.93)	− 0.07 (− 1.17, 1.02)	0.064**
Follow up	3.30 (0.99)	2.17 (2.18)		0.022†
P-value	0.011*	0.809*
Fish (not fried) (meals/week)
Baseline	0.21 (0.34)	0.21 (0.53)	0.00 (− 0.27, 0.27)	< 0.001**
Follow up	1.16 (0.83)	0.31 (0.53)		< 0.001†
P-value	< 0.001*	0.249*
Beans (meal/week)d
Baseline	1.00 (0.59)	1.10 (0.70)	0.10 (− 0.29,0.49)	< 0.001**
Follow up	2.93 (0.92)	1.65 (0.87)		< 0.001†
P-value	< 0.001*	< 0.001*
Poultry (not fried) (meal/week)e
Baseline	2.27 (1.03)	1.88 (1.00)	− 0.39 (− 1.00, 0.23)	0.871**
Follow up	3.06 (0.84)	2.63 (0.97)		0.301†
P-value	0.004*	< 0.001*
Red meat and products (meals/week)
Baseline	0.83 (0.57)	0.57 (0.38)	− 0.26 (− 0.56, 0.03)	0.013**
Follow up	0.44 (0.23)	0.54 (0.41)		0.069†
P-value	0.004*	0.602*
Fast fried foods (times/week)
Baseline	0.15 (0.27)	0.14 (0.43)	− 0.01 (− 0.23, 0.21)	0.405**
Follow up	0.02 (0.07)	0.07 (0.34)		0.263†
P-value	0.053*	0.064*
Pastries and sweets (serving/week)
Baseline	2.57 (1.81)	2.67 (2.75)	0.11 (− 1.31, 1.52)	0.821**
Follow up	2.15 (2.84)	2.11 (2.10)		0.857†
P-value	0.372*	0.230*

^aSpinach, lettuce, greens, collards. ^bCarrots, zucchini/summer squash/eggplant, green/red peppers, green beans, mushroom, garlic, onion, celery, cabbage, cucumber. ^cMulberry, dried berries, blueberry, blackberry, strawberry. ^dBeans, lentils, chickpease, soybeans. ^eChicken, duck, turkey, quail. ^* Obtained from paired sample t-test. ^** Obtained from independent sample t-test. ^† Obtained from linear regression after adjusted for baseline variables.

Changes in sleep quality

As it was shown in Table 4, the sleep quality score improved significantly in both groups after 12 weeks (p < 0.001). Although, the improvements in MIND diet group were statistically significant compared to the control group (β =-2.25; p = 0.006). Our results showed that the improvement in sleep latency (p = 0.011), sleep duration (p = 0.022) and sleep disturbance (p < 0.001) scores was significant in the MIND group vs. control group after adjusting for baseline variables.

Table 4.

Sleep quality scores at baseline and after 12 weeks of intervention.in study groups.

Variables	MIND diet group (n = 22) Mean (SD)	Control group (n = 22) Mean (SD)	Mean difference (95% CI)	P-value	Difference in outcome measures between MIND diet and control groups a
					β (95% CI)	P-value
Subjective sleep quality (score)
Baseline	1.55 (0.67)	1.77 (0.75)	0.23 (− 0.21, 0.66)	0.173**	− 0.29 (− 0.72, 0.14)	0.181¥
End of trial	0.95 (0.49)	1.41 (0.91)		0.074†
P-value	< 0.001*	0.002*
Sleep latency (score)
Baseline	2.36 (0.90)	2.55 (0.60)	0.18 (− 0.29, 0.65)	0.047**	− 0.29 (− 0.84, 0.26)	0.291¥
End of trial	1.36 (0.85)	2.05 (0.79)		0.011†
P-value	< 0.001*	0.002*
Sleep duration (score)
Baseline	1.68 (0.78)	2.09 (0.68)	0.41 (− 0.04, 0.86)	0.296**	− 0.39 (− 0.89, 0.12)	0.129¥
End of trial	0.73 (0.63)	1.36 (0.73)		0.022†
P-value	< 0.001*	< 0.001*
Sleep efficiency (score)
Baseline	0.91 (1.19)	1.14 (1.25)	0.23 (− 0.51, 0.97)	0.371£	− 0.15 (− 0.65, 0.34)	0.537
End of trial	0.41 (0.67)	0.86 (1.08)		0.359†
P-value	0.019€	0.201€
Sleep disturbance (score)
Baseline	1.45 (0.51)	1.55 (0.51)	0.09 (− 0.22, 0.40)	0.001**	− 0.59(− 0.90, − 0.28)	0.001¥
End of trial	0.91 (0.29)	1.50 (0.51)		< 0.001†
P-value	< 0.001*	0.329*
Use of sleep medication (score)
Baseline	0.27 (0.77)	0.09 (0.43)	− 0.18 (− 0.56, 0.20)	0.490**	0.08 (− 0.06, 0.22)	0.242¥
End of trial	0.05 (0.21)	0.00 (0.00)		0.302†
P-value	0.204*	0.329*
Daytime dysfunction (score)
Baseline	1.55 (1.01)	1.95 (0.95)	0.41 (− 0.19, 1.01)	0.708**	− 0.41 (− 1.00, 0.18)	0.163¥
End of trial	1.05 (0.90)	1.55 (0.91)		0.241†
P-value	0.013*	0.016*
PSQI score
Baseline	9.68 (3.68)	11.14 (3.40)	1.45 (− 0.70, 3.61)	0.026**	− 2.25 (− 3.79, − 0.71)	0.006¥
End of trial	5.45 (1.87)	8.73 (3.10)		< 0.001†
P-value	< 0.001*	< 0.001*

