Fatty acid status and behavioural symptoms of Attention Deficit Hyperactivity Disorder in adolescents: A case-control study

Attention deficit hyperactivity disorder (ADHD) is primarily characterized by a "persistent pattern of inattention and/or hyperactivity-impulsivity that is more frequent and severe than is typically observed in individuals at a comparable level of development" [1, 2]. The American Psychiatric Association estimates that 3–5% of school aged children have ADHD (DSM-IV), while other sources report higher prevalence rates ranging from 5–13% [3‐6]. ADHD is the most common psychiatric disorder in children and is diagnosed in males two to nine times as often as in females. ADHD shows high comorbidity with several other conditions including learning differences, oppositional defiance disorder (ODD), obsessive compulsive disorder (OCD) and depression [7, 8] For up to 60% of these children, ADHD symptoms and difficulties will persist into adulthood [9, 10].

The cause of ADHD is generally acknowledged to be multifactorial, involving both biological and environmental influence [2, 11]. In the past two decades, there has been an increasing focus particularly on the effects of diet in hyperactivity in children. Researchers have reported that various aspects of a child's diet including food additives, refined sugars, food allergies, minerals and fatty acid metabolism may have adverse effects on behaviour[7, 12, 13]. While there is no definitive proof that any of these is responsible for the spectrum of ADHD symptoms, there is a compelling argument for a role for long-chain polyunsaturated fatty acids.

The processes of elongation and desaturation occur mainly in the liver, but also in the central nervous system, placenta, glial tissue and choroid plexus vasculature[14]. Within the brain, four fatty acids are particularly important; dihomogammalinolenic acid (20:3n-6, DGLA), arachidonic acid (20:4n-6, AA), eicosapentaenoic acid (20:5n-3, EPA) and docosahexaenoic acid (22:6n-3). AA and DHA play a major structural role in neuronal membranes and make up 20% of the dry mass of the brain[11]. In addition the eicosanoid and other fatty acid metabolites of various LC-PUFAs, though at much lower concentrations, could play important roles in brain function [15‐19]. EPA and DGLA play a more minor structural role but are also crucial for normal brain function. Since optimal requirements are not fully known, definitive dietary reference intakes (DRIs) for the omega-3 and omega-6 fatty acids have not yet been determined[20]. However, Petrie and colleagues published recommendations for adequate intake (AI) for boys 9–16 yr as 12–16 g linoleic acid (LA)/d and 1.2–1.6 g α-linolenic acid (ALA)/d. For girls the corresponding amounts were 10–11 LA g/d and 1.0–1.1 ALA g/d[21]. In order to ensure the best biological functions, Bjerve suggests an intake of 900 mg/d EPA and 400 mg/d DHA[22].

A number of the physical and behavioural symptoms of essential fatty acid deficiency mimic some of the symptoms described in typical ADHA patients, therefore it is conceivable, that either dietary deficiency of omega-3 fatty acids, or altered metabolic handling of these fatty acids, could contribute to the abnormalities observed in those affected by ADHD. Several studies have examined fatty acid status in patients with ADHD, but only recently have researchers begun to examine efficacy of high dose supplementation on ADHD behaviours [13, 23‐25].

The purpose of the present study was to compare several parameters in an adolescent ADHD population versus an adolescent control population. Parameters included comparisons of dietary patterns between the groups based on 7-day diet records and analysis of red blood cell fatty acid composition and serum hormone levels as well as the frequency of symptoms of fatty acid deficiency and ADHD-associated behaviours in both populations.

Subject characteristics at visit #1 are given in Table 1. There were no significant differences between the groups for age and sex characteristics or anthropometric measures. A health questionnaire was filled out at the start of the study by the parent/guardian for each subject. The questionnaire collected information regarding ADHD diagnosis, medication use, co-morbid disorders/conditions, vitamin/supplement use, family history of ADHD, allergies, duration of breastfeeding and prevalence of fatty acid deficiency symptoms. Six of eleven subjects in the ADHD group were taking medications (55%), five of eleven presented with a co-morbid learning disorder (45%), and eight of eleven reported a history of ADHD within the family (73%). These three variables were significantly different from the control group with p = 0.016, p = 0.016 and p = 0.001, respectively.

The duration of breastfeeding in infancy was also reported. Subjects in the ADHD group were breastfed an average of 4.14 ± 2.6 (0–7) months, whereas subjects in the control group were breastfeed an average of 10.4 ± 10.9 (3–39) months. With respect to symptoms of fatty acid deficiency, subjects in the ADHD group reported an average of 2.4 ± 3.5 (0–9) symptoms versus 1.6 ± 1.6 (0–4) symptoms in the control group. Neither of these differences between groups was significant (p = 0.077 and p = 0.461, respectively).

