Empirically Derived Dietary Patterns and Attention Deficit/Hyperactivity Disorder (AD/HD) in Children CURRENT

Attention deficit/hyperactivity disorder (AD/HD) is the most common chronic mental and behavioral disorder among children. We aimed to derive major dietary patterns in relation with ADHD through a case-control study. Participants were selected from age-gender matched children and adolescents who were categorized into case (n = 120) and control groups (n = 240). Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition was used to diagnose ADHD. Food frequency questionnaire and principal component analysis were used to measure food intake and identify major dietary patterns, respectively. The snack-fast food dietary pattern significantly increased odds of ADHD in fully adjusted model (odds ratio [OR], 3.30; 95% Confidence Interval [CI], 1.39-7.84; P for trend < 0.001). Fish and low fat dairy products dietary pattern is protectively associated with ADHD (OR, 0.42; 95% CI, 0.19-0.91; P for trend = 0.02). Vegetable and nut dietary pattern showed no significant relation with possibility of ADHD (OR, 0.88; 95% CI, 0.40-1.90; P for trend = 0.53). Children are suggested to reduce intake of snack and fast food dietary pattern and increase fish and low fat dairy products, and legumes to reduce the chance of ADHD.


Introduction
Attention deficit/hyperactivity disorder (AD/HD) is the most common chronic mental and behavioral disorder among children (Izquierdo-Pulido et al.). The prevalence of AD/HD around the world was 7.2% in 2015 and 17% in Iran (Meysamie et al., 2011) that has been growing every year (Thomas et al., 2015, Visser et al., 2010. This disorder is caused by abnormalities in frontal brain regions (Cubillo et al., 2012). Attention deficit, learning disorders, depression, anxiety, obesity, and communication problems are some examples of ADHD comorbidities caused by brain dysfunction (Barbaresi et al., 2013).
A study found that ADHD children have more tendency to eat unhealthy foods, low in nutrients than non ADHD ones (Al-Muammar et al., 2015). As reported in the literature, zinc (Bilici et al., 2004), iron (Konofal et al., 2008), magnesium (Mousain-Bosc et al., 2004), copper (Kiddie et al., 2010), and vitamin D (Kamal et al., 2014) deficiency are common among these children. Previous studies conducted in this area proved the role of nutrients such as zinc (Bilici et al., 2004), iron (Konofal et al., This persistence should be maladaptive and inconsistent with the person's developmental level (Owens and Hoza, 2003). The reliability and validity of this questionnaire was proved previously (Owens and Hoza, 2003).

Assessment of dietary intake
Usual dietary intake was assessed using a validated semi quantitative food-frequency questionnaire (FFQ) (Asghari et al., 2012), which consisted of 186 food items commonly consumed by Iranians and local foods specific to children in Yazd. Participants were also asked about intake of supplements. All questionnaires were administered by a trained dietitian in a face-to-face interview. Parents were interviewed along with their children, they were asked to report the frequency of food items consumed by their children on a daily (i.e., fruits), weekly (i.e., cruciferous vegetables), and monthly (i.e., pizza) basis during the previous year. The reported frequency for each food item was then converted to a daily basis. Portion sizes of consumed foods were converted to grams using household measures. Each food and beverage was then coded according to the protocol and analyzed for the content of energy and other nutrients using Nutritionist IV (N-Squared Computing, Salem, OR, USA), which was designed for Iranian foods. The validity and reliability of the FFQ was evaluated in Tehran lipid and glucose study (TLGS) (Asghari et al., 2012). We categorized foods into 30 groups used in the dietary pattern analysis; the food groups were defined based on the nutrient similarity of foods (Table 1).

Assessment of other variables
To address the important confounding factors, we employed a pair-match design on age and gender.
Variables including age, medical history, current use of medications, family history of ADHD, and screen exposure were investigated through demographic questionnaires. Family history of ADHD was assessed by asking parents if they or any of their siblings used ADHD medications or their ADHD was diagnosed by a psychotherapist. Socioeconomic status was determined as low, moderate, or high based on the education (undergraduate, postgraduate, and doctorate) and acquisition (house ownership or not). Physical activity was also assessed through a valid questionnaire (Aadahl andJørgensen, 2003, Kelishadi et al., 2007) and categorized to less than once a week, 2-3 times a week, 3-5 times a week, and more than 5 times a week.