^a “Outcome measures” refers to the change in values of measures of interest between baseline and week 12. β [difference in the mean outcomes measures between intervention groups (MIND diet group = 1 and control group = 0)]. *Obtained from paired sample t-test. **Obtained from independent sample t-test. ^€ Obtained from Wilcoxon test. ^£ Obtained from Mann-Whitney test. ^† Obtained from linear regression after adjusted for baseline variables. ^¥ Obtained from linear regression after adjusted for baseline values, age, education, physical activity, weight reduction, waist circumference reduction, protein intake percentage, carbohydrate intake percentage, fat intake percentage. Abbreviations: PSQI, pittsburgh sleep quality index.

Changes in mental health

The mean ± SD of the DASS scores are shown in Table 5. Significant improvements in total DASS score and depression and anxiety subscales’ score were observed in both groups (p < 0.05), while stress scores decreased significantly only in the MIND diet group (p = 0.001). Changes of depression (p = 0.03) and anxiety (p = 0.008) scores were also significant in MIND group vs. control group after adjusting the baseline variables. However, after adjusting for other variables including baseline values, age, education, physical activity, weight changes and calorie intake these results remained significant only for depression (β =-2.64; p = 0.033).

Table 5.

Mental health scores of study groups at baseline and after 12 weeks of intervention.

Variables	MIND Diet Group (n = 22) Mean (SD)	Control Group (n = 22) Mean (SD)	Mean difference (95% CI)	P-value	Difference in outcome measures between MIND diet and control groups a
					β (95% CI)	P-value
Depression (score)
Baseline	9.91 (10.98)	12.73 (9.96)	2.82 (− 3.56, 9.20)	0.276 £	− 2.64 (− 5.05, − 0.23)	0.033¥
End of trial	6.27 (7.64)	10.27 (8.10)		0.030†
P-value	0.001 €	0.033 €
Anxiety (score)
Baseline	9.91 (9.83)	10.73 (6.10)	0.82 (− 4.16, 5.79)	0.155**	− 1.46 (− 3.91, 0.98)	0.233¥
End of trial	4.55 (5.49)	7.82 (4.49)		0.008†
P-value	0.001*	0.006*
Stress (score)
Baseline	15.64 (10.67)	17.82 (9.26)	2.18 (− 3.89, 8.26)	0.242**	− 1.31 (− 5.77, 3.14)	0.553¥
End of trial	10.64 (9.25)	15.27 (8.83)		0.095†
P-value	0.001*	0.128*
DASS score
Baseline	35.45 (27.22)	41.27 (16.85)	5.82 (− 8.04, 19.67)	0.062**	− 5.07 (− 11.10, 0.95)	0.096¥
End of trial	21.45 (20.29)	33.36 (15.06)		0.003†
P-value	< 0.001*	0.001*

^a “Outcome measures” refers to the change in values of measures of interest between baseline and week 12. β [difference in the mean outcomes measures between intervention groups (MIND diet group = 1 and control group = 0)]. ^* Obtained from paired sample t-test. ^** Obtained from independent sample t-test. ^€ Obtained from Wilcoxon test. ^£ Obtained from Mann-Whitney test. ^† Obtained from linear regression after adjusted for baseline variables. ^¥ Obtained from linear regression after adjusted for baseline values, age, education, physical activity, weight reduction, waist circumference reduction, calorie intake. Abbreviations: DASS, depression anxiety stress scales.

