The MIND diet, cognitive function, and well-being among healthy adults at midlife: a randomised feasibility trial

Abstract

Background

The Mediterranean and Dietary Approaches to Stop Hypertension (DASH) diets have been shown to slow cognitive decline. However, these diets were not originally developed with dementia prevention as their primary focus. In contrast, the Mediterranean-DASH Intervention for Neurodegenerative Delay (MIND) diet was specifically designed based on evidence linking individual dietary components to brain health and the prevention of cognitive decline. The aim of this research was to test the effectiveness of the MIND diet on cognitive function, mood, and quality of life using the Behaviour Change Wheel, and to design an intervention using the COM-B model to promote the MIND diet at midlife.

Methods

An online pilot randomised control trial (RCT) was conducted. Forty-one participants (aged 40–55 years, male and female) were randomised into a MIND diet with support group (n = 15), MIND diet group with no support (n = 14) or control group (n = 12) for 12 weeks. Baseline and follow-up measures of cognitive function, mood, quality of life (QOL) and adherence to MIND diet was assessed in each group. Capability, opportunity, and motivation (COM-B) towards MIND diet behaviour were also assessed pre and post intervention.

Results

A repeated measures ANOVA showed that in comparison to the control group, both intervention groups significantly improved mood, quality of life, MIND diet score and all COM-B components at follow-up (p < 0.05). No significant differences or interactions in cognitive function were found between groups.

Conclusion

There are only two RCT’s that test the effectiveness of the MIND diet on cognitive function, and this is the first RCT to assess the usefulness of the COM-B in increasing adherence to the MIND diet. Future interventions with longer duration are needed to establish an association with MIND diet and cognitive function in adults at midlife. This study recommends using less Behaviour Change Techniques (BCTs) with a focus on self-monitoring, goal setting and education on diet as an effective strategy for promoting adherence to the MIND diet.

Trial registration

Trial registered at ClinicalTrials.gov Protocol Record NCT04654936, May 2019.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40795-025-01020-6.

Background

Adopting healthy dietary behaviours are widely recognised as a key factor in improving population health and reducing the risk of non-communicable diseases (NCDs) such as cardiovascular disease, diabetes, and dementia. The most recent statistics show that around 850,000 people in the UK are living with dementia, and this number is projected to rise to 1 million by 2025 due to increased longevity, as well as a combination of factors such as genetic predisposition, lifestyle choices, and comorbidities including hypertension, diabetes, and obesity [1]. This underscores the importance of addressing modifiable lifestyle factors—such as physical activity, smoking cessation, cognitive engagement, and dieting promoting healthy ageing. Among these, dietary patterns like the Mediterranean diet (Med diet), DASH, and MIND diets are of particular interest for their potential neuroprotective effects, as they have been associated with reduced inflammation, improved vascular health, and enhanced cognitive function [2].

Recent research has shown that the MIND diet may slow cognitive decline, particularly in older adults. One study found that the MIND diet was more predictive of cognitive decline over time than the Mediterranean or DASH diets [3]. Additionally, an RCT on the MIND diet found significant improvements in cognitive function in obese middle-aged adults after just three months. MRI scans showed increased surface area in the inferior frontal gyrus in the MIND diet group, as well as decreases in cerebellum white and grey matter, with the MIND diet group showing greater reductions than the control group. Furthermore, longer adherence to the MIND diet was associated with better verbal memory and a lower risk of cognitive decline [4–8].

While the Mediterranean and DASH diets have shown positive effects on cognitive health, they are not specifically designed for this purpose. The MIND diet, however, is a hybrid of both the Mediterranean and DASH diets, with components tailored to promote cognitive health [9–11]. The MIND diet recommends limiting saturated fat, which is detrimental to brain health, and increasing intake of leafy greens and berries, which have been shown to have the strongest associations with slowing cognitive decline [3, 12–14].

However, there is a significant gap in the literature regarding the MIND diet, as most studies to date have been cross-sectional or longitudinal and have focused primarily on elderly populations. This study aims to address this gap by targeting mid-life adults, a population where dietary interventions may have the most potential to mitigate cognitive decline. Research suggests that cognitive function peaks as early as the 20 s or 30 s, followed by a gradual decline [15]. Preventive measures and lifestyle modifications can reduce risk factors even before the preclinical stage, and interventions targeting cognitive health may yield significant benefits when implemented early—potentially in the preclinical phase of cognitive decline [16]. This period presents a critical window for preventive measures or early intervention strategies to slow or prevent cognitive deterioration.

To date, only two randomised controlled trial (RCT) has explored the causal relationship between the MIND diet and cognitive decline, highlighting a notable evidence gap with substantial potential to generate new knowledge. This study aims to evaluate not only the cognitive outcomes of the MIND diet but also its broader impact on quality of life and well-being. Secondary measures of mood and quality of life were included in this study. Previous research has demonstrated that dietary patterns such as the Mediterranean and MIND diets are associated with improved mood, enhanced self-reported functional health, and reduced risk of depression [17–20]. These findings underscore the broader implications of such dietary interventions, highlighting their potential to support both mental and physical health.

Theoretical framework

Moderate adherence to the MIND diet has been reported to be beneficial to cognitive health, but in order to promote the MIND diet more effectively, and to change behaviour, we need to understand the underlying mechanisms of action that impact on the behaviour. This is what the BCW allows us to do within a theoretical framework that can be used to map intervention functions and components to best promote the behaviour (see Fig. 1).

Fig. 1.

Stages involved in the development of an intervention using the Behaviour Change Wheel [21]

However, poor adherence is a common challenge in lifestyle interventions [22] and understanding the factors that influence adherence is key to improving outcomes [23]. To design interventions that are effective, sustainable, and scalable, it is essential to identify which components of the intervention most positively impact the desired outcomes. This can be achieved through a theoretical framework that guides intervention design and evaluation [24].

Behaviour change is a critical component of dietary interventions, as individuals need to alter their eating habits to achieve the benefits of the MIND diet. Psychological theory-based interventions have been shown to be particularly effective in promoting behaviour change [25], with a systematic review indicating that theory-based interventions are more successful in improving dietary behaviours than non-theory-based approaches [26, 27].

In this study, the intervention was informed by previous research by the authors [27], which identified key barriers and facilitators to adhering to the MIND diet across nine domains of the Theoretical Domains Framework (TDF): knowledge, behaviour regulation, skills, memory, attention and decision processes, environmental context and resources, social influence, beliefs about capabilities, beliefs about consequences, and emotion. To address these barriers and enhance facilitators, the Behaviour Change Wheel (BCW) was used as a guide for mapping TDF domains to specific intervention functions and Behaviour Change Techniques (BCTs) (see Fig. 2). These techniques were then embedded into the 12-week online dietary intervention programme to improve adherence and, ultimately, the effectiveness of the MIND diet (see supplementary data 1).

Fig. 2.

TDF domains and corresponding mapping onto the COM-B component [28]

Aim

The aim of this research was to test the effectiveness of the MIND diet on cognitive function, mood and quality of life using the Behaviour Change Wheel.

We hypothesise that.

1. By promoting capability, opportunity and motivation, we will enhance adherence to a 12-week MIND diet intervention in 40–55-year-old community dwelling adults.

2. A 12-week MIND diet intervention will improve cognitive function, mood and quality of life in 40–55-year-old community dwelling adults.

3. Adherence to the MIND diet is expected to be higher in the intervention group receiving face to face and web-based support, compared to intervention group with no support (information only) and control (Eat well plate).

4. Adherence to the MIND diet will improve nutritional intake over the 12-week intervention period in 40 -55-year-old community dwelling adults.

Methods

Design

The study employed a randomised controlled prospective follow-up study, employing a 3 (condition) X 2 (time) factorial design. Outcome measures were recorded for MIND diet score, cognitive function, everyday mood, quality of life and COM-B components for all groups. This study follows the Consolidated Standards of Reporting Trials (CONSORT) guidelines (see supplementary data 2).

