Hostname: page-component-cd9895bd7-p9bg8 Total loading time: 0 Render date: 2024-12-13T08:44:08.844Z Has data issue: false hasContentIssue false

Frequency of fruit consumption and savoury snacking predict psychological health; selective mediation via cognitive failures

Published online by Cambridge University Press:  26 May 2022

Nicola-Jayne Tuck*
Affiliation:
School of Psychology, Aston University, Birmingham B4 7ET, UK
Claire V. Farrow
Affiliation:
School of Psychology, Aston University, Birmingham B4 7ET, UK
Jason Michael Thomas
Affiliation:
School of Psychology, Aston University, Birmingham B4 7ET, UK
*
*Corresponding author: Nicola-Jayne Tuck, email n.tuck@aston.ac.uk
Rights & Permissions [Opens in a new window]

Abstract

While there is growing interest in the link between diet and psychological health, there is a surprising lack of studies investigating the precise associations between nutrient-rich foods (such as fruit and vegetables) v. nutrient-poor foods (such as energy-dense savoury and sweet snacks), and psychological health. Similarly, the psychological processes underpinning the relationship between dietary intake and psychological health remain unclear. Hence, the present study aimed to explore the relationship between dietary consumption and psychological health, with cognitive processes as a theoretical mediator. This cross-sectional online study included 428 healthy adults (53 % female; mean age = 39·7 years, sd = 13·0), with participants completing a range of validated questionnaires measuring dietary habits and psychological health. Stepwise multiple regression revealed that more frequent consumption of fruit was associated with reduced symptoms of depression (β = –0·109, P = 0·025) and greater positive psychological wellbeing (β = 0·187, P < 0·001). Conversely, more frequent savoury snacking was associated with increased anxiety (β = 0·127, P = 0·005). Further, mediation analyses revealed that more frequent consumption of savoury snacks was associated with increased symptoms of depression, stress, anxiety and reduced psychological wellbeing, via an increase in cognitive failures (ps < 0·001). These results provide new insights on the independent associations between certain types of food and psychological health, and the psychological mechanisms that may mediate these. Further work is now required to establish causality and determine whether these may represent modifiable dietary targets that can directly (and indirectly) influence our psychological health.

Type
Research Article
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
© The Author(s), 2022. Published by Cambridge University Press on behalf of The Nutrition Society

Psychological health includes the absence of mental health complaints and the experience of positive psychological wellbeing(1). Given the health, social and economic burden of impaired psychological health, there are calls for new preventative public health approaches(Reference Jané-Llopis, Anderson and Stewart-Brown2), with recent research suggesting that dietary intake is a potential target for improving psychological health(Reference Głąbska, Guzek and Groele3).

Indeed, the consumption of nutrient-rich (unprocessed) foods, such as fruit and vegetables, have been associated with fewer psychological health issues(Reference Nicolaou, Colpo and Vermeulen4,Reference Lai, Hiles and Bisquera5) and reduced cognitive impairment(Reference Gomez-Pinilla and Nguyen6). Fruit and vegetable intake (FVI) are also associated with a reduced risk of symptoms of depression, stress and anxiety(Reference Gibson-Smith, Bot and Brouwer7,Reference Mikolajczyk, El Ansari and Maxwell8) , and there is accumulating evidence that FVI is positively related to psychological wellbeing(Reference Mujcic and J Oswald9,Reference Ocean, Howley and Ensor10) . Although studies have begun to explore fruit and vegetable consumption as separate predictors of psychological health(Reference Mihrshahi, Dobson and Mishra11,Reference Nguyen, Ding and Mihrshahi12) , few have evaluated the impact of frequency. However, a recent study reported that the frequency with which fruits and vegetables are consumed may be more important than quantity of consumption, suggesting a more nuanced approach to consumption may be required(Reference Ocean, Howley and Ensor10).

Conversely, habitual consumption of nutrient-poor (processed) foods, such as sweet and savoury snacks, is associated with increased risk of depression, anxiety, stress(Reference Sangsefidi, Lorzadeh and Hosseinzadeh13) and lower psychological wellbeing(Reference Smith and Rogers14). A prospective study reported that trans-unsaturated fatty acid in processed food, such as snack foods (crisps, cookies, cakes, etc.), was associated with a higher depression risk at follow-up of 8–10 years later risk(Reference Sánchez-Villegas, Verberne and De Irala15), illustrating the potential long-term effects of diet on psychological health. Further, evidence shows that a decrease in frequency of fruit intake and an increase in frequency of snack food intake independently coincided with increased perceived stress(Reference Liu, Xie and Chou16). Thus, the frequency of nutrient-poor food consumption may also distinctively contribute to psychological health.

While little is known about the psychological mechanisms by which diet may affect psychological health, the role of cognitive processes (e.g. memory and executive function) has been implicated(Reference Monti, Moulton and Cohen17,Reference Stevenson18) . Dietary intake high in energy content or low in micronutrients, antioxidants or fibre consumption may reduce optimal brain function, such as neurotransmitter regulation and inflammatory pathways, leading to poorer psychological health(Reference Young, Gauci and Scholey19). FVI have been shown to enhance cognition(Reference Polidori, Praticó and Mangialasche20), whereas saturated fat and sugar intake decrease cognitive performance(Reference Beilharz, Maniam and Morris21). Specifically, diet-induced changes to memory and inhibitory control have been consistently observed(Reference Francis and Stevenson22,Reference Attuquayefio and Stevenson23) . Frequent cognitive failures (memory errors) have been associated with increased perceived stress and sadness(Reference Boals and Banks24,Reference Payne and Schnapp25) . Additionally, reduced inhibitory control is a risk factor for depression(Reference Kaiser, Unger and Kiefer26,Reference Gotlib and Joormann27) . Even subtle cognitive deficits are significantly associated with reduced psychological wellbeing(Reference Gates, Valenzuela and Sachdev28); therefore, dietary intake may impact cognition and in turn psychological health.

Research has begun to identify relationships between the consumption of nutrient-rich or nutrient-poor foods and psychological health. However, it is important to evaluate the independent associations of the frequency and quantity of FVI, frequent snacking on energy-dense foods, and the role of cognitive processes as a potential mediator. This study aimed to assess the direct and indirect relationship between dietary intake and depression, anxiety, stress and wellbeing. It was predicted that greater nutrient-rich FVI (as separate frequency and portions variables) would be associated with increased psychological health, whereas greater nutrient-poor sweet and savoury snacking would be associated with decreased psychological health. It was also predicted that cognitive failures and inhibitory control would mediate the relationship between dietary intake and psychological health, whereby FVI would negatively predict cognitive failures and positively predict inhibitory control scores and thus increase psychological health, whereas, sweet and savoury snacking would positively predict cognitive failures, negatively predict inhibitory control, and thus reduce psychological health.