Changes in anthropometric measurements

As shown in Table 6, although, weight and BMI decreased significantly in both groups after 12 weeks of the intervention, there was not a significant difference between groups (p > 0.05). However, waist circumference (β =-1.73; p = 0.004), WHR (β =-0.02; p = 0.015) and WHtR (β =-0.11; p = 0.005) decreased significantly in the MIND diet group compared to the control group after adjusting for the baseline values, calorie intake, physical activity and body weight.

Table 6.

Comparison of the anthropometric measurements between groups at baseline and after 12 weeks of intervention.

Variables	MIND diet group (n = 22) Mean (SD)	Control group (n = 22) Mean (SD)	Mean difference (95% CI)	P-value	Difference in outcome measures between MIND diet and control groups a
					β (95% CI)	p-value
Weight (kg)
Baseline	78.00 (9.29)	79.05 (10.47)	1.05 (− 4.98, 7.08)	0.118**	− 0.68 (− 1.43, 0.08)	0.079§
End of trial	75.11 (9.29)	76.74 (10.35)		0.112†
P-value	< 0.001*	< 0.001*
BMI (kg/m2)
Baseline	30.94 (3.52)	31.31 (3.79)	0.38 (− 1.85, 2.60)	0.120**	− 0.26 (− 0.55, 0.04)	0.088§
End of trial	29.80 (3.58)	30.40 (3.75)		0.120†
P-value	< 0.001*	< 0.001*
Waist circumference (cm)
Baseline	101.76 (8.87)	104.62 (5.97)	2.86 (− 1.75, 7.46)	0.014**	− 1.73 (− 2.87, − 0.59)	0.004¥
End of trial	98.62 (8.41)	102.88 (5.61)		0.004†
P-value	< 0.001*	< 0.001*
WHR
Baseline	0.95 (0.07)	0.96 (0.07)	0.02 (− 0.02, 0.06)	0.102**	− 0.02 (− 0.04, − 0.004)	0.015¥
End of trial	0.92 (0.06)	0.95 (0.06)		0.029†
P-value	0.010*	< 0.001*
WHtR
Baseline	0.64 (0.07)	0.66 (0.04)	0.02 (− 0.02, 0.05)	0.017**	− 0.11(− 0.02, − 0.004)	0.005¥
End of trial	0.62 (0.06)	0.65 (0.04)		0.004†
P-value	< 0.001*	< 0.001*

^a “Outcome measures” refers to the change in values of measures of interest between baseline and week 12. β [difference in the mean outcomes measures between intervention groups (MIND diet group = 1 and control group = 0)]. ^* Obtained from paired sample t-test. ^** Obtained from independent sample t-test. ^† Obtained from linear regression after adjusted for baseline variables. ^§ Obtained from linear regression after adjusted for baseline values, calorie intake and physical activity. ^¥ Obtained from linear regression after adjusted for baseline values, calorie intake, physical activity and weight. Abbreviations: BMI, body mass index; WHR, waist-to-hip ratio; WHtR, waist to height ratio.

Changes in biochemical indices

The mean ± SD of serum levels of BDNF and cortisol are shown in Table 7. In the MIND group, the increase in BDNF (p = 0.004) and the decrease in cortisol (p = 0.022) were significant after 12 weeks of intervention compared to the control group. After adjusting for the baseline values, age, physical activity, weight loss, calorie intake, and duration of diabetes, these results also remained significant: β for BDNF = 0.37 (p = 0.004) and β for cortisol=-81.58 (p = 0.009).

Table 7.

Comparison of the biochemical tests between groups at baseline and after 12 weeks of intervention.