Participants, recruitment, and randomization

The software program G*Power 3 [29] was used to conduct a power analysis to determine what sample size would be sufficient to detect a difference between groups. The effect size chosen was small 0.28 and based on previous research on nutrition education and dietary intake [30]. With power set at 0.80, and alpha = 0.05, G*Power indicated that a sample size of 36 would be required. Forty-one participants took part in the study (see Fig. 3 for the flow of participants). Participants were healthy males and females aged between 40–55 years old living in Northern Ireland. An advertisement was posted around the local communities for interested people to contact the researcher. Also, a notice was posted on social media and university global email. Booklets were handed out in local GP surgeries, GAA community centers, over 50’s club, schools, community activities such as dance classes and library. All interested participants were asked to contact the researcher by email, and dates and times were arranged for baseline data collection at participant’s home. Participants were emailed a participant information sheet (PIS) to gain further information on the study. All participants were asked for consent to be contacted by text for the duration of the study, for reminders to complete food charts and diaries. Participants were randomised into one of three groups (MIND diet with support, MDWS; MIND diet with no support, MDNS; control) by a JavaScript random number allocation and allocated by an external researcher [31]. Ethical approval was obtained from the School of Psychology Staff and Postgraduate Filter Committee at Ulster University, which is in accordance with The Code of Ethics of the World Medical Association (Declaration of Helsinki).

Fig. 3.

CONSORT flow chart: Enrollment, allocation of participants, follow up and analysis

Inclusion criteria

Healthy males and females aged 40–55 years old living in Northern Ireland.

Exclusion criteria

Anyone following a specific dietary pattern (veganism, vegetarian, Atkins). Anyone on a specific diet recommended by their GP (eating disorder), specific illnesses such as high cholesterol, diabetes, heart disease, dementia. Participants were screened for any underlying neurological or cognitive conditions that would exclude them from the study. Specifically, individuals were asked two exclusionary questions as part of the intake process:

1. “Do you have any diagnosed neurological or cognitive conditions (e.g., dementia, memory impairment)?”

2. “In the past year, have you experienced difficulties with memory or concentration that interfere with daily functioning?”

Participants who reported any positive responses to these questions were excluded from the study. It is important to note that no formal cognitive screening tool, such as the Mini-Mental State Examination (MMSE), was used to assess for conditions like dementia. Instead, cognitive health was assessed through self-reported symptoms.

Design and content of intervention

Design

In order to develop the MIND diet intervention, the factors identified from previous research [27], to promote the MIND diet, were then mapped onto the subsequent steps of the BCW (see Fig. 1). First, the factors were mapped onto intervention functions the BCW suggested are most likely to be effective in facilitating behaviour change. There are nine functions to choose from: education, persuasion, incentivisation, coercion, training, restriction, environmental restructuring, modelling, and enablement [21]. The BCW recommends evaluating intervention functions through the APPEASE (Affordability, Practicability, Effectiveness and cost-effectiveness, Acceptability, Side-effects/safety, and Equity) criteria [21], to assist in the decision making regarding the intervention delivery and content, and to make judgements on what would be most appropriate for the intervention. This allowed for the identification of appropriate and suitable intervention functions within the context of this RCT. As all COM-B components have been highlighted as potential to change behaviour, all intervention functions have potential for intervention design. However, after evaluation through the APPEASE criteria, we found that 6 of the 9 intervention functions were most relevant to take forward into the intervention. These were: education (increasing knowledge), training (imparting skills), persuasion (influencing attitudes and actions), modelling (using examples to inspire people), environmental restructure (changing physical or social environment), enablement (providing support to overcome barriers). These intervention functions were considered most likely to be effective to change behaviour in the target population.

To identify the intervention content, we determined which BCT’s would bring about change in the target behaviour. Two researchers (DT, EEAS) used the behaviour change technique taxonomy (BCTTv1) [21] to identify the most likely BCTs to bring about behaviour change and mapped these onto the TDF identified from the COM-B analysis in stage 1 and the 6 selected intervention functions. Eighteen BCT’s were identified for inclusion in the dietary intervention. Identifying BCT’s for inclusion was also based on the APEASE criteria recommended by the BCW [21] and the theory and techniques tool [32], which explores the link between BCT’s and mechanisms of action (MO).

Intervention group one: MIND diet with support (n = 15)

Over the 12 weeks of the intervention, each week addressed the barriers and facilitators identified in the COM-B analysis [27] by delivering tips and advice on how to adhere to the MIND diet in different situations such as the workplace, eating out and cooking for family and friends. Participants also received a self-monitoring chart to track their daily consumption of the MIND diet. A range of resources were developed to facilitate behaviour change in the target population and optimise delivery of the BCTs which included written educational material on the MIND diet elements and the benefits of the diet, alongside recipes, and peer support. To allow participants to engage and interact with each other and the researcher in the online environment, a chat room/forum was available on the website for participants to log in to, allowing the researcher to engage and monitor participants throughout the intervention period. This chat room was for social support, which is one of the BCTs identified by the BCW as possibly effective for behaviour change. Also, text messages were sent on a weekly basis over the 12 weeks of the intervention to check in with participants progress, which has been found to be effective in promoting adherence to dietary guidelines over 6 months [33, 34].

Intervention group two: MIND diet without support (n = 14)

Participants in the second group of the intervention did not have access to the website. Participants received information on the MIND diet elements, including what foods to eat, how often, and in what portion sizes, based on UK guidelines. Participants also received a self-monitoring chart to track their daily consumption of the MIND diet. The self-monitoring chart and basic information containing food frequency and portion sizes for both intervention groups were in paper form. It was suggested by the researcher that the MIND diet chart be placed in a prominent place to facilitate use, for example the participants refrigerator.

For both intervention groups, the MIND diet chart was developed as a self-monitoring tool to track daily adherence to the MIND diet. The chart was presented as a table which included a list of foods associated with the MIND diet, such as leafy greens, chicken, nuts, presented in rows down the left-hand side of the table. Days ran along the top of the columns. Participants were instructed to complete the chart daily, noting their consumption of MIND diet foods by ticking the relevant box in the table to indicate consumption of that food group on that day. As an aid to memory, they were encouraged to place it in a prominent location, such as on their refrigerator, for easy access. The completed charts were then submitted to the researcher regularly (e.g., weekly or monthly) for adherence monitoring. They were awarded one mark for each food consumed. In addition, the chart was used as part of the treatment fidelity process to ensure that participants were following the prescribed dietary intervention. This enabled the researcher to assess and verify adherence to the MIND diet throughout the study and provide feedback when necessary.

Control group (n = 12)

The control group were asked to follow their usual diet without any support but received basic information on general government dietary guidelines as set out in the ‘Eat Well Guide’[30]. The group were asked to complete all the same measures as the intervention group at baseline and follow up and asked to record their food intake at week 1 and 12.

Materials

Cognitive function

The Cambridge Neuropsychological Test Automated Battery (CANTAB) was chosen for its robust reliability, validated measures, and sensitivity to cognitive changes related to aging and dietary intake. CANTAB tests have demonstrated brain-to-behavior reliability [34], strong test–retest reliability in older adults [35], and construct validity in both clinical and non-clinical populations [36]. Its computerized format ensures standardized administration, reducing biases and allowing for the precise detection of subtle cognitive changes that traditional methods may miss. This makes it particularly suitable for studying early cognitive changes influenced by dietary interventions. Memory tests in the CANTAB activate regions such as the hippocampus, temporal lobes, amygdala, and prefrontal cortex, which are critical for cognitive domains affected by age and diet, including memory, attention, and executive function [37].