Method

Participants

Participants were recruited using the recruitment platform Toluna(29) to gain access to a nationally representative sample of the UK. Toluna is an online community website which invites members to complete paid surveys. Participants were compensated with 3000 Toluna points (redeemable towards retail vouchers) after completing the survey. A total of 977 participants provided informed consent to take part, with 442 participants completing this online cross-sectional study. Of these, fourteen were excluded from analyses due to having a go reaction time percentage accuracy lower than 60 % on the inhibitory control task (see below for more information on this task), thus the total n for analysis = 428 (53 % were female). Participants completed an initial screener questionnaire to ensure they met the inclusion criteria which included being aged 18–60 years of age, not colour blind (due to the demands of the inhibitory control task) and having English as their first language. Participants who rated their general health as poor over the last 12 months, who currently have or had diabetes, or an eating disorder and/or medically diagnosed food allergy, high blood pressure, a heart attack, or were experiencing medical illness were not eligible to participate (509 participants did not meet the screening criteria and 26 eligible participants left the survey incomplete). The study was approved by the College of Health & Life Sciences Ethics committee at Aston University.

Sample size

Using G × Power 3.1.9.2, with α set at 0·05, modelling a small effect size (f = 0·02), and power at 80 %, the minimum required sample size was 395 participants(Reference Faul, Erdfelder and Buchner30). However, to account for participants who might not complete the study in full, or whose data may need to be excluded from analysis (e.g. due to the threshold for the inhibitory control task), we aimed to enrol a minimum of 450 participants onto the study.

Procedure

The Gorilla Experiment platform was used to create and host the cross-sectional study. The survey and cognitive task were completed by participants online using a computer; phones or tablets were not permitted because the cognitive task required the use of a keyboard. Participants who expressed a willingness to take part provided informed consent and were screened for the inclusion criteria. Eligible participants then progressed through the series of questionnaires measuring demographic information, dietary intake, lifestyle behaviours, psychological health, cognition, mood and appetite, followed by the stop-signal task (SST), and were finally debriefed.

Measures

Lifestyle behaviours, health and demographic information

During an initial screener questionnaire, general health was assessed using a single item with a five-point Likert scale (ranging from poor to excellent) asking participants to rate: ‘Over the last 12 months, would you say that on the whole, your health has been…’. Participants were also asked if they previously or currently had diabetes, an eating disorder, a medically diagnosed food allergy, high blood pressure, a heart attack or any other medical illness in order to recruit healthy adults. Participants were asked their age and if they were colour blind (due to the demands of the cognitive task) during the initial screener questionnaire to meet the inclusion criteria. Sex, ethnic group, years and level of education and household income data were also collected to characterise the sample.

Within the short-form FFQ (see below), there were items that assessed average weekly alcohol intake, exercise and smoking status. Exercise (total minutes last week) was used in the present study as a control variable for the mediation models, given the evidence that physical activity has beneficial effects on cognitive function(Reference Hötting and Röder31,Reference Molteni, Wu and Vaynman32) and psychological health(Reference Schuch, Vancampfort and Rosenbaum33). Additionally, the twenty-one-item, four-point Likert Three-Factor Eating Questionnaire (TFEQ)(Reference Cappelleri, Bushmakin and Gerber34) was used to collect data on eating style in order to calculate an uncontrolled eating, cognitive restraint and emotional eating average score for the sample. Participants were also asked if they were vegetarian or vegan. To calculate BMI, participants were asked to report their height and weight in either metric or imperial units.

Dietary intake

The short-form FFQ was used to examine the consumption of foods in a ‘typical’ week, over the past month by measuring the frequency of food group consumption using an eight-point Likert scale (rarely or never to 5 + a day)(35). The food groups investigated in this study were as follows: (1) fruit consumption, which included fresh or tinned; (2) vegetable consumption, which included fresh, tinned or frozen, but not potatoes; (3) sweet snacking, which included biscuits, cakes, chocolate and sweets, and; (4) savoury snacking, which included crisps or savoury snacks. In addition to these measures of frequency of consumption, the short-form FFQ included two additional single items to collect data on average fruit and vegetable consumption per day in portions. One portion was quantified as approximately 80 g in weight and examples were provided for each item, such as ‘a handful of grapes’. The measure did not include similar items for sweet or savoury snacking; however, this questionnaire is valid and reliable for assessing diet in the UK population, hence, was used here(Reference Cleghorn, Harrison and Ransley36). Validation of this measure has shown that participant short-form FFQ responses for single food items are independently predictive of a participant’s diet quality score (as measured by a 217-item FFQ used in the UK Women’s Cohort Study)(Reference Cleghorn, Harrison and Ransley36). This highlights that individual food groups are significant for dietary assessment.

Psychological health

Symptoms of depression and anxiety during the past week were measured using the fourteen-item, four-point Likert, Hospital Anxiety and Depression Scale(Reference Zigmond and Snaith37). The Hospital Anxiety and Depression Scale is recommended for use in the general population as it was designed as a screening tool to identify possible and probable cases of anxiety and depression in outpatients(Reference Bjelland, Dahl and Haug38). Stress was measured using the fourteen-item, five-point Likert, Perceived Stress Scale(Reference Cohen, Kamarck and Mermelstein39). The items examine levels of stress experienced over the last month. The Perceived Stress Scale was designed for use within community samples(Reference Cohen, Kamarck and Mermelstein39). Visual analogue scales were used to assess current mood and appetite ratings(Reference Crichton40), and these were included immediately before the cognitive task. Participants indicated on a 100-mm horizontal line the point that represented their current experience of the following (where 0 = not at all, and 100 = very much): sad, happy, anxious, alert, drowsy, withdrawn and hungry. Positive psychological wellbeing was measured using the fourteen-item, five-point Warwick–Edinburgh Mental Wellbeing Scale, which is validated for use in the general adult population(Reference Tennant, Hiller and Fishwick41). The Warwick–Edinburgh Mental Wellbeing Scale evaluates subjective wellbeing and psychological functioning over the last 2 weeks(Reference Stewart-Brown and Janmohamed42).