Statistical analyses
To

Results
The principal component analysis retained three primary factors (dietary patterns). The factor loadings associated with each pattern are shown in Table 3. The high positive loadings indicate strong associations between the given food groups and patterns, whereas the negative loadings suggest inverse associations with the patterns. Each pattern was labeled according to the food groups with high absolute loadings. Factor 1, with high loads of snack, fast foods, refined grains, and sweetened beverages was labeled "snack-fast food" dietary pattern. Factor 2, with high loadings of fish, low fat dairy, and legumes was tagged "fish-low fat dairy" dietary pattern. Factor 3, with high positive loads of yellow-red vegetables, other vegetables, pickles, and nuts, but high negative loadings for refined grain and sweets was named "vegetable-nuts" dietary pattern. In general, the three dietary patterns accounted for 18.75% of the variance in food intake. Only items with correlation coefficients ≥ 0·20 are presented.
The general characteristics of study participants are shown in Table 2. The participants with ADHD were more likely to use medication and electronic screens such as computer, laptop, TV, etc.; they were also more physically active than the healthy controls. Moreover, they had higher likelihood of a family history of ADHD. The children with ADHD and controls weren't different in BMI, household income, parental education, and sleep hour per day. Sugar and iron intakes weren't significantly different between the case and control groups. As represented in Table 2, energy intake, including consumption of protein, carbohydrate, fat, and vegetable oil was significantly different between ADHD children and controls. The distribution of characteristics by dietary pattern score quartile is presented in Table 4. Increase of scores in the snack-fast food dietary pattern was significantly correlated with the decrease of age (Pvalue = 0.03). Increase of scores in the snack-fast food and fish-low fat dairy products dietary patterns were significantly correlated with a decrease in mean of refined grains intake (P-value < 0.001; Pvalue < 0.001, respectively). Furthermore, legumes' mean intake significantly decreased with fish-low fat dairy products dietary pattern scores and increased with snack-fast food dietary pattern scores (Pvalue < 0.001; P-value < 0.001, respectively). The mean intake of leafy green, yellow-red, other vegetables, and fruits significantly increased with fish-low fat dairy products dietary pattern scores (Pvalue < 0.001; P-value = 0.007, P-value = 0.03, P-value < 0.001, respectively). Table 4 Dietary intakes by quartile (Q) categories of dietary pattern scores in ADHD and non ADHD children aged (7)(8)(9)(10)(11)(12)(13)  Increase of scores in all dietary patterns was significantly correlated with decreased energy intake, carbohydrate, protein, and fat (P-value < 0.001, respectively).
Fatty acids were significantly associated with dietary pattern scores. The vegetable-nuts dietary pattern score was associated with a high intake of MUFAs and PUFAs, while the scores related to snack-fast food and fish-low fat dairy products dietary patterns were associated with low intakes of them.
Decrease of iron was significantly associated with vegetable-nuts and snack-fast food dietary patterns' scores, however, this relation is inverse in fish-low fat dairy products dietary pattern. Mean intake of zinc significantly decreased by increase of all dietary patterns scores.
The association between dietary food patterns and odds of developing ADHD is represented in Table 5. The univariate conditional logistic regression analyses showed a significant linear association between the snack-fast food dietary pattern and the likelihood of developing ADHD (P for trend = 0.004). It was also observed that the fish-low fat dairy products' dietary pattern significantly decreased the risk of ADHD (P for trend = 0.021) (  0.72 † Adjusted for energy intake. ‡ Adjusted for energy intake, socioeconomic status, and family history of ADHD. § Adjusted for energy intake, socioeconomic status, family history of ADHD, and BMI. * Adjusted for energy intake, socioeconomic status, family history of ADHD, BMI, sugar intake, monounsaturated fatty acids, polyunsaturated fatty acids, vitamins and minerals. Conditional logistic regression models were used to estimate odds ratios with 95% confidence intervals for all such values. **All comparisons were made in reference to the first quintile (Q) of the corresponding pattern.