CANTAB is widely applied in dietary intervention studies due to its ability to assess multiple cognitive domains and detect nuanced cognitive shifts. For this study, tasks included spatial working memory (SWM), spatial span (SSP), pattern recognition memory (PRM), and reaction time (RTI), with an average administration time of 25 min. Practice components for each task allowed participants to familiarize themselves with the CANTAB tablet. PRM is a two-alternative forced-choice task assessing visual recognition memory. Participants memorized abstract patterns and selected the correct one among distractors. The dependent variable was mean correct latency (ms). In the SWM task participants searched for hidden tokens across multiple boxes, avoiding previously searched locations, to fill up a column on each page. The outcome measure was the total number of search errors. SSP assessed short-term spatial memory, where participants replicated sequences of illuminated boxes. The dependent variable was the maximum span length. In RTI participants responded to visual stimuli under simple and five choice reaction conditions. Mean reaction time (ms) was the dependent variable.

These tasks are well-suited for assessing specific aspects of cognition that may be influenced by dietary interventions over time. They collectively provide a comprehensive evaluation of memory, executive function, processing speed, and attention, making them robust tools for studying cognitive decline.

PANAS-20

Everyday mood was measured in this study with the use of the Positive and Negative Affect Schedule (PANAS), [38]. This scale developed to measure levels of negative affect (e.g. feelings of distress, guilt, displeasure) and positive affect (e.g. levels of enthusiasm and alertness), has been found to show good reliability and validity, [39]. Cronbachs alpha for positive affect range from 0- 86–0.90 and for negative affect 0.84–0.87, showing high reliability. The PANAS scale was administered 4 times a day (on rising in the morning, lunch time, dinner time and before bed) for 4 days a week (2 weekdays and 2 weekend days) at baseline and at follow up, to measure mood overtime.

QOL

WHOQOL BREF: (Whoqol.group) assesses measure of quality of life. It consists of 24 items to assess perception of quality of life in four domains, including physical health, psychological, social relationships, and environment. Cronbach’s alpha values were 0.91 for the overall scale. All values were above 0.70 and thus showed adequate internal consistency. The WHOQOL-BREF is suitable for use in New Zealand with samples from the general population [40].

COM-B questionnaire

Designed by the researcher based on the most common barriers and facilitators derived from stage 1 qualitative study [27]. The questionnaire is based on the COM-B Self-Evaluation Questionnaire V1 [41]. This questionnaire is recommended to use for the intervention development process of the BCW and is applicable to a range of health behaviours and populations. In this study, the COM-B questionnaire was scored using a 5-point Likert scale, where participants rated each of the 24 statements as follows:1 = Strongly Disagree, 2 = Disagree, 3 = Neither Agree nor Disagree, 4 = Agree, 5 = Strongly Agree (see supplementary data 2). For positive statements, a higher score (i.e., “Strongly Agree”) reflects stronger alignment with the desired behaviour or outcome, indicating higher motivation or capability. Conversely, for negative statements, the scale was reverse scored: “Strongly Agree” was assigned 1 point, and “Strongly Disagree” was assigned 5 points. This reverse-scoring ensures that higher scores consistently represent greater alignment with the desired behaviour across all items, regardless of whether the statement was framed positively or negatively.

The scores for each domain (Capability, Opportunity, and Motivation) were calculated by summing the relevant items within each domain. Higher domain scores indicate a stronger alignment with the desired behaviour. This scoring approach is commonly used in behaviour change research [21] to facilitate the interpretation of data and identify areas where interventions may be needed to improve behaviour change outcomes. Cronbach’s Alpha was calculated for each of the COM-B components with multiple questions. Items that affected Cronbach’s alpha negatively were deleted resulting in Cronbach alpha for Capability (0.657), opportunity (0.527) and motivation (0.630). Similar Cronbach’s Alpha for COM-B components were seen in previous research [42].

MIND diet chart

Designed by the researcher to allow participants to record their MIND diet foods daily (see Table 1).

Table 1.

Weekly food chart for self-monitoring MIND diet consumption

Date:	Monday	Tuesday	Wednesday	Thursday	Friday	Saturday	Sunday
Leafy greens (6/wk): Tick the box if you have eaten at least one portion = 100 g
Other vegetables (1/d): Tick the box if you have eaten at least one portion 100 g
Berries (2/wk). Have you eaten berries today? Tick the box if you have = 1 portion = 50 g
Beans and legumes (3/wk). If you have eaten a portion of beans today, tick the box. 1 portion ½ cup
Fish (1/wk). If you have eaten a portion of fish today 75-100 g tick the box
Poultry (2/wk) If you have eaten a portion of poultry 100-150 g today, tick the box
Nuts and seeds (5/wk) If you have eaten a portion of nuts today, tick the box 1 portion 25 g
Whole grains (3/d) 70 g If you have eaten a portion of wholegrains today, tick the box: brown bread/rice/pasta/cereal etc
Olive oil as main oil: If your main cooking oil is olive oil, please tick the box each day
Red meat (max 4/wk). Have you eaten red meat today? If so, tick the box, includes pork
Cheese (max 1/wk). Have you eaten a portion of full fat cheese today 50 g? If so, tick the box 1 portion 1 thumb size
Fast/fried food (max 1/wk); Tick box if you have eaten a portion of fast/fried food. If you have eaten 2 portion, tick box twice. Takeaways, deep fat fried foods, crisps etc
Sweets/pastries (max 5/wk). Tick box if you have eaten a portion of sweets/pastries. If you have eaten 2 portion, tick box twice
Butter (max 1tbsp/d). Tick the box if you have eaten none or less than a tbsp of butter. If you have eaten more than a tbsp mark with an X

d day, wk week

Food diary

All participants were asked to record their food intake for a seven-day period at both baseline and week 12. Types of food as well as brand names were recorded, and portion sizes consumed were estimated using general household measures. Total energy, macronutrient and micronutrient intakes at baseline and week 12 were analysed using the dietary analysis software programme Nutritics [43]. The food diary information was also used to validate the scores recorded on the MIND food chart.

MIND diet score

MIND diet score was computed from the 7-day food diaries recorded on Nutritics [43]. The MIND diet has 15 dietary components including 10 food groups heathy for the brain, (green leafy vegetables, other vegetables, nuts, berries, beans, whole grains, fish, poultry, olive oil and wine) and 5 food groups unhealthy for the brain (red meats, butter and margarine, cheese, pastries and sweets, and fried/fast food). Participants received 1 point if olive oil was identified as their main cooking oil. For all other dietary components, each food group was assigned a score depending on the frequency of consumption (0, 0.5, 1). The total MIND diet score was calculated by summing all the 15 components (see Table 2).

Table 2.

MIND diet component servings and scoring

	Examples	0	0.5	1
Green Leafy Vegetables	Kale, spinach, cabbage	≤ 2 servings/wk	> 2 to < 6/wk	≥ 6 servings/wk
Other Vegetables	Broccoli, carrots, potatoes, Tomatoes, onion,	< 5 serving/wk	5 – < 7 wk	≥ 1 serving/day
Berries	Blueberries, strawberries raspberries	< 1 serving/wk	1 /wk	≥ 2 servings/wk
Nuts	Walnuts, almonds, brazil	< 1/mo	1/mo – < 5/wk	≥ 5 servings/wk
Olive Oil	Olive oil	Not primary oil		Primary oil used
Butter, Margarine	Butter, margarine	> 2 T/d	1–2 /d	< 1 T/d
Cheese	Cheddar,	7 + servings/wk	1–6 /wk	< 1 serving/wk
Whole Grains	Bread, pasta, rice, cereals	< 1 serving/d	1–2 /d	≥ 3 servings/d
Fish (not fried)	Salmon, cod, tuna	Rarely	1–3 /mo	≥ 1 meals/wk
Beans	Kidney, butter, chickpea, black bean	< 1 meal/wk	1–3/wk	> 3 meals/wk
Poultry (not fried)	Chicken, turkey	< 1 meal/wk	1 /wk	≥ 2 meals/wk
Red Meat and products	Red meat, bacon, sausages	7 + meals/wk	4–6 /wk	< 4 meals/wk
Fast Fried Foods	French fries, pizza, KFC	4 + times/wk	1–3 /wk	< 1 time/wk
Pastries & Sweets	Cake, ice-cream, biscuits, chocolate	7 + servings/wk	5 − 6 /wk	< 5 servings/wk