Cognition

The twenty-five-item self-report Cognitive Failures Questionnaire (CFQ) measured attentional, memory, perceptual and action-related mental lapses in everyday tasks over the past 6 months(Reference Broadbent, Cooper and FitzGerald43). One item was adapted: ‘Do you leave important letters unanswered for days?’ whereby ‘letters’ was changed to ‘emails’. The CFQ was used to assess the frequency of global cognitive dysfunction. Scores can range from 0 to 100, whereby higher scores indicate more subjectively experienced cognitive failures. The CFQ has high internal validity (α = 0·91) and test–retest reliability(Reference Wallace, Kass and Stanny44). The CFQ has been correlated with the Everyday Memory Questionnaire (r = –0·64) which assesses memory errors(Reference Martin45).

The SST is a behavioural measure of inhibitory control which was used to provide a Stop-signal Reaction Time (SsRT); this measure represents a participant’s efficiency to inhibit an already initiated response(Reference Logan46). Higher SsRT indicate poorer inhibitory control, and the SST has been shown to be one of the most sensitive and reliable measures of executive control(Reference Padilla, Perez and Andres47). This task-based measure of inhibitory control may provide insight into the specific neural and behavioural impairments of cognition associated with both eating and the development or maintenance of poor psychological health(Reference Bartholdy, Dalton and O’Daly48Reference Marazziti, Consoli and Picchetti50). An online SST was created using The Gorilla Experiment Builder(Reference Anwyl-Irvine, Massonnié and Flitton51) based on the stop-signal paradigm(Reference Verbruggen and Logan52). The SST included a practice task with sixteen trials whereby participants were required to reach an accuracy threshold of 50 % to proceed to the main task. The main task comprised six blocks of sixteen trials (ninety-six trials in total, 75 % go trials, 25 % stop-signal trials). Each trial started with the presentation of a fixation cross which was replaced with an arrow pointing in the left or right direction inside a white circle (the inter-stimulus interval was 500 ms). Participants indicated the direction of the arrow using the keyboard responses: ‘b’ for left and ‘n’ for right (the maximum response time was 1500 ms before the trial moved along). On stop-signal trials, the white circle surrounding the directional arrow changed to red which occurred after a variable delay ranging from 250 to 400 ms, with 50 ms incremental increases in difficulty, and participants were required to withhold a response on these trials.

Statistical analyses

Descriptive data were calculated as means with standard deviations, or as frequencies (%) for categorical variables. For the frequency of dietary intake, the median and range are presented. For processing the cognitive task raw data, a published protocol was applied to estimate SsRT and the average stop-signal delay(Reference Congdon, Mumford and Cohen53). Simple linear and stepwise multiple regression analyses were conducted to explore the relationship between dietary intake and psychological health (before and after including the covariates: age, sex, BMI, exercise, general health rating, smoking status and alcohol intake). The selection of these covariates was informed by previous literature(Reference Gibson-Smith, Bot and Brouwer7,Reference Ocean, Howley and Ensor10,Reference Mihrshahi, Dobson and Mishra11) . The stepwise analysis put predictors into competition with each other by applying the ≤ 0·050 to enter and ≥ 0·1 to remove criteria. Separate models were conducted for dietary intake (food frequency) for fruit, vegetable, sweet and savoury snack consumption as predictors. FVI (separately) quantified as portions consumed were also explored as separate predictors. The outcome variables included four different measures of psychological health: symptoms of depression, anxiety, stress and positive psychological wellbeing scores. Multiple mediation analyses were performed using PROCESS, version 3.5(Reference Hayes54) in SPSS version 26 using bootstrapping over 5000 samples. The mediators included in the models were either cognitive failures score or SsRT (the index of inhibitory control). All mediation models included covariates that were consistently, significantly associated with the psychological health outcomes at the P < 0·01 level, which were general health rating and exercise (see Fig. 1 for the mediation paths).

Fig. 1. Mediation paths. All models controlled for general health rating and exercise. SsRT, Stop-signal Reaction Time.

Results

Participants characteristics

Descriptive statistics regarding demographic and lifestyle behaviour information for the healthy adult sample are presented in Table 1. On average, the participants were middle aged, 52·8 % of the sample were female and 89·7 % identified as White for ethnicity. The mean BMI for the sample was 26·0 (sd = 5·6) and 53·5 % were normal weight. On average, participants exercised for 2·5 h per week and most of the sample rated their general health as good to excellent (86 %). Further, 63·1 % were not smokers and 79·2 % consume less than 14 units of alcohol per week or rarely/never drink. Finally, education and household income information were provided by some of the sample (n 207), and of these, the majority achieved university-level education (66 %), and their household income was between £20 000 and £49 999 (73 %).

Table 1. Sample demographic information and lifestyle behaviour

(numbers and percentages; mean values and standard deviation, n 428)

TFEQ, Three-Factor Eating Questionnaire.

BMI n 422 because six participants selected ‘prefer not to say’.

NHS classifications for BMI ranges were used; underweight < 18·5, normal weight 18·5–24·9, overweight 25–29·9, obese ≥ 30; eating style (TFEQ) average scores for each dimension (uncontrolled eating, cognitive restraint and emotional eating) indicate a non-disordered sample.

Dietary intake

The frequency of dietary intake showed that on average, both sweet and savoury snacks were each consumed 2–3 times a week, while both fruit and vegetables were each consumed 4–6 times a week. The range of responses was from rarely or never to 5+ times a day; overall participants were not consuming fruit and vegetables frequently enough to reach recommended intake. On average, participants consumed 1·9 portions of fruit per day (sd = 1·5) and 2·3 portions of vegetables per day (sd = 1·5); thus, combined consumption was short of the recommended daily intake. FVI ranged from 0 to 10 portions per day and 10% of the sample identified as vegetarian or vegan.

Cognition

For the CFQ, the average frequency, variance and range of cognitive failures scores were as follows: mean = 34·8, sd = 16·2 and range = 0–97. For the SST, fifty-three participants (12 % of the sample) were required to complete the practice trials a second time to meet the accuracy threshold of 50 % and proceed to the test phase. However, descriptive results from the test phase showed that on average, the percentage accuracy on stop-signal and go trials was high (stop-signal trials, mean = 81 %, sd = 23 %; go trials, mean = 93 %, sd = 9 %). Finally, calculation of the SsRT revealed a mean of 151·4 ms (sd = 276·1 ms).

Psychological health

Descriptive statistics for mood ratings and psychological health are presented in Table 2. Visual analogue scale scores show that on average participants were happy, alert, low in sadness and anxiousness and not highly drowsy or withdrawn. Overall, mean depression and anxiety scores were considered non-clinical and stress levels were moderate(Reference Bjelland, Dahl and Haug38,Reference Andreou, Alexopoulos and Lionis55) . The mean psychological wellbeing score (46·0) is similar to other UK general population groups(Reference Bartram, Yadegarfar and Sinclair56). The Warwick–Edinburgh Mental Wellbeing Scale does not a have a ‘cut-off’ level to indicate those who have ‘good’ and those who have ‘poor’ psychological wellbeing, but the minimum score is 14 and the maximum is 70(Reference Stewart-Brown and Janmohamed42).

