To the best of our knowledge, this is the first comparison of erythrocyte GSH and plasma CoQ10, retinyl palmitate, IL-5, IL-8 and IgM between ADHD patients and controls. Significantly increased plasma MDA levels were found in ADHD patients as compared to controls (p < 0.05), as well as a trend for higher urinary 8-OHdG levels. In addition, significantly increased erythrocyte GSH as well as plasma retinyl palmitate and total IgE and IgG levels were found in ADHD patients as compared to controls, along with a trend for more diagnosed allergies. Finally, a trend for lower plasma IL-5 levels was observed. These data are consistent with a role of oxidative stress that affects the immune system in children with ADHD as compared to controls. As a result, pediatric ADHD patients might develop more IgE and non-IgE-mediated allergies.
No differences were found in demographic variables between patients and controls. The expected higher scores on various SEQ (sub)conditions in the ADHD group compared to controls is a confirmation of more comorbidity in ADHD, though high comorbidity rates could also be due to selection of the patient group via a university hospital [40, 41]. However, as conduct disorder for example was linked before to lower antioxidant levels, these comorbidities could influence the results of this study [14].
Dietary habits of patients and controls appeared very similar, as the only significant differences found concern less consumption of sweet milk drinks and fish by ADHD patients (p < 0.05). Moreover, as some p-values might represent a false-positive significant difference due to multiple testing, Bonferroni correction should be taken into account (in this case: 35 questioned items, so significant if p < 0.05/35). Therefore, these differences can be neglected and dietary habits probably do not explain differences observed regarding the biomarkers analyzed. In addition, similar dietary habits reflect a similar socioeconomic status of both groups, despite different recruitment sources [42].
Significantly higher adjusted plasma MDA levels in ADHD patients were observed, though MDA correlated weakly with only the SEQ impulsivity score. After Bonferroni correction for multiple testing, the adjusted difference was only nominally significant. Another study also reported more plasma MDA in pediatric ADHD, but no correlation with ADHD subtype was found [13]. Increased lipid peroxidation in ADHD is also evidenced by raised urinary acrolein-lysine levels as well as by exhaled ethane levels, a marker of omega-3 fatty acid oxidation [15, 16]. Other studies, however, found lower plasma or serum MDA levels [29, 30].
A trend for higher adjusted urinary 8-OHdG levels was observed in ADHD. Literature on oxidative DNA damage in ADHD is contradictory, as increased levels of total DNA damage were found, determined by 8-oxoG in lymphocytes [28], as well as reduced 8-OHdG levels in serum [30].
When correcting for processing time, ADHD patients had significantly higher erythrocyte GSH levels than controls, even after Bonferroni correction for multiple testing, though GSH levels did not correlate with any of the SEQ ADHD scores (p > 0.05). Higher antioxidant levels do not necessarily imply less oxidative stress, as these could be a compensation mechanism for increased oxidative stress [13]. In fact, levels of GSH S-transferase and GSH peroxidase, both necessary for GSH activities, were found to be lower in ADHD before [13, 14, 23]. These reduced levels, leading to low GSH consumption and thus lack of an efficient antioxidant defense, could also explain higher GSH levels.
As retinyl palmitate is the storage form of vitamin A, and low levels were found in plasma, it is questionable whether the borderline significant difference that was observed is biologically relevant. No difference was found for adjusted plasma retinol levels as well as for its precursor ”-carotene, as confirmed by Spahis et al. [29]. No difference regarding retinol is not surprising as retinol concentrations in plasma are strictly regulated due to toxicity at high levels [54-56]. Spahis et al. found significantly higher plasma ‘- and ”-tocopherol levels in ADHD patients [29], but another study, like the present study when accounting for processing time, found no difference regarding serum tocopherol levels [48].
Research generally points towards more oxidative damage in ADHD. Although non-enzymatic antioxidant levels generally do not appear reduced in ADHD, levels or activities of antioxidant enzymes were found to be lower in various studies [13, 14, 23, 49-51]. Therefore, oxidative damage biomarkers appear more reflective of the actual oxidative stress situation compared to antioxidant levels, since even relatively high antioxidant levels can still be too low to balance high oxidant levels [47][52].
In patients as compared to controls and when accounting for processing time, no significantly different plasma cytokine levels were found, despite a trend for lower IL-5 levels. In autism, IL-6 and IFN-” levels correlated negatively with full-scale IQ, verbal comprehension index and working memory index [53], but these cognitive abilities were not assessed in the present study. Nevertheless, significantly higher adjusted plasma IgG and IgE levels in ADHD were observed (nominally significant after Bonferroni correction), without correlation with any of the SEQ ADHD scores [54].
Oxidative and inflammatory processes are closely related [23]. An oxidant/antioxidant imbalance is responsible for changes in the nervous and immune system [23, 55-61]. Moreover, immune cells are an important source of both oxidant and proinflammatory compounds, like reactive oxygen and nitrogen species (ROS and RNS) and inflammatory cytokines, which stimulate NF-”B activation, leading to production of more oxidants and inflammatory compounds, and thereby establishing a vicious circle [23, 61].