Morris et al., (2015) [15]. Wk = week, mo = month, T = tablespoon, d = day

Procedure

Following ethical approval, participants were approached by research staff, student e-mail, social media, and face to face. The invitation email/social media message contained some brief information about the study. Participants approached face to face were given a recruitment booklet on the MIND diet. All Interested participants were asked to contact the researcher by email and sent a participant information sheet (PIS) and consent form. Date and time were arranged for data collection at participant’s home. Prior to week 1 of the study, all participants were asked to complete a 7-day food diary. The researcher visited each participant prior to week 1 to complete baseline measures of CANTAB tests, COM-B questionnaire, personal information sheet, consent form and weight and height measurements. Those in the intervention groups were given written information on the MIND diet elements and portion sizes and a self-monitoring resource (weekly chart) to monitor intake of MIND diet foods daily. The intervention group with support were guided through the website with an in-depth discussion on the MIND diet, the website and what was required of them (access the website regularly to access hints and tips on how to adhere to MIND diet). The researcher sent weekly text messages to the intervention group with support only, over the 12-week intervention to remind participants to complete daily food charts and access website. The control group were given information on the eat well plate and recommended national dietary guidelines. On week 12, all participants were asked to complete food diaries again. In the 2 weeks following the intervention, the researcher visited each participant’s home to collect follow-up data on COM-B, CANTAB tests and weight measurements.

Data analyses

All statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS) with significance set at P < 0.05 throughout (IBM SPSS Statistics for Windows, version 24.0, IBM Corp, Armok NY). Prior to analyses, the data was checked for normality by first examining statistics for skewness, kurtosis and to check that the scores are normally distributed. Skewed variables (TDF: behaviour regulation; TDF: emotion; free sugars; riboflavin; folates B6) were log transformed to attain a normal distribution prior to analyses, which is common in nutrition studies [37, 44]. A series of one-way ANOVA’s were conducted to look at baseline differences in cognitive function, MIND diet score, capability, opportunity, and motivation. To determine the effect of the 12-week dietary intervention on outcome measure, a 3(group: intervention with support vs intervention with no support vs control) X 2 (time: baseline and follow up) repeated measures ANOVA was used. Sphericity was checked and if it was supported, sphericity assumed was reported and if Box’s M was significant, Greenhouse Geisser was reported [45]. In case of any significance between-group differences, the ANOVA was followed by pairwise between-group comparisons using the Bonferroni post hoc test to appropriately control for multiple comparisons.

Results

Participants characteristics and outcomes variables for pre and post intervention are reported to explore the impact of the dietary intervention cognitive function, mood, QOL, MIND diet adherence and COM-B. Recruitment period was from April to June 2019, with the intervention period between June and November 2019. All participants that were recruited, completed the intervention, which ended in November 2019.

Participants characteristics

Sample characteristics are presented in Table 3. The mean age of participants was 45.5 years. There were more females than males recruited onto the study (62.5%). A large percentage of the participants were professional (58.5%), married (70%), with a higher education (56%), and earning over £50,000 a year (32%).

Table 3.

Participants characteristics by group

Demographic information	MIND diet with support n = 15	MIND diet no support n = 14	Control n = 12
Age M (SD)	46 (5.2)	45.5 (6.0)	42.5 (4.3)
Gender %
Male	14	50	50
Female	86	50	50
Occupation%
Professional	46	60	20
Skilled	27	12	7
Unskilled	27	14	2
Retired	0	14	0
Unemployed	0	0	0
Income%
Under £10,000	6	15	0
£10–20,000	0	0	17
£20–30, 000	19	20	17
£30–40,000	33	15	41
£40–50,000	14	6	0
Over £50,000	28	44	25
Marital status %
Married	83	67	83
Co-habit	0	6	0
Separated	0	6	0
Widowed	0	15	0
single	17	6	17
Education %
Primary	6	15	34
Secondary	6	21	0
Further education	20	6	24
Higher education	68	58	42

Descriptive information on age, occupation, income, marital status, and education were computed to identify participants characteristics

M = mean, % = percentage, N = number (n = 41)

Exploring group differences at baseline

A series of one-way ANOVA’s were conducted to compare baseline differences for all groups for cognitive function, mood, QOL, COM-B components, and MIND diet score. The groups did not differ on any of the measures except for TDF domain belief about capabilities showing a higher mean score for the intervention without support group (see supp data 4). A series of one-way ANOVA’s were conducted to compare baseline differences for all groups for all nutrients (energyKCal, energy (kj), carbohydrates, protein, fat, fibre, sugar, free sugar, saturated fat, poly-saturated fat, omega 3, omega 6, sodium, calcium, iron, zinc, iodine, vitamin A, vitamin D, vitamin B12, vitamin C, riboflavin, vitamin B6, folates B9). The groups did not differ in any measures except for fibre showing a higher mean score for the intervention with support group (see supp data 5).

Intervention effects on outcome variables

In order to determine if there is a significant difference on mean scores on cognitive function, MIND diet score and nutritional information over time (baseline to follow up), between the control and intervention groups, and interactions between group and time, a 3 (control Vs intervention (MIND diet with support) Vs intervention (MIND diet no support) × 2 (baseline Vs follow up) repeated measures factorial analysis of variance was conducted (see Tables 4 and 5). A series of graphs were plotted to show the time by group interaction for MIND diet score, TDF knowledge, memory/attention, behaviour regulation, belief about capabilities, belief about consequences, skills, environmental context and resources and emotion (see Fig. 4 A-I). A further series of graphs were plotted to show either time or group effects for carbohydrates, sugar, free sugar, and saturated fat and vitamin A vitamin b6, 9, iron, vitamin C, fibre, and omega 6 (see supplementary data 6).

Table 4.

Pre and post mean scores and standard deviations for intervention groups (MDWS, MDNS) and control group for mood, QOL, MIND diet score and COM-B components