Discussion
Having investigated the relationships between dietary patterns and ADHD in gender and age matched case-control study; we observed that the fish-low fat dairy dietary pattern was protectively associated with ADHD. In addition, the snack-fast food dietary pattern increased the odds of ADHD. These associations persisted in multivariate models when a wide range of potential confounding variables were adjusted. To the best of our knowledge, the present study is among the first investigations We found that the snack-fast food dietary pattern was related with increased odds of hyperexcitability in children and remained significant after adjustment of energy and other potential confounders. In the same regard, a study on Chinese children (Zhou et al., 2016) reached the same association in the fast food-sweet dietary pattern. Moreover, in line with results of the current research, a study carried out over 1046 Australian children concluded that a western dietary pattern, which is, similar to our snack-fast food dietary pattern highly loaded with red meat, processed meat, animal fat, refined grains, and soft drinks, was associated with a higher likelihood of ADHD symptoms (Howard et al., 2011). Another study found that western dietary pattern with increased intakes of take-away foods, confectionary, and red meat was significantly associated with internalizing and externalizing behavior disorders in 1324 Australian adolescents (Oddy et al., 2009). In the same line with our conclusions, a cross-sectional Norwegian study found that junk-convenient dietary pattern, consisting of energy-dense and processed foods increased the likelihood of ADHD in adolescents (Oellingrath et al., 2014). Moreover, consumption of sugar-sweetened beverages highly prevalent in snack-fast food dietary patterns were adversely associated with childhood ADHD (Yu et al., 2016). In addition, a cross-sectional research among Iranian children indicated that a sweet and fast food dietary pattern increases the odds of ADHD. But no significant associations were observed between the western dietary pattern and ADHD in this study (Azadbakht and Esmaillzadeh, 2012). This can be due to more similarity of fast food dietary pattern ingredients highly loaded with junk foods and snacks with our snack-fast food dietary pattern than their western pattern, which contained red meat isoprostanes) (Simic and Karel, 2013) are abundant in snack-fast food dietary pattern ingredients.
They are related with the membrane-associated pathologies in the central nervous system such as ADHD and affect the neurotransmitter functioning that has an important role in ADHD (Bulut et al., 2007, Yumru et al., 2009). The snack-fast food dietary pattern contains refined carbohydrates that can be related with elevated C-reactive protein (Maes, 1999). This pattern also consists of junk foods that contain high amounts of food additives. There is some overlap between allergy and ADHD mechanisms (Melamed and Heffron, 2016). So, ADHD symptoms may exacerbate by releasing non immunoglobulin E-dependent histamine from circulating basophils (Supramaniam and Warner, 1986, Murdoch et al., 1987b, Murdoch et al., 1987a. In addition, the histamine risk alleles make the child more vulnerable to the effects on behavior of food additives in the diet (Stevenson et al., 2010).
Consumption of artificial food additives, especially food colors in the sweetened beverages, candies, and biscuits in snack-fast food dietary pattern has also been associated with gene polymorphisms of histamine HNMT T939C and HNMT Thr105Ile expression. Furthermore, food additives affect brain function directly by increasing beta1 activity in the frontotemporal regions of the brain (Uhlig et al., 1997).  (Patrick and Ames, 2015). This might contribute to the favorable inverse association of the fish-low fat dairy dietary pattern and ADHD. Moreover, folate and B vitamins are cofactors in the methylation process of homocysteine to methionine that plays a key role in the production of monoamine transmitters (Crider et al., 2012, Kamphuis et al., 2008. This dietary pattern may also increase the likelihood of ADHD, since it is loaded with low fat dairy products and milk as one of the most common foods provokes allergy and consequently ADHD (Wright andTruelove, 1965, Marshall, 1989). Results of the current study regarding this dietary pattern showed that the provoking effect of milk may be masked by fish consumption. The nutrient-adjusted model for ADHD across quartile categories of the fish-low fat dairy dietary pattern revealed that this association was modified by minerals and fatty acids to some extent but still remained significant. Thus, the effect of this pattern is independent of the PUFAs' effect related with decreased ADHD.