Table 2. Descriptive results for mood ratings and psychological health (Mean values and standard deviation, n 428)

VAS, Visual analogue scale; HADS, Hospital Anxiety and Depression Scale; PSS, Perceived Stress Scale; WEMWBS, Warwick–Edinburgh Mental Wellbeing Scale.

Regression results

Simple linear regression

Prior to the inclusion of covariates, simple linear regression was conducted to analyse sweet and savoury snacking, fruit, and vegetables (frequency or portions) consumed as predictors of either depression, anxiety, stress or wellbeing scores. Frequency of fruit consumption and portions of vegetables consumed negatively predicted depression scores (β = –0·140, P < 0·05, R2 = 0·020; β = –0·108, P < 0·05, R² = 0·012, respectively). The results also revealed that frequency and portions of fruit consumption positively predicted psychological wellbeing scores (β = 0·192, P < 0·001, R² = 0·037; β = 0·120, P < 0·05, R² = 0·014, respectively). Furthermore, sweet and savoury snacking positively predicted anxiety scores (β = 0·098, P < 0·05, R² = 0·010; β = 0·186, P < 0·001, R² = 0·035, respectively), while only savoury snacking positively predicted stress (β = 0·136, P < 0·05, R² = 0·019). Frequency of vegetable consumption did not predict psychological health (all ps > 0·05).

Stepwise regression

To evaluate the relative contribution of dietary intake (sweet and savoury snacking, frequency and portions of fruit and vegetable consumption) as predictors of psychological health, compared with other known predictors, the following covariates were included in a stepwise multiple regression: age, sex, BMI, exercise, general health rating, smoking status and alcohol intake. The significant stepwise regression models’ β values, P values, change in , and the adjusted for each step of the analysis are presented in Table 3. Firstly, for depression scores, the significant stepwise model selected exercise for entry first (β = –0·112, P = 0·021, R² = 0·027), then age was added (β = –0·180, P < 0·001, R² = 0·047), followed by general health rating (β = –0·128, P = 0·008, R² = 0·068), smoking status (β = –0·112, P = 0·019, R² = 0·084) and finally fruit frequency (β = –0·109, P = 0·025, R² = 0·095). Overall, the model predicted 8·4 % of the variance in depression scores (F (5, 416) = 8·735, P < 0·001, adjusted total = 0·084). Secondly, for wellbeing scores, the significant stepwise model selected general health rating for entry first (β = 0·255, P < 0·001, R² = 0·064), then fruit frequency (β = 0·187, P < 0·001, R² = 0·096) and finally age (β = 0·140, P = 0·003, R² = 0·116). Overall, the model predicted 10·9 % of the variance in wellbeing scores (F (3, 418) = 18·199, P < 0·001, adjusted total = 0·109). Finally, for anxiety scores, the significant stepwise model selected age for entry first (β = –0·311, P < 0·001, R² = 0·105), then general health rating was added (β = –0·136, P = 0·003, R² = 0·130), followed by smoking status (β = –0·139, P = 0·002, R² = 0·148), savoury snacking (β = 0·127, P = 0·005, R² = 0·163) and finally sex (β = 0·101, P = 0·029, R² = 0·172). Overall, the model predicted 16·2 % of the variance in anxiety scores (F (5, 416) = 17·296, P < 0·001, adjusted total = 0·162).

Table 3. The significant stepwise regression models for the relationship between fruit frequency, savoury snacking and psychological health, including covariates (n 428)

Note: Depression: Step 1 exercise; Step 2 age was added; Step 3 general health rating added; Step 4 smoking status added; Step 5 fruit frequency added. Wellbeing: Step 1 general health rating; Step 2 fruit frequency was added; Step 3 age added. Anxiety: Step 1 age; Step 2 general health rating was added; Step 3 smoking status added; Step 4 savoury snacking added; Step 5 sex added.

Focussing on dietary intake within these models, fruit frequency was a significant negative predictor of depression scores (β = –0·109, P = 0·025) and independently contributes 1·1 % to the variance explained ( change value at step 5 = 0·011), with an adjusted, standardised β estimate of –0·099 (after adjusting for exercise, age, general health rating and smoking status retained in the significant stepwise model). Fruit frequency was also a significant positive predictor of positive psychological wellbeing scores (β = 0·187, P < 0·001) and independently contributes 3·2 % to the variance explained ( change value at step 2 = 0·032), with an adjusted, standardised β estimate of 0·180 (after adjusting for general health rating and age retained in the significant stepwise model). The unstandardised β values (presented in Table 3) show that for every 1 unit increase in the frequency of fruit consumption (e.g. from 4–6 times a week to 1–2 times a day), depression scores decrease by 0·188, while positive wellbeing scores increase by 0·916. Furthermore, savoury snacking was a significant positive predictor of anxiety scores (β = 0·127, P = 0·005) and independently contributes 1·4 % to the variance explained ( change value at step 4 = 0·014), with an adjusted, standardised β estimate of 0·131 (after adjusting for age, general health rating, smoking status and sex retained in the significant stepwise model). For every 1 unit increase in the frequency of savoury snacking (e.g. from 2–3 times a week to 4–6 times a week), anxiety scores increase by 0·362 (unstandardised β value). After including the covariates in the stepwise regression, all other models for snacking, vegetable and fruit portions were no longer significant predictors of psychological health (all ps > 0·05).

Mediation results

Mediation analyses were applied to follow-up on the significant regression analyses presented above. Cognitive failures mediated the relationship between savoury snacking and psychological health whilst controlling for general health rating and exercise. Specifically, savoury snacking significantly positively predicted cognitive failures (a pathway; B = 1·93, P < 0·001). As cognitive failures increased, so did depression (b pathway; B = 0·10, P < 0·001), stress (b pathway; B = 0·23, P < 0·001) and anxiety scores (b pathway; B = 0·16, P < 0·001), whereas wellbeing scores decreased (b pathway; B = –0·23, P < 0·001). Savoury snacking was not a direct significant predictor of psychological health when cognitive failures were held constant (for all c’ pathways, P > 0·05); however, savoury snacking was a significant predictor of stress (P < 0·05) and anxiety (P < 0·001) when the indirect and direct pathways were combined (c total pathway). Further, examining the indirect effect (ab pathway) indicated that mediation had occurred for: depression (ab = 0·188, 95 % CI [0·082, 0·309]); anxiety (ab = 0·295, 95 % CI [0·134, 0·472]); stress (ab = 0·436, 95 % CI [0·196, 0·691]) and wellbeing (ab = –0·433, 95 % CI [–0·721, –0·188]). No further significant mediation was identified, either for the other predictors (dietary predictors) or for the other theoretical mediator (inhibitory control). The models did reveal that higher SsRT (indicative of poorer efficiency inhibiting a response) was associated with a significant increase in depression (P < 0·001) and stress (P < 0·05) scores.