Variable	Baseline Mean(sd)	Follow-up Mean (sd)	Time	Group	TimexGroup Interaction
	MDWS MDNS CONTROL	MDWS MDNS CONTROL		DWS, MDNS)
PA	23.42 25.05 25.01 (4.92) (5.27) (4.98)	26.88 29.22 24.18 (5.63) (6.48) (4.40)	F (2,38) = 7.488, p = 0.009,np2 = 1.66	F(2,38) = 1.109, p = 0.340, nP2 = 0.55	F(2,38) = 3.326, p = 0.047,np2 = 1.49
NA	11.45 12.29 11.04 (1.52) (2.98) (1.36)	11.58 12.40 11.36 (2.20) (3.26) (1.91)	F(2,38) = .191, p = .665,np2 = .005	F(2,38) = 1.283, p = .289,np2 = .001	F(2,38) = .023, p = .977, np2 = .001
QOL: Environment	32.73 32.00 31.14 (3.88) (5.05) (4.43)	33.20 32.00 32.16 (4.14) (4.70) (4.24)	F(2,38) = .477, P = .494, np2 = .012	F(2,38) = .297, p = .745,np2 = .015	F(2,38) = .885, p = .885,np2 = .006
QOL: Physical	27.26 29.07 30.58 (3.63) (4.58) (2.71)	29.20 29.57 29.58 (2.45) (2.40) (3.62)	F(2,38)) = 1.354, P = .252,np2 = .034	F(2,38) = 1.250, p = .298,np2 = .062	F(2,38) = 4.196, p = .023, np2 = .181
QOL: Psychological	20.80 22.28 22.33 (2.95) (3.49) (2.99)	22.46 23.35 22.25 (2.97) (3.05) (2.89)	F(2,38) = 6.261, P = .17,np2 = .141	F(2,38) = .630, p = .538, np2 = .032	F(2,38) = 2.038, p = .144, np2 = .097
QOL: Social	10.53 11.50 11.83 (2.29) (2.02) (1.46)	11.53 11.57 11.83 (2.26) (2.27) (1.89)	F(2,38) = 90.081, p < .001, np2 = .703	F(2,38) = 7.368, p = .002, np2 = .279	F(2,38) = 31.684, p < .001, np2 = .625
MIND score	6.46 6.71 7.16 (1.15) (2.18) (1.60)	11.50 10.42 6.66 (1.62) (1.85) (1.73)	F(2,38) = 410.50, p < .001, np2 = .915	F(2,38) = 67.188, p < .001, np2 = .780	F(2,38) = 67.06, p < .001, np2 = .779
TDF: Knowledge	2.80 3.28 2.58 (0.94) (1.20) (1.08)	10.00 10.00 3.41 (0.00) (0.00) (2.35)	F(2,38)) = 77.16, p < .001, np2 = .670	F(2,38) = 14.001, P < .001, nP2 = .425	F(2,38) = 17.100, p < .001, nP2 = .474
TDF: behaviour regulationa	4.26 4.00 3.83 (2.18) (2.14) (.717)	9.93 9.57 4.16 (.25) (1.34) (2.03)	F(2,38) = 49.229,p < .001, nP2 = .564	F(2,38) = 6.20, p = .005, np2 = .246	F(2,38) = 6.812, p = .003, np2 = .264
TDF: Memory/attention	5.53 6.28 5.50 (1.92) (1.32) (1.38)	8.93 8.28 6.25 (1.22) (1.20) (1.65)	F(2,38)) = 24.39, p < .001, nP2 = .391	F(2,38) = .945, P = 0.398, np2 = .047	F(2,38) = 9.98, p < .001, np2 = .344
TDF: Skills	7.26 7.92 7.83 (1.94) (2.26) (1.69)	9.66 9.14 7.66 (0.89) (1.70) (1.77)	F(2,38) = 70.69, p < .001, np2 = .650	F(2,38) = 6.024, p = .005, np2 = .241	F(2,38) = 9.189, p = .001, np2 = .326
TDF: Resources/context	11.13 12.92 11.75 (2.50) (3.12) (2.89)	17.33 17.85 12.91 (2.49) (2.03) (3.31)	F(2,38) = 9.89, p = .003, np2 = .207	F(2,38) = 3.81, p = .031, np2 = .167	F(2,38) = .626, p = .540,np2 = .032
TDF: Social influence	2.13 2.07 1.58 (1.24) (1.26) (0.99)	3.46 2.85 2.16 (1.30) (1.23) (1.46)	F(2,38) = 50.51, p < .001, np2 = .571	F(2,38) = 15.97, p < .001, np2 = .457	F(2,38) = 8.021, p = .001, np2 = .297
TDF: Belief about capability	6.33 7.28 6.25 (1.63) (2.61) (1.42)	12.60 12.21 7.08 (2.16) (3.01) (2.84)	F(2,38) = 116.572, p < .001, np2 = .754	F(2,38) = 12.066, p < .001, np2 = .388	F(2,38) = 13.969, p < .001, np2 = .424
TDF: Belief about consequences	11.46 12.28 11.50 (1.35) (1.85) (2.23)	18.33 18.14 12.91 (2.28) (1.70) (3.42)	F(2,38) = 62.05, p < .001, Np2 = .620	F(2,38) = 9.801, p < .001 Np2 = .340	F(2,38) = 4.85, p = .013 Np2 = .203
TDF: EMOTIONa	6.20 6.42 6.16 (.77) (1.08) (.389)	9.66 9.85 6.66 (1.04) (.363) (1.55)	F(2,38) = .048, P = .828, np2 = .001	F(2,38) = 1.348, p = .272, np2 = .066	F(2,38) = 1.984, p = .151, np2 = .095
PRMMCLI	1602.07 1467.19 1521.39 (254.87) (182.99) (303.62)	1514.10 1452.10 1654.08 (222.47) (249.88) (317.07)	F(2,38) = .001, p = .980, np2 = .000	F(2,38) = .363, p = .698, np2 = .019	F(2,38) = .41, p = .63, np2 = .02
SSPFSL	6.06 6.21 6.25 (1.43) (.89) (1.71)	5.86 6.42 6.25 (.83) (1.28) (1.71)	F(2,38) = 8.51, p = .006, np2 = .183	F(2,38) = .856, p = .433, np2 = .043	F(2,38) = .653, p = .526, np2 = .033
SWMTE468	10.73 11.21 16.58 (8.20) (10.44) (12.39)	8.13 8.07 10.33 (7.98) (8.63) (11.08)	F(2,38) = 1.629, p = .210, np2 = .041	F(2,38) = .170, p = .845, np2 = .009	F(2,38) = .240, p = .788, np2 = .012
RTIFMRT	405.68 403.19 400.64 (39.30) (48.72) (39.37)	413.27 419.04 404.52 (51.09) (56.79) (26.86)

PA positive affect, NA negative affect, QOL quality of life, PRMMCLI Pattern Recognition Memory: Mean Correct Latency, SSPFSL Spatial Span: Forward Span Length, SWMTE468 Spatial Working Memory: Total Errors, RTIFMRT Reaction Time Task: Mean Five Choice Reaction Time

Significant results are presented in bold

^a log transformed

Table 5.

Pre and post mean scores and standard deviations for both intervention groups (MDWS, MDNS) and control group for micro/macro nutrients