ADHD-associated increased oxidative stress might, in case of a chronic state, lead to immune dysfunction resulting in elevated TH2 induction and thereby increased IgE levels. Although IgE is found in every individual, and levels increase with age until reaching a stable level at adulthood, elevated concentrations of total IgE reflect a predisposition to develop IgE-mediated allergic diseases, despite not being directly related to allergic status [62, 63]. Oxidative stress might thus facilitate the development of allergic conditions in ADHD patients [11, 58]. Indeed, the observed higher IgE levels were supported by a trend for more diagnosed allergies in patients than in controls.
It should be noted that the use of medication was an exclusion criterium of this study, so that patients with intense allergy symptoms were likely to be excluded. It is therefore no surprise that no significantly aberrant systemic cytokine levels were found, which would indicate active inflammation. In addition, cytokine concentrations in mucosal secretions are predominantly the result of local cytokine production, causing a potential lack of correlation between low plasma IL-5 levels and, presumably, high concentrations in mucosal secretions reflecting local TH2 expansion [64]. In addition, as the analyzed cytokine levels in both groups were very low, it is questionable whether any observed difference would be biologically relevant. Though elevated oxidative stress can cause higher basal (without antigen stimulation) levels of inflammatory cytokines [65], oxidative stress in ADHD might be too limited for clear effects on plasma cytokine levels.
The shifted immune balance due to oxidative stress , results in a modified humoral immune response. Though IgE levels are more strongly T-cell regulated and have a much shorter half-life than IgG, under chronic conditions also IgG levels will be changed [66]. The observed higher IgG levels in ADHD support involvement of inflammation [54].
Due to the use of various methodologies in literature, it is hard to compare the results of different studies. For example, despite the increased prevalence of atopies in ADHD in literature [17, 67], their association appears based on a non-IgE-mediated mechanism [25, 74-76]. In addition, a role for IgG in ADHD pathophysiology was countered before as well [25]. Moreover, indications of more TH1 cytokines have been found in previous research, e.g. higher detection rates for IFN-‘ and TNF-‘ in cerebrospinal fluid [77]. Another study however reports no significant difference regarding serum TNF-”, IL-1’, IL-6 and IL-10, but did find higher levels of IFN-‘ in ADHD patients than in controls [78]. Further research with consistent methodology is thus required to draw a final conclusion. Nevertheless, cytokines could be important in ADHD as they can pass the blood-brain barrier and affect synaptic plasticity and neurogenesis and can even cause T-cell mediated neuroinflammation [73, 78, 79].
Taking into account the strength of the correlations between SEQ scores and biomarkers, especially impulsivity appeared related to higher levels of oxidative lipid damage. This could imply that the ADHD spectrum not only varies in terms of behavioral manifestations, but also regarding underlying mechanisms.
This case-control comparison has limitations. For example, , the nature of this study does not address the investigation of causality. It is therefore unknown whether increased oxidative stress or immune dysbalance are causative factors in the pathophysiology of ADHD, or a consequence of the disorder [23]. For instance, increased oxidative stress in ADHD could be the result of the restlessness related with ADHD [81]. Still, case-control studies are a good start, though not providing solid proof, to unravel underlying mechanisms of action. .
Furthermore, though all included patients in this study were diagnosed by a child neurologist as having ADHD according to DSM criteria, several patients would not be classified as such based solely on the SEQ scores of their parents. In addition, several controls had a positive SEQ ADHD score, without ADHD diagnosis or parental complaints. This underscores the subjectivity of questionnaires and might explain the lack of correlations between biomarkers and SEQ (sub)scores.
Another limitation is that the FFQ questioned consumption frequency but not portion size, which makes it impossible to draw conclusions on actual intake.
An important strength of this study is the use of regression analysis correcting for differences in processing time due to practical issues (e.g. sample transport to the analytical facility from the neighboring UZA was faster than from schools). Systematically recording processing time and correcting for this potential confounder was found crucial to obtain valid results and should thus be implemented in future research. In addition, the comparison of erythrocyte GSH and plasma CoQ10, retinyl palmitate, IL-5, IL-8 and IgM between ADHD patients and controls has not been published before.
Conclusion
Significantly higher plasma MDA levels were found in ADHD patients as compared to controls, and a trend for higher urinary 8-OHdG levels. Erythrocyte GSH and plasma retinyl palmitate, as well as plasma IgG and IgE levels were significantly higher in patients than in controls. Finally, a trend for lower plasma IL-5 levels was observed. Though only slightly elevated, more oxidative damage was thus found in ADHD. Antioxidant levels however did not differ or were even higher than in the control group. In addition, higher IgE levels were supported by a trend for more diagnosed allergies in patients than in controls. However, due to the nature of this study, it is unknown whether oxidative stress and immune dysbalance have a causative role in ADHD. Dietary habits do not appear to explain the observed biomarker differences. Further confirmation of these results is required, as well as further investigation of potential differences between ADHD subtypes. Finally, systematically correcting for processing time is crucial to obtain valid results.