Discussion

This study assessed the direct relationship between dietary intake and psychological health, while extending existing literature by evaluating both frequency and portions of fruit and vegetable consumption, separately. It also explored a novel mediation model to evaluate the indirect relationship between dietary intake and psychological health via cognitive processes. Results showed that after including a range of covariates, fruit frequency negatively predicted depression scores and positively predicted psychological wellbeing scores. By contrast, savoury snacking positively predicted anxiety scores. Although, cognitive failures did not mediate the relationship between either FVI or sweet snacking and psychological health, mediation was observed for savoury snacking, whereby more frequent consumption of savoury snacks was associated with greater cognitive failures, and in turn, reduced psychological health, including increased symptoms of depression, stress and anxiety, and lower positive psychological wellbeing. Inhibitory control did not mediate any of the relationships between dietary intake and psychological health. The results emphasise that frequency of fruit consumption and savoury snacking could be potential targets for improving psychological health at the individual level, which in turn could have larger gains (e.g. health, social and economic) at a population level.

The positive relationship between FVI and psychological health has been consistently reported in the literature(Reference Saghafian, Malmir and Saneei57,Reference Blanchflower, Oswald and Stewart-Brown58) ; however, few studies have analysed FVI separately(Reference Tuck, Farrow and Thomas59,Reference Dharmayani, Juergens and Allman-Farinelli60) , and even fewer have compared frequency and portions (quantity)(Reference Ocean, Howley and Ensor10). The current results provide further nuance by revealing that the frequency with which fruit is consumed, but not the portions consumed during a typical week, negatively predicted depression and positively predicted psychological wellbeing, after including covariates. This suggests that how often we consume fruit may be more important than the total amount we consume. Indeed, poor mental health has been associated with less frequent intake of fresh fruits among women and men(Reference Sarlio-Lähteenkorva, Lahelma and Roos61). Additionally, frequent consumption of fruit snacks over a 2-week period has been shown to improve positive psychological wellbeing(Reference Conner, Brookie and Carr62). Hence, the present findings support the notion that frequently consuming nutrient-rich fruits may be more important than the quantity of consumption for psychological health(Reference Ocean, Howley and Ensor10); however, experimental studies, varying frequency and portions of fruit consumed, are required to test this directly.

Although fruit frequency predicted psychological health, it is notable that neither the frequency of consumption, or the portions of vegetables consumed, were significant predictors when including age, sex, BMI, exercise, general health rating, smoking status and alcohol intake as covariates. This highlights not only the importance of covariate analysis but also the significance of assessing FVI individually. Furthermore, FVI are often consumed in different environmental contexts as vegetables are typically found to be consumed with family members at home during meal times, while fruits are typically consumed as snacks outside the home and throughout the day(Reference Chawner, Blundell-Birtill and Hetherington63). Although the null association for vegetables and psychological health was not predicted, it is not necessarily surprising. Recent research suggests that the relationship between FVI and psychological health is stronger for raw fruit and vegetables compared with cooked or canned(Reference Brookie, Best and Conner64), and that frequent intake of fruit may involve greater consumption in raw form (for instance, snacking on whole fruits), which may maximise the absorption of nutrients with antioxidant properties, thus having a more potent influence on psychological health(Reference Rooney, McKinley and Woodside65).

Further, precise quantities (daily portions) of vegetable intake may be required to observe an influence on psychological health. Previous research demonstrated that an association between vegetable portions consumed and reduced symptoms of depression only occurred at higher levels of intake every day (≥ 5·0 portions/d)(Reference Mihrshahi, Dobson and Mishra11). Thus, the current sample’s low average vegetable consumption (2·3 portions/day), offers an alternative explanation of the null association, here. Further, recent findings suggest that certain types of fruits and vegetables may be more effective in reducing symptoms of depression. For instance, intakes of tomatoes, dark-green vegetables, berries and fruits were more strongly negatively related to symptoms of depression, than other vegetables and dried fruits(Reference Sun, Li and Li66). Therefore, higher quantities of certain categories of vegetables may need to be consumed in order to observe benefits to psychological health.

In contrast to the patterns described above, savoury snacking positively predicted anxiety scores (but neither FVI or sweet snacking predicted anxiety scores). Previous research has explored snacking more broadly by combining sweet and savoury intake(Reference Sangsefidi, Lorzadeh and Hosseinzadeh13); thus, the current results shed light on a specific link between savoury snacking and anxiety, after including covariates. These results support previous work demonstrating that poorer dietary intake including higher salty snack and fast-food consumption is associated with greater anxiety(Reference Zahedi, Kelishadi and Heshmat67). Of course, it is necessary to point out that the direction of this relationship is uncertain, as a strategy to cope with anxiety often involves increased consumption of nutrient-poor foods(Reference Adan, van der Beek and Buitelaar68). However, it may be the case that savoury snacking is more problematic for psychological health than sweet snacking because the current results suggest that a robust direct and indirect relationship exists. Hence, similar to fruit and vegetables, assessing the individual contributions of sweet and savoury snacking to our psychological health appears to be warranted.

Following on from this finding, we also revealed that more frequent savoury snacking was associated with an increase in cognitive failures, which were in turn associated with increased symptoms of depression, stress and anxiety, but also a decrease in wellbeing. This mediation supports previous findings that snacking on nutrient-poor processed foods, such as crisps, is associated with increased cognitive and psychological problems(Reference Sangsefidi, Lorzadeh and Hosseinzadeh13,Reference Smith and Rogers14) and builds on findings that reported a positive relationship between frequent unhealthy snacking and higher cognitive failures and stress(Reference Chaplin and Smith69). Research shows that subjective memory concern during daily activities is a significant predictor of psychological wellbeing, anxiety and symptoms of depression(Reference Gates, Valenzuela and Sachdev28). It is possible that savoury snacking may have a negative effect on psychological health, via a decrease in cognition related to general memory lapses due to saturated fat content(Reference Francis and Stevenson22,Reference Gibson, Barr and Jeanes70) . There is an abundance of evidence from animal models showing that high saturated fat diets specifically reduce memory function(Reference Wu, Ying and Gomez-Pinilla71) which speaks to the causal link between diet and cognition. Hence, the memory errors reflected in the cognitive failures measure may be the cognitive mechanisms by which processed food intake indirectly reduces psychological health. Indeed, that inhibitory control did not mediate this effect may suggest that specific cognitive processes are involved, relating to memory. Thus, future work should focus on delineating the precise cognitive mechanisms involved, and establishing causality, using an experimental approach.