Variable	Baseline Mean(SD)	Follow-up Mean (SD)	Time	Group	TimexGroup Interaction
	MDWS MDNS Control	MDWS MDNS Control
EnergyKcal	1682 1552 1537 (399) (352) (318)	1504 1440 1495 (326) (417) (437)	F (2,38) = 3.939,p = 0.054, np2 = .094	F(2,38) = 0.339, p = 0.714,nP2 = 0.018	F(2,38) = 0.484, p = 0.620, np2 = 0.025
Energy (Kj)	7057 6522 6449 (1683) (1478) (1332)	6313 6050 6276 (1372) (1745) (1831)	F (2,38) = 3.908,p = 0.055, np2 = .093	F(2,38) = 0.333, p = 0.719,nP2 = 0.017	F(2,38) = 0.487, p = 0.618, np2 = 0.025
Carbohydrate(g) -as %energy	183(52) 175(49) 152(40) 44(5.5) 45(5) 40(6)	149(46) 158(43) 56(52) 39(7) 44(7) 42(4)	F (2,38) = 5.010,p = 0.031, np2 = 0.116	F(2,38) = 0.342, p = 0.713,nP2 = 0.017	F(2,38) = 2.317, p = 0.112, np2 = 0.018
Protein (g) - as %energy	76(21) 73(16) 75(14) 18(3) 19(4) 20(4.5)	3(25) 74(20) 73(14) 22(5) 21.5(7) 20(3.5)	F (2,38) = 0.464,p = 0.500, np2 = 0.012	F(2,38) = 0.521, p = 0.598, 00np2 = 0.027	F(2,38) = 0.773, p = 0.469, np2 = 0.039
Fat(g) - as %energy	67(16) 57(14) 61(18) 36.5(5) 33(5) 35.5(5.5)	62(16) 54(21) 60(21) 37.5(7) 33(5.5) 35.5(5)	F(2,38) = 1.311, p = 0.259, np2 = 0.033	F(2,38) = 1.192, p = 0.315, np2 = 0.059	F(2,38) = 0.115, p = 0.892, np2 = 0.006
Fibre(g)	18(5) 15(4) 13(3)	21(5) 18(6) 13(3)	F(2,38) = 4.939, p = 0.032, np2 = 0.115	F(2,38) = 7.765, p = 0.001, np2 = 0.290	F(2,38) = 1.008, p = 0.374, np2 = 0.050
Sugars(g)	64(25) 68(26) 48(23)	55(26) 53(20) 48(25)	F(2,38) = 6.193, p = 0.017, np2 = 0.140	F(2,38) = 1.286, p = 0.288, np2 = 0.063	F(2,38) = 1.366, p = 0.267, np2 = 0.067
Free sugarsa(g)	22(11) 28 (16) 22(22)	13(9) 17(12) 23(24)	F(2,38) = 9.146, p = 0.004, np2 = 0.194	F(2,38) = 0.618, p = 0.545, np2 = .031	F(2,38) = 2.893, p = 0.068, np2 = .132
Saturated fat (g)	22(6) 21(7) 21(6)	18(6) 18(7) 19(8)	F(2,38) = 6.758, p = 0.013, np2 = 0.151	F(2,38) = 0.103, p = 0.903, np2 = 0.005	F(2,38) = 0.488, p = 0.618, np2 = 0.025
Poly-unsaturated fat	10(3) 8(4) 9(3)	12(5) 9(4) 9(3)	F(2,38) = 1.274, p = 0.266, np2 = 0.032	F(2,38) = 2.879, p = 0.069, np2 = 0.132	F(2,38) = 0.413, p = 0.664, np2 = 0.021
Omega n3(mg)	0.9(0.4) 0.9(0.4) 0.6(0.3)	1.3(1.4) 1(0.7) 0.6(0.2)	F(2,38) = 1.29, p = 0.264, np2 = 0.033	F(2,38) = 2.328, p = 0.111, np2 = 0.109	F(2,38) = 0.588, p = 0.577, np2 = 0.029
Omega 6(g)	5.2(2.3) 3.6(2.8) 3.1(1.6)	7.2(4.6) 4.6(3.3) 4.0(1.7)	F(2,38) = 5.709, p = 0.022, np2 = 0.131	F(2,38) = 4.315, p = 0.020, np2 = 0.185	F(2,38) = 0.412, p = 0.665, np2 = 0.021
Sodium(mg)	1863 1768 1756 (556) (535) (453)	1618 1608 1731 (466) (539) (451)	F(2,38) = 2.492, p = 0.123, np2 = 0.062	F(2,38) = 078, p = 0.925, np2 = 0.004	F(2,38) = 0.481, p = 0.622, np2 = 0.025
Calcium(mg)	629(211) 676(233) 695(241)	643(233) 653(208) 671(319)	F(2,38) = 0.065, p = 0.801, np2 = 0.002	F(2,38) = 0.198, p = 0.821 ,np2 = 0.010	F(2,38) = 0.084, p = 0920., np2 = 0.004
Iron 9 mg)	9(3.5) 9(3.5) 7(2)	10(2,4) 10 (2.2) 7(2.6)	F(2,38) = 1.025, p = 0.318, np2 = 0.026	F(2,38) = 4.771, p = 0.014, np2 = 0.201	F(2,38) = 0.468, p = 0.630, np2 = 0.024
Zinc(mg)	7.6(2.3) 7.5(2.3) 7.3(2.3)	7.8(2) 8.4(2.2) 7.5(2.4)	F(2,38) = 1.630, p = 0.209, np2 = 0.041	F(2,38) = 0.335, p = 0.718, np2 = 0.017	F(2,38) = 0.397, p = 0.675, np2 = 0.020
Iodine(mcg)	108(60) 115(37) 116(45)	98(31) 114(46) 103(61)	F(2,38) = 0.772, p = 0.385, np2 = 0.020	F(2,38) = 0.339, p = 0.717, np2 = 0.018	F(2,38) = 0.156, p = 0.856, np2 = 0.008
Vitamin A(ug)	694(484) 602(333) 542(310)	891(245) 790(351) 683(388)	F(2,38) = 8.032, p = 0.007, np2 = 0.174	F(2,38) = 1.195, p = 0.314, np2 = 0.059	F(2,38) = 0.079, p = 0.925, np2 = 0.004
Vitamin Da(ug)	2.3(1.9) 2.7(1.3) 2.6(1.4)	3(2.8) 2.8(1.2) 2.6(1.7)	F(2,38) = .953, p = .335, np2 = .024	F(2,38) = .831, p = .443, np2 = .042	F(2,38) = 1.283, p = .289, np2 = .063
Riboflavin B2a (mg)	1.4(0.5) 1.4(0.4) 1.3(0.5)	2.3(3.9) 1.3 (0.3) 1.2(0.4)	F(2,38) = .048, p = .828, np2 = 0.001	F(2,38) = 1.161, p = .324, np2 = .058	F(2,38) = .918, p = .408, np2 = .046
Vitamin B6(mg)	1.7(0.5) 1.7(0.3) 1.3(0.3)	1.8(0.6) 1.5(0.3) 1.3(0.3)	F(2,38) = 0.080, p = 0.779, np2 = 0.002	F(2,38) = 4.55, p = 0.017, np2 = 0.193	F(2,38) = 0.526, p = 0.595 ,np2 = 0.027
Folates B9a(ug)	223(73) 233(59) 179(56)	251(71) 246(67) 169(53)	F(2,38) = 1.351, p = .252, np2 = .034	F(2,38) = 5.056, p = .011, np2 = .210	F(2,38) = .911, p = .411, np2 = .046
Vitamin B12(ug)	3.7(2.3) 4.2(1.7) 4.1(1.5)	3.5(1.6) 4.2(1.8) 4.1(2.1)	F(2,38) = 0.030, p = 0.864, np2 = 0.001	F(2,38) = 0.545, p = 0.584, np2 = 0.028	F(2,38) = 0.037, p = 0.964, np2 = 0.002
Vitamin C(mg)	70(49) 71(38) 42(32)	113(61) 83(55) 48(26)	F(2,38) = 9.616, p = 0.004 ,np2 = 0.202	F(2,38) = 4.160, p = 0.023 ,np2 = 0.180	F(2,38) = 3.017, p = 0.061, np2 = 0.137

Significant results are presented in bold. ^a = log transformed

Given that energy intakes did not differ significantly between the groups at baseline or follow up, micronutrient intakes were not adjusted for energy intakes

Fig. 4.

A-I Interaction for time and group for TDF components and MIND diet score. A-I shows that there is a significant improvement in the above variables for both intervention groups but no change in the control group. No significant difference between intervention groups. MIND diet with support n = 15(MDWS), MIND diet no support n = 14(MDNS) control n = 12

Adherence to the MIND diet

There was an increase in mean scores over time (baseline to post intervention) across all 3 groups for MIND diet score (p < 0.001) (see Table 4). There was a difference in mean scores between groups for MIND diet score (p < 0.001). Post hoc tests showed there was a significant difference in mean scores between both intervention groups and the control group, with mean scores being higher in the intervention groups than the control group (p < 0.001). However, no significant difference was shown between the two intervention groups (see Table 4), thus rejecting hypothesis 3, that the highest adherence rate would be seen in the MIND diet with support group.

Nutritional data

Given that energy intakes did not differ significantly between the groups at baseline or follow up, micronutrient intakes were not adjusted for energy intakes. There was a decrease in mean scores over time across all 3 groups for carbohydrate (p = 0.031), sugars (p = 0.017), free sugar(p = 0.004), and saturated fat (p = 0.013) (See Table 5). There was a difference in mean scores between groups for fibre(p = 0.001), omega 6(p = 0.020), iron(p = 0.014), vitamin B6(p = 0.017), folates B9 (p = 0.011) and vitamin C (p = 0.023). Post hoc tests show that there was a significant difference between MIND diet with support group and the control group for fibre, NSP, omega 3, vitamin B6 and vitamin C, with mean scores being higher in the MIND diet with support group (p < 0.05). No significant difference was shown between the two intervention groups for these nutrients.