Variables	MIND diet group (n = 22) Mean (SD)	Control group (n = 22) Mean (SD)	Mean difference (95% CI)	P-value	Difference in outcome measures between MIND diet and control groups a
					β (95% CI)	P-value
BDNF (pg/ml)
Baseline	1.16 (0.12)	1.18 (0.11)	0.01 (− 0.06, 0.08)	0.003**	0.37 (0.13, 0.61)	0.004¥
End of trial	1.67 (0.46)	1.33 (0.27)		0.004†
P-value	< 0.001*	0.009*
Cortisol (nmol/L)
Baseline	560.90 (169.83)	593.41 (216.52)	32.51 (− 85.88, 150.91)	0.251**	− 81.58 (− 141.14, − 22.01)	0.009¥
End of trial	416.82 (125.60)	498.45 (128.77)		0.022†
P-value	< 0.001*	0.009*

^a “Outcome measures” refers to the change in values of measures of interest between baseline and week 12. β [difference in the mean outcomes measures between intervention groups (MIND diet group = 1 and control group = 0)]. *Obtained from paired sample t-test. **Obtained from independent sample t-test. ^† Obtained from linear regression after adjusted for baseline variables. ^¥ Obtained from linear regression after adjusted for baseline values, age, physical activity, weight reduction, calorie intake, and duration of diabetes. Abbreviations: BDNF, brain-derived neurotrophic factor.

Discussion

This randomized clinical trial was the first to examine the effects of 12 weeks of adherence to the MIND diet on sleep quality, mental health, and some biochemical and anthropometric indices in type 2 diabetic women with insomnia. Our results showed that following the MIND diet significantly improved the sleep quality and depression as compared to the control group. Moreover, a significant decrease of waist circumference and serum levels of cortisol, and increase of BDNF levels were also revealed vs. control group.

At the beginning of the study, the demographic and anthropometric characteristics of the participants showed no significant differences. There were also no significant differences between the groups in terms of initial calorie intake, macronutrient distribution and fiber consumption. The study design ensured that both groups received a low-calorie diet, and the subsequent analyses showed that there were no statistically significant differences between groups in calorie intake or the percentage of protein, fat and carbohydrate intake after the intervention. However, the MIND diet group exhibited a significant reduction in the percentage of saturated fat and a significant increase in fiber intake.

As it was expected, the MIND diet group showed significant modifications in their dietary habits as the increased intake of green leafy vegetables, berries, nuts, olive oil, whole grains, fish, beans and poultry as well as decreased intake of butter, cheese and red meat. These alterations were reflected a significant improvement in MIND dietary scores in the MIND diet group compared to the control group.

Calorie restriction was implemented in both groups, which led to a significant improvement in anthropometric measurements. Especially, the MIND diet group exhibited a more significant reduction in waist circumference, WHR and WHtR compared to the control group. Previous research has demonstrated an association between adherence to the MIND diet and a reduction in the odds of general obesity^53,54. Specifically, Arjmand et al. investigated the effects of MIND diet on weight loss, BMI, waist circumference and body fat percentage in overweight /obese women and observed significant improvements in these outcomes⁵⁵. However, not all studies have consistently shown this relationship⁵⁶. The disparity in results of studies may be due to different study designs, characteristics of subjects including age, health status, and the sample size. The MIND diet is characterized by the high content of polyphenols, which have been contributed to weight management through several mechanisms like the inhibition of digestive enzymes, regulation of lipid metabolism, modulation of gut microbiota, inhibition of adipocyte differentiation, suppression of appetite and stimulation of energy expenditure⁵⁷.

Our participants were type 2 diabetics women suffered from the insomnia and impaired sleep quality. Following the MIND diet significantly improved sleep quality scores especially the sleep latency, sleep duration and sleep disturbance. These results suggest that the MIND diet may be a valuable adjunct therapy for improving sleep quality in diabetic patients with insomnia. This finding is supported by previous studies that have demonstrated a positive association between the MIND diet and the Mediterranean diet with improved sleep quality^42,58. In addition, the results of studies included in the systematic review indicated a negative association between Dietary Inflammatory Index (DII) scores and sleep outcomes⁵⁹.

It seems that some components of MIND diet contribute to the improvement of sleep quality. In particular, the high proportion of green leafy vegetables in the MIND diet plays a crucial role in this regard. Green leafy vegetables are rich source of magnesium, an essential mineral involved in the nerve function and the regulation of sleep quality^60,61. Magnesium acts as a cofactor for the enzyme N-acetyltransferase in the melatonin synthesis pathway, which is critical for the regulation of sleep-wake rhythm⁶². Moreover, magnesium acts as an N-methyl-D-aspartate (NMDA) antagonist and GABA agonist and modulates sleep behavior through its effect on neurotransmitter activity^63,64. Another important component of the MIND diet are berries. Berries are a rich source of melatonin and flavonoids, which have been shown to play an important role in sleep regulation. Flavonoids have been shown to reduce inflammation and oxidative stress in neurons, protect the blood-brain barrier (BBB) from inflammatory cells, and improve endothelial function and blood pressure control^65–69. The cumulative effects of these compounds may contribute to the observed improvement in sleep quality.