Strengths of this study included that it was a large national sample of the UK population, including a wide participant age range with each sex well-represented. Also, instead of a general measure of psychological health, multiple aspects were measured as discrete outcomes (depression, anxiety, stress and positive psychological wellbeing), and similarly, mutiple aspects of dietary intake were assessed (fruit and vegetables separately; frequency and portions, etc.) providing greater precision. Nevertheless, there are limitations to be considered. This study provides a credible model by which we can infer causal relationships, but due to the cross-sectional design, further prospective and experimental work is required to test causality and temporality robustly. Furthermore, the food frequency measure did not identify the number of snacks consumed; therefore, the portions of sweet and savoury snacks consumed were not evaluated. Relatedly, this measure could also benefit from validation with biomarker levels to establish the level of agreement. Finally, the testing environment could not be precisely controlled, as this was an online study; thus, the results must be considered with that caveat in mind. However, the results suggest that psychological health can be directly and indirectly influenced by specific nutrient-rich (fruit) and nutrient-poor foods (savoury snacks) which contribute to an evidence base for developing effective preventive strategies in public health. Promoting dietary habits that lead to better psychological health could improve individual wellbeing, alleviate strain on healthcare systems and reduce the economic cost associated with both poor psychological health and cognitive failure(Reference Adan, van der Beek and Buitelaar68).

In conclusion, this study identified that frequent fruit consumption has a direct positive relationship, whereas savoury snacking has a direct negative relationship, with elements of our psychological health. This study also revealed that cognitive processes may be one of the mechanisms by which our dietary intake affects our psychological health, but this appears to be limited to savoury snack foods. Given that dietary intake is associated with psychological health, a more precise understanding of how, and to what extent, our diet affects psychological health could help to inform novel nutritional approaches to enhance it. Future work should look to experimentally test causality and examine the potential options for intervention.

Acknowledgements

The authors would like to thank Gorilla (www.gorilla.sc) for the grant awarded to use the research platform to collect the data.

This research was supported and funded by Aston University. Nicola-Jayne Tuck was awarded a Gorilla (www.gorilla.sc) Early Career Researcher Grant to use the platform to complete this research. Gorilla had no role in the design, analysis or writing of this article.

N. J. T., C. F. and J. M. T. formulated the research questions and designed the study. N. J. T. carried out the study. N. J. T. and J. M. T. analysed the data. N. J. T., C. F. and J. M. T. interpreted the findings and wrote the paper.

There are no conflicts of interest.