Post hoc tests also showed a mean difference in scores between both intervention groups and the control group for iron and folates B9, with mean scores being higher in the intervention groups than the control group (p < 0.05). No significant difference shown between the two intervention groups and the control group (see Table 5). There were no significant interaction effects for time by group for any of the individual nutrients. Thus, accepting hypothesis 4 that increased adherence to the MIND diet improves nutritional data over the intervention period.

Cognitive function

There was a decrease in mean scores over time across all 3 groups for spatial working memory (p = 0.006). No difference in mean scores between groups or any significant interaction effects for cognitive function tests (see Table 4), thus, rejecting hypothesis 2, that adherence to the MIND diet improves cognitive function.

Mood

There is an increase in mean scores over time (baseline to post intervention) across all 3 group for positive affect. There are significant interaction effects for PA (p = 0.047). This shows that compared to the control group, the intervention groups at completion of the dietary intervention showed an increase in mean scores for PA (see Table 4) thus, accepting hypothesis 2, that adherence to the MIND diet improves mood in 40–55-year-olds.

Quality of life

There are significant interaction effects for physical quality of life (p = 0.023). This shows that compared to the control group, the intervention groups at completion of the dietary intervention showed an increase in mean scores for physical quality of life (see Table 4) thus, accepting hypothesis 2, that adherence to the MIND diet improves quality of life in 40–55-year-olds.

COM-B/TDF components

There is an increase in mean scores over time (baseline to post intervention) across all 3 group for all TDF components (P = 0.003 to P < 0.001) (see Table 4). There was a difference in mean scores between groups for all TDF components: Knowledge (p < 0.001), behaviour regulation (p = 0.05), skills (p = 0.05), environmental context and resources (p = 0.031), social influences (p < 0.001), belief about capabilities((p < 0.001), and belief about consequences ((p < 0.001). Post hoc tests showed that for all the above variables, there was a significant difference in mean scores between both intervention groups and the control group, with mean scores being higher in the intervention groups than the control group (p < 0.05). However, no significant difference was shown between the two intervention groups (see Table 4). There are significant interaction effects for all the TDF components except for Social Influences (P = 0.003 to P < 0.001). This shows that compared to the control group, the intervention groups at completion of the dietary intervention showed an increase in mean scores for capability, opportunity and motivation towards adherence to the MIND diet (See Table 4). Thus, accepting hypothesis 1, that we will improve adherence to the MIND diet, by improving capability, opportunity and motivation.

Discussion

To our knowledge, this is the first RCT evaluating the effects of the MIND diet on cognitive function, mood and quality of life. It is also the first study to assess the usefulness of a 12-week behaviour change intervention aimed at increasing capability, opportunity, and motivation for adopting the MIND diet in adults aged 40–55 years old. While there is a growing body of research using the BCW in designing interventions for a range of behaviours including enhancing self-care adherence in patients with heart failure [46], improving hearing aid use [47] and improving early detection of atrial fibrillation [48], to our knowledge, this is the first study to apply the BCW framework to design and evaluate an intervention promoting adherence to the MIND diet. The 12-week dietary intervention significantly improved MIND diet score and components of the COM-B model. Additionally, MIND diet adherence led to improvements in various nutrient intake over time/and or between groups, including fibre, NSP, omega-6, vitamin A, C, B6, iron, folate (B9), carbohydrates, protein, sugar, free sugar, and saturated fat. These findings suggest that adherence to the MIND diet not only enhances intake of key nutrients associated with brain health (e.g., folate B9, iron, and vitamin C) but also contributes to overall improved nutrient intake, which may have broader implications for general health.

The current study found no association between higher MIND diet adherence and cognitive function. Previous observational studies found no association with MIND diet adherence and cognitive decline [49], or with older adults with cognitive impairment [50]. The findings from the current study are also very similar to other dietary patterns such as the Med diet that did not find an association with higher adherence to diet and executive function, memory, processing speed and visual-spatial memory over a short 6-month RCT in adults aged over 65 years [44]. The findings from the current study are in contrast with recent research, that found an improvement in cognitive function in middle aged adults over a 3-month period [4]. However, the target group in this study were also obese adults, and this study used a calorie deficit to encourage participants to lose weight also. Research has found a link between BMI and cognitive function, and that an obesity related intervention may improve cognitive performance [51]. The majority of current literature focuses on cross-sectional or prospective cohort studies that assess dietary intake during midlife and track cognitive changes over decades. This study targeted a middle-aged cohort, where cognitive performance is typically robust, creating a potential ceiling effect that could mask detectable changes from dietary interventions. As a result, the MIND diet’s potential benefits may not manifest within the short-term intervention timeframe used here but rather over longer periods through its ability to delay the onset of cognitive decline. Discussing the long-term adherence to the MIND diet, its ability to reduce neurodegeneration markers, and its impact on later-life cognition is important for future research.

A range of cross-sectional and longitudinal studies that found that higher adherence to the MIND diet was associated with reduced risk of cognitive decline [52–54] and better cognitive function [55]. Similarly, studies evaluating the Med diet have reported improvements in cognitive performance [56, 57]. The present study analysed a relatively small sample (n = 41) over 12 weeks, whereas prior RCTs and observational studies involved larger cohorts followed over several years. It is likely that the study’s small sample size and short duration lacked the statistical power to detect an effect of MIND diet adherence on cognitive function. Furthermore, differences in cognitive assessment methods could explain the discrepancies in findings. Previous studies measured different cognitive domains, such as visuospatial ability and perceptual orientation, which may be more sensitive to dietary influences. Additionally, given that prior research primarily focused on older populations, it is possible that the relationship between diet and cognitive function is not as pronounced in midlife [58].

The current study aligns with previous research indicating that whole dietary patterns, such as the Med and MIND diets, can improve mood-related outcomes, including contentment, alertness, and reductions in depressive symptoms [59–61]. Research has also shown that everyday health behaviours, such as fruit and vegetable consumption, are associated with better mood [62]. In contrast, one study reported that an 8-week Med diet intervention did not improve mood [63]. However, this study used the Profile of Mood States (POMS) scale, which assesses mood retrospectively over the previous month, whereas the current study employed the PANAS scale to measure mood multiple times per day over a four-day period at both baseline and follow-up. By capturing real-time fluctuations in mood, the PANAS scale may provide a more accurate reflection of mood changes over the intervention period. Additionally, the POMS scale measures only six distinct mood states, potentially overlooking aspects of mood that could be influenced by dietary changes, such as alertness [64]. One study measuring daily mood fluctuations found that a dietary supplement improved alertness over four weeks [65]. Furthermore, a 16-week Med diet intervention did not improve mood scores in a study where participants already had low depressive symptom scores at baseline, suggesting a limited capacity for improvement [66].

The literature on quality of life shows that in general, dietary patterns are associated with quality of life and well-being [67, 68], with adherence to specific dietary patterns such as the Med diet leading to a better quality of life [69]. Similar to previous research, the findings from this study found that the intervention significantly improved physical quality of life [28] but with no improvement in the psychological, social or environment components [28]. Considering the growing population of older people around the globe, and the imperative to maintain QoL in older age, these findings are important, as they highlight that changes in diet promote aspects of well-being. This is also aligns with other research that did not find an association with diet and the psychological component of quality of life [70]. In contrast, some research has reported that higher adherence to the Mediterranean diet was associated with better physical and psychological quality of life [28, 71]. However, both these studies were conducted in older adults and quality of life was assessed with the Short Form Health Survey (SF-12), which research has shown to measure different QOL constructs, with the SF-12 measuring health related quality of life and WHOQOL measuring global QOL [72].