Vegetable-nut dietary pattern containing different kinds of vegetables, nuts, and pickles did not have any significant associations with ADHD. The results of a cohort study (Howard et al., 2011) also showed that a healthy dietary pattern including vegetables, fruits, and fish did not have any protective relationship. In addition, the findings of a case-control study on Korean children (Woo et al., 2014) did not indicate any significant association between their traditional dietary pattern and ADHD.
This was similar to results of our vegetable-nut dietary pattern that contained high amounts of vegetable, fruit, and condiments and had low amounts of fish and Kimchi (a Korean traditional food).
This study used three nonconsecutive 24-h recalls for assessing dietary intake. Another case-control study showed that vegetable-fruit dietary pattern containing different kinds of vegetables and fruits did not have any significant association with ADHD (Zhou et al., 2016). The considerable difference between this pattern and the one investigated in the present study is that nuts are not highly loaded in this study but the results of them are similar. In contrast to our results, an Australian cohort study showed that healthy dietary pattern with high intakes of leafy green vegetables and fresh fruit was significantly associated with improvement of behavioral scores in adolescents (Oddy et al., 2009). This discrepancy was due to different study designs and different components of healthy dietary patterns. Further, another study showed that children whose infancy diet was characterized by high consumption of fruits, vegetables, and home-prepared foods ('infant guidelines' dietary pattern) had better memory performance at age 4. But this study population consisted of infants who were followed until preschool age (Gale et al., 2009). The non-significant relation between this pattern and ADHD may be due to negative effects caused by high amounts of salt in pickle which counteracts with the improving effects of vegetables and nuts.
As reported in a systematic review, the associations of dietary patterns with psychological disorders are gender dependent (Andersen and Teicher, 2000). This brain disorders have a higher male incidence, so we matched this variable in this study.
The present study enjoys several strengths. To the best of our knowledge, it is the first case-control study dealing with the association between major dietary patterns and ADHD in a Middle Eastern country. Furthermore, we controlled for a wide range of confounders, especially age and gender that might affect psychological conditions. Moreover, cases with ADHD were recently diagnosed.
Some limitations in the interpretation of our findings should also be taken into account. First of all, due to the case-control design of the study, we cannot confer causality. We used factor analysis to identify dietary patterns. This method includes several subjective decisions, such as the consolidation of food items into food groups, the number of extracted factors, the rotation method, and factors' labeling. As a case-control study, dietary intakes can be affected by an individual's health status and social background. Results could differ by ADHD types, however, Information about ADHD type was not collected for subgroup analysis due to small sample size.
Another potential limitation is the measurement error, which is a recognized feature of all dietary assessment methods. Due to the use of an FFQ, misclassification of study participants is another concern. Furthermore, we could not exclude the possibility of residual confounding in the analysis due to unmeasured or imprecisely measured factors.

Conclusion
We observed that the fish-low fat dairy dietary pattern was protectively associated with ADHD. In addition, the snack-fast food dietary pattern increased the odds of ADHD. General population is recommended to consume fish, low fat dairy products, and legumes to reduce the chance of ADHD.
Also, they are suggested to reduce intake of snack and fast food dietary pattern. However, prospective studies are needed to establish this result.

Declarations
Ethics approval and consent to participate: The ethics committee of ShahidSadoughi University of Medical Sciences and Health Services, Yazd, Iran also approved the study with the code number IR.SSU.SPH.REC.1395.158. Furthermore, informed consents were taken from all the participants.

Consent for publication: Not applicable.
Availability of data and material: Not applicable.
Competing interests: Author disclosures no Competing interest.

Funding:
The study was financially supported by the Deputy for Research, ShahidSadoughi

University of Medical Sciences (SSUMS).
Authors' contributions: MH created the study concept and design and edited the manuscript, ESH collected data, and prepared the manuscript; MH and ASA statistical analyses, HM and MM managed subjects and edited the manuscript; HM and MM was involved in the design of the study, and edited the manuscript.