The ADHD group was also more likely to have allergies (27% vs. 8%) and less likely to be taking vitamins (18% vs. 50%) when compared to the control group. Again, neither of these differences were significant.

Most clinical studies have focused on fatty acid and other abnormalities associated with ADHD in children (6–12 years) and adult populations (18–65 years) [25‐28]. One recent study used a slightly older population where the average age of subjects was 11 years [24]and in the current study our subjects had a mean age of 14 years. The primary objective of this study was to determine whether abnormalities typically observed in younger and older populations are also observed in adolescents with ADHD when compared to controls.

The Conners' Parent Rating Scale (CPRS:L) was utilized for behaviour assessment in this study. The CPRS:L is well respected and has been used widely among studies assessing ADHD behaviours in children [28‐32]. The Conners' scale used provides an appropriate instrument for measuring behaviours in youths aged 3 to 17 and conveys information that corresponds to the official ADHD criteria in the DSM-IV [33]. In general, T-scores of 65 and above are usually taken to indicate a clinically significant problem. In this study, the ADHD group presented with significantly higher mean T-scores than the control group on ten of the fourteen scales on the CPRS:L. This significant difference was expected as individuals were assigned to the ADHD group based on whether or not they had been previously diagnosed by a physician as having the disorder. It is interesting to note however, that the mean T-scores for the ADHD group did not reach 65 or above on 4 of the 14 measures. Furthermore, there were subjects within the ADHD group who did not present with 65 or higher on any of the scales. Perhaps, these particular individuals have become better able to cope with their symptoms over time, which contributed to the presence of less problematic behaviours. It is also possible that some of these individuals were misdiagnosed with ADHD or "grew out of it". Regardless, had these potentially misdiagnosed individuals not been in the ADHD group here, it is possible that there may have been more significant differences between the groups. There were also subjects in the control group who had multiple scales with T-scores of 65 or more. To this effect, one subject in the control group presented with T-scores greater than 65 on 11 of 14 scales, the fifth most out of all 23 subjects in the study. This same individual was subsequently diagnosed with ADHD shortly after this study ended. As such, all of the data relating to this subject was transferred to the ADHD data pool.

As a group, the ADHD subjects in this study presented with mean T-scores that are comparable to other studies [27‐32], although slightly higher on most scales. The higher T-scores in the current study may simply reflect the fact that similar behaviours in a nine-year old would be given a much higher T score, if exhibited by a 13-year old. The behaviour assessment performed here identified that the ADHD group had higher mean T-scores than controls on over seventy-five percent of the Conners' scales, indicating that problematic behaviours persist into adolescence for the majority of these individuals. The Conners' Parent Rating Scale appears to be an extremely reliable tool when comparing ADHD behaviours among individuals of different gender and age.

The red blood cell phospholipid fatty acid composition data reported here, indicates that adolescent ADHD subjects have lower total omega-3 fatty acids, lower DHA levels and higher linoleic acid levels when compared to age-matched control subjects. The absolute values and degree of difference between groups are similar to those reported elsewhere in studies of younger children and adults with and without ADHD[12, 30, 32, 34] However, we did not see several differences reported by Stevens, which included lower AA and adrenic acid levels, and higher docosapentaenoic acid levels in ADHD subjects when compared to controls[29]. Stevens did report lower DHA levels in the ADHD group but this was not statistically significant in their study[29]. In 2004, Chen and colleagues reported significantly lower levels of AA, LA, DHA and total n-3 in red blood cell phospholipids of young children with ADHD when compared to controls[12]. Similarly, when fatty acid composition of red blood cells were compared in an adult population, Young and her colleagues reported significantly lower DHA and significantly higher total n-6 fatty acids among other measures, in the individuals with ADHD[34]. Thus, while there are some differences in the specific fatty acid changes, the data we report here for adolescents indicates that childhood patterns persist through adolescence and into adulthood with few alterations. In several other studies, authors have reported higher AA/EPA ratios in subjects with ADHD. In our study there was a numerical difference but this was not statistically different; a larger sample size may have been able to pick up this difference as the standard deviation for this estimate was quite large. If indeed the levels of AA and EPA are not different between groups, this would suggest that delta-5 desaturase activity may be normal in the ADHD subjects. Rather, the fact that DHA levels were significantly lower in ADHD subjects may suggest a higher rate of oxidation of this fatty acid in these patients. None-the-less, a lower level of circulating DHA may indicate lower levels in the brain as well.