The presence of anxiety and depression in these patients may also be attributed to the concomitant sleep disturbances^3,70. In the present study, the MIND diet intervention resulted in a statistically significant reduction in the level of anxiety and depression. In consistent with this finding, a previous study by Daneshzad et al. showed that following the DASH diet for 3 months in diabetic patients caused a significant reduction in anxiety and depression²⁸. Another study by Bayes et al. showed the positive effects of the Mediterranean diet on reducing depression⁷¹. In another study, a significant association was found between adherence to the MIND dietary pattern and improved quality of life and a reduction in depression symptoms in adolescent girls⁷². The observed effects of MIND diet on mental health may be mediated by its influence on several physiological pathways, including inflammation, oxidative stress, HPA axis activity, gut microbiota, tryptophan-kynurenine metabolism, and BDNF levels⁷³.

In fact, adherence to a healthy diet such as the MIND diet can regulate cortisol secretion by modulating HPA axis function, modulating the gut-brain axis, and reducing oxidative stress, which can directly enhance BDNF secretion⁷³. Consistent with this hypothesis, our study found that a 12-week intervention with the MIND diet in diabetic patients with insomnia resulted in a significant reduction of the cortisol levels. Furthermore, we observed a significant increase in BDNF levels, an important neurotrophic factor involved in neuronal survival and differentiation. Our findings are also consistent with those of Sánchez-Villegas et al. which reported a statistically significant increase in BDNF levels in patients with depression under the Mediterranean diet plus nuts (Med + Nuts) group compared to the control group⁷⁴. In addition, in a previous study by Arjmand et al. a 3-month intervention with the MIND diet resulted in an increase in BDNF levels, although this increase was not significant⁷⁵. This may be attributed to differences in adherence levels to the dietary pattern among participants, as well as specific characteristics of the individuals involved in the study.

Fig. 2 represents some default mechanisms of MIND diet for improving sleep quality and mental health.

Fig. 2.

Mechanisms by which the MIND diet impacts sleep quality and mental health. Abbreviations: BDNF, brain-derived neurotrophic factor; HPA, hypothalamic-pituitary-adrenal.

Overall, to the best of our knowledge, this study was the first RCT to examine the effects of the MIND diet on sleep quality, mental health and some anthropometric and biochemical indices in type 2 diabetic patients with insomnia. However, the notable limitations of the present study were the relatively small sample size, and short duration of study as well as being the sex specific (women), which may affect the generalizability of the results. The fourth limitation of this study is that subjective measures were used to assess sleep quality rather than using objective tools such as polysomnography (PSG) and actigraphy, which are considered the gold standard for sleep assessment. Finally, the self-reported dietary intake assessment may be also prone to bias.

Conclusion

In conclusion, adherence to the MIND diet resulted in significant improvement in sleep quality and mental health in type 2 diabetic women with insomnia. Moreover, the MIND diet group exhibited a significant increase in serum BDNF levels and a significant decrease in cortisol levels and measurements of waist circumference, WHR and WHR after a 12-week intervention compared to the control group. These results suggest a promising therapeutic potential of the MIND diet in T2DM. However, future well-designed studies with larger sample sizes and objective measurement tools such as PSG and actigraphy are needed to fully clarify the effects of the MIND diet on these variables. Furthermore, the assessment of serum levels of melatonin, serotonin, tryptophan and kynurenine is recommended to better understand the underlying mechanisms of action.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

M.G.: Conceptualization, Data Curation, Methodology, Formal analysis, Writing. V.E.A.: Review and editing. Y.S.: Data Curation, Methodology. L.S.: Acquisition of Data. M.S.: Conceptualization, Supervising, Review and editing. S.M.N.: Conceptualization, Supervising, Review and editing. All authors reviewed the manuscript.

Funding

This trial was financially supported by Kermanshah University of Medical Sciences, Kermanshah, Iran (Grant numbers: 4020342).

Data availability

The data associated with the paper are not publicly available but are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.