References

World Health Organization (2019) The WHO Special Initiative for Mental Health (2019–2023): Universal Health Coverage for Mental Health. World Health Organization. https://apps.who.int/iris/handle/10665/310981 (accessed May 2021).Google Scholar
Jané-Llopis, E, Anderson, P, Stewart-Brown, S, et al. (2011) Reducing the silent burden of impaired mental health. J Health Commun 31, 5974.CrossRefGoogle Scholar
Głąbska, D, Guzek, D, Groele, B, et al. (2020) Fruit and vegetable intake and mental health in adults: a systematic review. Nutrients 12, 115149.CrossRefGoogle ScholarPubMed
Nicolaou, M, Colpo, M, Vermeulen, E, et al. (2020) Association of a priori dietary patterns with depressive symptoms: a harmonised meta-analysis of observational studies. Psychol Med 50, 18721883.CrossRefGoogle ScholarPubMed
Lai, JS, Hiles, S, Bisquera, A, et al. (2014) A systematic review and meta-analysis of dietary patterns and depression in community-dwelling adults. Am J Clin Nutr 99, 181197.CrossRefGoogle ScholarPubMed
Gomez-Pinilla, F & Nguyen, TT (2012) Natural mood foods: the actions of polyphenols against psychiatric and cognitive disorders. Nutr Neurosci 15, 127133.CrossRefGoogle ScholarPubMed
Gibson-Smith, D, Bot, M, Brouwer, IA, et al. (2020) Association of food groups with depression and anxiety disorders. Eur J Nutr 59, 767778.CrossRefGoogle ScholarPubMed
Mikolajczyk, RT, El Ansari, W & Maxwell, AE (2009) Food consumption frequency and perceived stress and depressive symptoms among students in three European countries. Nutr J 8, 18.CrossRefGoogle ScholarPubMed
Mujcic, R & J Oswald, A (2016) Evolution of well-being and happiness after increases in consumption of fruit and vegetables. Am J Public Health 106, 15041510.CrossRefGoogle ScholarPubMed
Ocean, N, Howley, P & Ensor, J (2019) Lettuce be happy: a longitudinal UK study on the relationship between fruit and vegetable consumption and well-being. Soc Sci Med 222, 335345.CrossRefGoogle Scholar
Mihrshahi, S, Dobson, AJ & Mishra, GD (2015) Fruit and vegetable consumption and prevalence and incidence of depressive symptoms in mid-age women: results from the Australian longitudinal study on women’s health. Eur J Clin Nutr 69, 585591.CrossRefGoogle ScholarPubMed
Nguyen, B, Ding, D & Mihrshahi, S (2017) Fruit and vegetable consumption and psychological distress: cross-sectional and longitudinal analyses based on a large Australian sample. BMJ Open 1, 110.Google Scholar
Sangsefidi, ZS, Lorzadeh, E, Hosseinzadeh, M, et al. (2020) Dietary habits and psychological disorders in a large sample of Iranian adults: a population-based study. Ann Gen Psychiatr 19, 110.CrossRefGoogle Scholar
Smith, AP & Rogers, R (2014) Positive effects of a healthy snack (fruit) v. an unhealthy snack (chocolate/crisps) on subjective reports of mental and physical health: a preliminary intervention study. Front Nutr 16, 110.Google Scholar
Sánchez-Villegas, A, Verberne, L, De Irala, J, et al. (2011) Dietary fat intake and the risk of depression: the SUN Project. Plos One 26, 162168.Google Scholar
Liu, C, Xie, B, Chou, CP, et al. (2007) Perceived stress, depression and food consumption frequency in the college students of China Seven Cities. Physiol Behav 92, 748754.CrossRefGoogle ScholarPubMed
Monti, JM, Moulton, CJ & Cohen, NJ (2015) The role of nutrition on cognition and brain health in ageing: a targeted approach. Nutr Res Rev 28, 167180.CrossRefGoogle Scholar
Stevenson, RJ (2017) Psychological correlates of habitual diet in healthy adults. Psychol Bull 143, 5390.CrossRefGoogle ScholarPubMed
Young, LM, Gauci, S, Scholey, A, et al. (2020) Self-reported diet quality differentiates nutrient intake, blood nutrient status, mood, and cognition: implications for identifying nutritional neurocognitive risk factors in middle age. Nutrients 12, 120.CrossRefGoogle ScholarPubMed
Polidori, MC, Praticó, D, Mangialasche, F, et al. (2009) High fruit and vegetable intake is positively correlated with antioxidant status and cognitive performance in healthy subjects. J Alzheimers Dis 17, 921927.CrossRefGoogle ScholarPubMed
Beilharz, JE, Maniam, J & Morris, MJ (2015) Diet-induced cognitive deficits: the role of fat and sugar, potential mechanisms and nutritional interventions. Nutrients 7, 67196738.CrossRefGoogle ScholarPubMed
Francis, HM & Stevenson, RJ (2011) Higher reported saturated fat and refined sugar intake is associated with reduced hippocampal-dependent memory and sensitivity to interoceptive signals. Behav Neurosci 125, 943955.CrossRefGoogle ScholarPubMed
Attuquayefio, T & Stevenson, RJ (2015) A systematic review of longer-term dietary interventions on human cognitive function: emerging patterns and future directions. Appetite 95, 554570.CrossRefGoogle ScholarPubMed
Boals, A & Banks, JB (2012) Effects of traumatic stress and perceived stress on everyday cognitive functioning. Cogn Emot 26, 13351343.CrossRefGoogle ScholarPubMed
Payne, TW & Schnapp, MA (2014) The relationship between negative affect and reported cognitive failures. Depress Res Treat 1, 17.Google Scholar
Kaiser, S, Unger, J, Kiefer, M, et al. (2003) Executive control deficit in depression: event-related potentials in a Go/Nogo task. Psychiatr Res Neuroimaging 122, 169184.CrossRefGoogle Scholar
Gotlib, IH & Joormann, J (2010) Cognition and depression: current status and future directions. Annu Rev Clin Psychol 6, 285312.CrossRefGoogle ScholarPubMed
Gates, N, Valenzuela, M, Sachdev, PS, et al. (2014) Psychological well-being in individuals with mild cognitive impairment. Clin Interv Aging 9, 779.CrossRefGoogle ScholarPubMed
Toluna. https://www.toluna-group.com/ (accessed May 2021).Google Scholar
Faul, F, Erdfelder, E, Buchner, A, et al. (2009) Statistical power analyses using G × Power 3.1: tests for correlation and regression analyses. Behav Res Meth 41, 11491160.CrossRefGoogle ScholarPubMed
Hötting, K & Röder, B (2013) Beneficial effects of physical exercise on neuroplasticity and cognition. Neurosci Biobehav Rev 37, 22432257.CrossRefGoogle ScholarPubMed
Molteni, R, Wu, A, Vaynman, S, et al. (2004) Exercise reverses the harmful effects of consumption of a high-fat diet on synaptic and behavioral plasticity associated to the action of brain-derived neurotrophic factor. Neuroscience 123, 429440.CrossRefGoogle Scholar
Schuch, FB, Vancampfort, D, Rosenbaum, S, et al. (2016) Exercise improves physical and psychological quality of life in people with depression: a meta-analysis including the evaluation of control group response. Psychiatr Res 241, 4754.CrossRefGoogle ScholarPubMed
Cappelleri, JC, Bushmakin, AG, Gerber, RA, et al. (2009) Psychometric analysis of the Three-Factor Eating Questionnaire-R21: results from a large diverse sample of obese and non-obese participants. Int J Obes 33, 611620.CrossRefGoogle ScholarPubMed
Cade (2008). Nutritional Epidemiology Group, School of Food Science and Nutrition. SFFFQ: Short Form Dietary Questionnaire. SFDietQuest. Leeds: University of Leeds Google Scholar
Cleghorn, CL, Harrison, RA, Ransley, JK, et al. (2016) Can a dietary quality score derived from a short-form FFQ assess dietary quality in UK adult population surveys? Public Health Nutr 19, 29152923.CrossRefGoogle ScholarPubMed
Zigmond, AS & Snaith, RP (1983) The hospital anxiety and depression scale. Acta Psychiatr Scand 6, 361370.CrossRefGoogle Scholar