The MIND diet is hypothesised to exert its neuroprotective effects through multiple mechanisms, including reducing systemic inflammation and improving cardiovascular health. Chronic low-grade inflammation has been implicated in the pathogenesis of cognitive decline and Alzheimer’s disease, with dietary patterns such as the MIND diet potentially modulating inflammatory pathways [3]. Furthermore, the diet emphasizes nutrient-dense foods like leafy greens, berries, and nuts, which contain bioactive compounds (e.g., polyphenols and omega-3 fatty acids) known to reduce inflammation. Improved cardiovascular health is another proposed mechanism, as vascular risk factors, including hypertension, atherosclerosis, and diabetes, are significant contributors to cognitive impairment. The MIND diet’s overlap with the DASH diet supports improved vascular health, which may prevent microvascular damage and support cognitive longevity.

All the COM-B/TDF domain scores significantly improved at the end of the intervention, except for social influence, indicating that participants enhanced their capability, opportunity, and motivation to adhere to the MIND diet. The two intervention groups differed in their resources: one group received a self-monitoring resource, access to a website with 18 BCTs and 6 intervention functions, while the second group received only a self-monitoring chart and basic dietary information. Both groups significantly improved their MIND diet adherence, with no significant difference between them, suggesting that self-monitoring, education, and goal setting were sufficient to promote adherence. A systematic review found that, dietary interventions that use BCT’s to deliver the interventions, found that goal setting, self-monitoring, increasing knowledge were the most utilised in app based mobile interventions and reported positive effects [73], and that BCT’S that self-regulate behaviour such as self-monitoring and goal setting, explain intervention effects [74].

Participant engagement with the intervention varied across its components. While all participants accessed the website to obtain foundational information about the MIND diet, adherence targets, and recipes, only a small subset utilized the supplementary weekly content addressing specific barriers and facilitators to adherence, such as navigating workplace dining or eating out. The majority of participants identified self-monitoring, goal setting, and the provision of clear dietary targets as the most impactful elements supporting their adherence. These findings align with Control Theory [75] which emphasizes the importance of goal-directed feedback mechanisms in driving behavioural change. A chat room feature, designed to provide peer support and monitored by the researcher, offered an additional measure of engagement. Weekly adherence tracking, submitted directly to the researcher, confirmed that all participants engaged with the intervention to some extent, as they needed to review adherence guidelines on the website and report their progress regularly. These results suggest that the primary behavioural change techniques incorporated into the intervention were sufficient to promote high adherence rates, even with limited use of supplementary materials.

Strengths and limitations

A major strength of the study is the design (RCT) and the inclusion of a theory driven intervention. This study also used CANTAB for measuring cognitive function, which is a highly sensitive battery of tests recognised as the gold standard cognitive assessment and data collection. Furthermore, CANTAB tests are sensitive to subtle changes with age and neurodegeneration. A further strength of the study is the use of food diaries and food charts, which minimises reliance on memory assessing usual food intake [76]. Importantly, the high completion rates in the intervention group point to the acceptability of the dietary intervention. Furthermore, the fact that participants were able to increase their MIND diet score suggests adherence is achievable in healthy adults at midlife.

However, there are several limitations of the study which include the lack of ability to track participants engagement on the website and their interaction with the website materials. However, weekly text messages were sent to those in the intervention with support group to remind them to engage with the website. Future research should use an analytical software programme that can assess engagement on the website [77]. As a pilot study, the sample size was not large enough to provide sufficient power for detecting significant effects across multiple cognitive measures. The power analysis was based on a small effect size (0.28) for the primary outcome, dietary uptake, and the study was not explicitly powered to assess cognitive outcomes. As a result, the findings related to cognitive function should be interpreted with caution and are exploratory in nature. Additionally, the study focused on a single effect size from previous dietary interventions, which may not fully capture the potential variability in the effects of the MIND diet on cognitive function. Future studies with larger sample sizes and focused on cognitive outcomes are needed to confirm these preliminary results and provide more robust data on the relationship between dietary interventions and cognitive function. However, the current RCT is a pilot study and not a full-scale RCT. According to the MRC, preparatory work prior to embarking on a full scale RCT is important to assess the feasibility and acceptability of complex health interventions. Assessing feasibility and acceptability is crucial in uncovering potential issues related to acceptability, compliance, recruitment, retention, and delivery of the intervention [73]. Furthermore, assessing the effectiveness of a 12-week dietary intervention on cognitive function outcomes may be insufficient to detect a significant improvement in cognitive function. The potential interaction between baseline diet quality and intervention outcomes warrants further exploration. While participants with healthier baseline diets may have found it easier to adhere to the MIND diet, those with lower baseline diet quality may have experienced greater improvements. Future studies could stratify analyses by baseline diet quality to better understand these interactions. The lack of blinding in this study represents a limitation as it could introduce biases affecting both participants and researchers. Participants’ awareness of their group allocation might have influenced their adherence to the dietary intervention or their self-reported outcomes, while researchers may have been susceptible to unintentional confirmation bias during data collection or analysis. Future research could mitigate such biases by implementing blinding methods, such as using placebo-like control diets or employing objective, automated data collection tools. A further limitation of this study is the absence of formal cognitive screening for conditions such as dementia or other cognitive impairments. While participants were asked to self-report any known cognitive conditions or difficulties with memory and concentration, no standardized assessment tools, such as the Mini-Mental State Examination (MMSE), were used. As such, the cognitive health status of participants was assessed based on self-reported data, which may not have captured individuals with subclinical cognitive impairments. Future research will incorporate standardized cognitive assessments of dementia at baseline, to strengthen methodological rigor.

Future RCT’s examining the effectiveness of the MIND diet over a longer duration is warranted.

Conclusions

This is the first RCT to test the effectiveness of the MIND diet on cognitive function, mood, and quality of life. It is also the first RCT to assess the usefulness of the COM-B in increasing adherence to the MIND diet. The findings from this study show that the BCW is an acceptable framework to design and deliver effective interventions to increase capability, opportunity, and motivation to adhere to the MIND diet. Our study showed that increased adherence to the MIND diet improved mood, quality of life and diet quality. Future interventions with longer duration are needed to establish an association with MIND diet and cognitive function in adults at midlife. This study recommends emphasising self-monitoring, goal setting and education on diet as an effective strategy for promoting adherence to the MIND diet. This study also recommends using less BCT’s with a less complex intervention focussing on a well-structured text messaging service with more contact and advice directly from the researcher. These findings may be of interest to health professionals wanting to find effective ways to design lifestyle interventions and promote health and well-being in individuals and communities.

Abbreviations

ANOVA

Analysis of variance

APPEASE

Acceptability, Practicability, Effectiveness, Affordability, Spill-over effects, and Equity.

BCW

Behaviour Change Wheel

BCT

Behaviour Change Techniques

CANTAB

Cambridge Neuropsychological Test Automated Battery

DASH

Dietary Approaches to Stop Hypertension

GAA

Gaelic Athletics Association

GP

General Practitioner

MED DIET

Mediterranean diet

MIND DIET

Mediterranean-DASH Intervention for Neurodegenerative Delay

MO

Mechanism of Action

MRI

Magnetic resonance imaging

NI

Northern Ireland

PRM

Pattern Recognition Memory

QOL

Quality of Life

RCT

Randomised Controlled Trial

RTI

Reaction Time

SSP

Spatial Span

SWM

Spatial Working Memory

TDF

Theoretical Domains Framework

UK

United Kingdom

WHOQOL

World Health Organisation Quality of Life

Authors’ contributions

D.T./ E.E.A.S. /J.M. designed the intervention. D.T. conducted all baseline and follow-up measures. D.T. drafted the manuscript. M.K. and L.K advised on all nutritional data analysis and interpretation, and nutritics. E.E.A.S., J.M., M.K., L.K. revised the manuscript critically for intellectual content. All authors read and approved the final manuscript.

Funding

The Department for the Economy (DfE) sponsored the research for this paper.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Ethical approval was obtained from the School of Psychology Staff & Postgraduate Filter Committee, Ulster University, which is in accordance with The Code of Ethics of the World Medical Association (Declaration of Helsinki). Consent was provided by all participants.

Consent for publication

Not applicable, no personal identification data is included.

Competing interests

The authors declare no competing interests.