Two previous studies reported dietary patterns in children with ADHD using three day diet records[12, 29]. Using a complete seven day diet record, we examined the intake patterns of our adolescent subjects. We demonstrate that ADHD subjects consumed higher levels of at least 10 different nutrients, than their control counterparts. This included 25% more energy, and more grams of carbohydrate, total fat, omega-6 fatty acids and trans fatty acids in addition to others. Stevens reported significantly higher intakes of total fat (g) and polyunsaturated fatty acids (g) in ADHD subjects when compared to control subjects[29]. In 2004, Chen and colleagues reported significantly increased iron (mg) and vitamin C (mg) consumption in subjects with ADHD versus controls[12]. As has been previously proposed, it is possible that over-ingestion of fats could interfere with the conversion of parent essential fatty acids (EFAs) to long-chain(LC)-PUFAs[11]. Despite the many dietary differences between our subject groups, they did not differ in either total omega-3 fatty acid consumption, or consumption of the LC-PUFAs. Thus, it is unlikely that dietary differences in omega-3 consumption can explain the significant differences in red blood cell membrane fatty acids observed in the ADHD subjects herein.

Whatever the reason for lower levels of DHA in ADHD subjects, or abnormal LC-PUFA status, an important question is whether this pattern can be normalized through dietary intervention or supplementation. Early supplementation trials with modest levels (eg. 300–500 mg DHA/day) of LC-PUFAs were largely neutral or showed only modest improvement in biochemical or behavioural parameters in ADHD subjects[27, 28, 35, 36]. However, two recent studies using very high levels of long chain omega-3 PUFAs suggest that the plasma fatty acid abnormalities can be reversed and that there may be significant impact on behavioural parameters [24, 25]. Germano and coworkers examined the effects of 0.23 g/kg body weight/day, fish-oil distillate containing a 2:1 ratio of EPA:DHA in 16 ADHD and 31 control children after an 8-week supplementation period[25]. The ADHD group had a significantly higher AA:EPA ratio than the controls at time zero and the supplementation decreased both ratios substantially so that there was no difference between the groups after 8 weeks supplementation. They also reported significant improvements in Inattention and Hyperactivity using the Conner's short-version inventory [25]. Sorgi and colleagues examined the effects of a high dose (initially 16.2 g) EPA/DHA concentrate on nine ADHD children over 8 weeks with dosage adjustment at week 4 depending on AA:EPA ratio [24]. They showed a similar, dramatic decrease in the AA:EPA ratio in the subjects and improvements using the ADHD Symptom Checklist 4, and the short version of the Conners' Parents Rating Scale{Sorgi, 2007 2845/id. These two latest studies suggest that fatty acid supplementation at high levels may be necessary to normalize blood fatty acids, and by assumption brain fatty acids, to achieve improvements in ADHD symptomology. Clearly additional large scale interventions are now warranted.

Regardless of diagnosis in the current study, dietary fat intake was significantly and positively correlated with scores on five of the Conners' scales, positively correlated with total n-6 red blood cell content and negatively correlated with total n-3 red blood cell content. Also, as Chen and colleagues have reported, there was a significantly higher intake of iron in the ADHD group [12]. In the present study, iron intake also displayed a significant and positive correlation to scores on eight of the Conners' scales. It is possible that iron levels in children and adolescents have an effect on behaviours and should therefore be investigated in future studies. Reports have also indicated that children with ADHD may be deficient in magnesium [7, 37]. However, in the present study, ADHD subjects reported slightly higher (but not significant) intakes of magnesium when compared to controls. Previous research had also suggested lower intakes of tyrosine, tryptophan and phenylalanine in individuals with ADHD compared to controls[38] however, this was not the case in the current study where again, if anything intakes for these amino acids were higher in ADHD subjects. From these data, a clear role for minerals and amino acids in ADHD behaviours can not be established and will require further investigation.

This study investigated differences between adolescent populations with and without ADHD with respect to a variety of measures. Despite the small sample size and an imbalance of genders within the groups, several significant differences were reported with respect to overall health, dietary patterns and red blood cell fatty acid compositions. Adolescents with ADHD appear to consume diets more rich in total energy, in addition to specific fats, minerals and other constituents. Although there was no difference in the dietary consumption of n-3 or n-6 fatty acids, adolescents with ADHD did present with significantly lower levels of DHA, total n-3 fatty acids and a lower n-3: n-6 ratio in red blood cell phospholipids. Abnormalites in fatty acid profile were also positively correlated with higher ratings on the Conner's scales. Further research is required to determine the mechanisms by which these fatty acid anomalies occur, and whether indeed supplementation for extended periods with high concentrations of omega-3 fatty acids will positively influence ADHD behaviours in patients of all ages.