Bjelland, I, Dahl, AA, Haug, TT, et al. (2002) The validity of the hospital anxiety and depression scale: an updated literature review. J Psychosom Res 52, 6977.CrossRefGoogle ScholarPubMed
Cohen, S, Kamarck, T & Mermelstein, R (1983) A global measure of perceived stress. J Health Soc Behav 1, 385396.CrossRefGoogle Scholar
Crichton, N (2001) Visual Analogue Scale (VAS). J Clin Nurs 10, 697706.Google Scholar
Tennant, R, Hiller, L, Fishwick, R, et al. (2007) The Warwick-Edinburgh mental well-being scale (WEMWBS): development and UK validation. Health Qual Life Outcomes 5, 13.CrossRefGoogle ScholarPubMed
Stewart-Brown, S & Janmohamed, K. (2008) Warwick-Edinburgh Mental Well-Being Scale. User Guide. Version 1. http://www.ocagingservicescollaborative.org/wp-content/uploads/2014/07/WEMWBS-User-Guide-Version-1-June-2008.pdf (accessed May 2021).Google Scholar
Broadbent, DE, Cooper, PF, FitzGerald, P, et al. (1982) The cognitive failures questionnaire (CFQ) and its correlates. Br J Clin Psychol 21, 16.CrossRefGoogle ScholarPubMed
Wallace, JC, Kass, SJ & Stanny, CJ (2002) The cognitive failures questionnaire revisited: dimensions and correlates. J Gen Psychol 129, 238256.CrossRefGoogle ScholarPubMed
Martin, M (1983) Cognitive failure: everyday and laboratory performance. Bull Psychon Soc 21, 97100.CrossRefGoogle Scholar
Logan, GD (1994) On the ability to inhibit thought and action: a users’ guide to the stop signal paradigm. In Inhibitory Processes in Attention, Memory and Language, pp. 184–239 [D Dagenbach and TH Carr, editors]. San Diego, CA: Academic Press.Google Scholar
Padilla, C, Perez, L, Andres, P, et al. (2013) Exercise improves cognitive control: evidence from the stop signal task. Appl Cogn Psychol 27, 505511.CrossRefGoogle Scholar
Bartholdy, S, Dalton, B, O’Daly, OG, et al. (2016) A systematic review of the relationship between eating, weight and inhibitory control using the stop signal task. Neurosci Biobehav Rev 1, 3562.CrossRefGoogle Scholar
Coumans, JM, Danner, UN, Intemann, T, et al. (2018) Emotion-driven impulsiveness and snack food consumption of European adolescents: results from the I. Family study. Appetite 123, 152159.CrossRefGoogle ScholarPubMed
Marazziti, D, Consoli, G, Picchetti, M, et al. (2010) Cognitive impairment in major depression. Eur J Pharmacol 626, 8386.CrossRefGoogle ScholarPubMed
Anwyl-Irvine, A, Massonnié, J, Flitton, A, et al. (2019) Gorillas in our Midst: gorilla.sc, a new web-based experiment builder. Behav Res Methods 52, 388407.CrossRefGoogle Scholar
Verbruggen, F & Logan, GD (2008) Response inhibition in the stop-signal paradigm. Trends Cogn Sci 12, 418424.CrossRefGoogle ScholarPubMed
Congdon, E, Mumford, JA, Cohen, JR, et al. (2012) Measurement and reliability of response inhibition. Front Psychol 3, 110.CrossRefGoogle ScholarPubMed
Hayes, AF (2017) Introduction to Mediation, Moderation, and Conditional Process Analysis: a Regression-Based Approach. New York, NY: Guilford Publications.Google Scholar
Andreou, E, Alexopoulos, EC, Lionis, C, et al. (2011) Perceived stress scale: reliability and validity study in Greece. Int J Environ Res Public Health 8, 32873298.CrossRefGoogle ScholarPubMed
Bartram, DJ, Yadegarfar, G, Sinclair, JM, et al. (2011) Validation of the Warwick–Edinburgh Mental Well-being Scale (WEMWBS) as an overall indicator of population mental health and well-being in the UK veterinary profession. Vet J 187, 397398.CrossRefGoogle ScholarPubMed
Saghafian, F, Malmir, H, Saneei, P, et al. (2018) Fruit and vegetable consumption and risk of depression: accumulative evidence from an updated systematic review and meta-analysis of epidemiological studies. Br J Nutr 119, 10871101.CrossRefGoogle ScholarPubMed
Blanchflower, DG, Oswald, AJ & Stewart-Brown, S (2013) Is psychological well-being linked to the consumption of fruit and vegetables? Soc Indic Res 114, 785801.CrossRefGoogle Scholar
Tuck, NJ, Farrow, C & Thomas, JM (2019) Assessing the effects of vegetable consumption on the psychological health of healthy adults: a systematic review of prospective research. Am J Clin Nutr 110, 196211.CrossRefGoogle ScholarPubMed
Dharmayani, PN, Juergens, M, Allman-Farinelli, M, et al. (2021) Association between Fruit and vegetable consumption and depression symptoms in young people and adults aged 15–45: a systematic review of cohort studies. Int J Environ Res Public Health 18, 780800.CrossRefGoogle ScholarPubMed
Sarlio-Lähteenkorva, S, Lahelma, E & Roos, E (2004) Mental health and food habits among employed women and men. Appetite 42, 151156.CrossRefGoogle ScholarPubMed
Conner, TS, Brookie, KL, Carr, AC, et al. (2017) Let them eat fruit! The effect of fruit and vegetable consumption on psychological well-being in young adults: a randomized controlled trial. PLoS One 12, 171206.CrossRefGoogle ScholarPubMed
Chawner, LR, Blundell-Birtill, P & Hetherington, MM (2020) Predictors of vegetable consumption in children and adolescents: analyses of the UK National Diet and Nutrition Survey (2008–2017). Br J Nutr 15, 12.Google Scholar
Brookie, KL, Best, GI & Conner, TS (2018) Intake of raw fruits and vegetables is associated with better mental health than intake of processed fruits and vegetables. Front Psychol 9, 114.CrossRefGoogle ScholarPubMed
Rooney, C, McKinley, MC & Woodside, JV (2013) The potential role of fruit and vegetables in aspects of psychological well-being: a review of the literature and future directions. Proc Nutr Soc 72, 420432.CrossRefGoogle ScholarPubMed
Sun, J, Li, Z, Li, Y, et al. (2020) Intakes of specific categories of vegetables and fruits are inversely associated with depressive symptoms among adults. J Epidemiol 31, 210219.CrossRefGoogle ScholarPubMed
Zahedi, H, Kelishadi, R, Heshmat, R, et al. (2014) Association between junk food consumption and mental health in a national sample of Iranian children and adolescents: the CASPIAN-IV study. Nutrition 30, 13911397.CrossRefGoogle Scholar
Adan, RA, van der Beek, EM, Buitelaar, JK, et al. (2019) Nutritional psychiatry: towards improving mental health by what you eat. Eur Neuropsychopharmacol 29,13211332.CrossRefGoogle ScholarPubMed
Chaplin, K & Smith, AP (2011) Breakfast and snacks: associations with cognitive failures, minor injuries, accidents and stress. Nutrients 3, 515528.CrossRefGoogle ScholarPubMed
Gibson, EL, Barr, S & Jeanes, YM (2013) Habitual fat intake predicts memory function in younger women. Front Hum Neurosci 7, 112.CrossRefGoogle ScholarPubMed
Wu, A, Ying, Z & Gomez-Pinilla, F (2004) The interplay between oxidative stress and brain-derived neurotrophic factor modulates the outcome of a saturated fat diet on synaptic plasticity and cognition. Eur J Neurosci 19, 16991707.CrossRefGoogle ScholarPubMed
Figure 0

Fig. 1. Mediation paths. All models controlled for general health rating and exercise. SsRT, Stop-signal Reaction Time.

Figure 1

Table 1. Sample demographic information and lifestyle behaviour(numbers and percentages; mean values and standard deviation, n 428)

Figure 2

Table 2. Descriptive results for mood ratings and psychological health (Mean values and standard deviation, n 428)

Figure 3

Table 3. The significant stepwise regression models for the relationship between fruit frequency, savoury snacking and psychological health, including covariates (n 428)