Plasma brain-derived neurotrophic factor (BDNF) in Egyptian children with attention deficit hyperactivity disorder

To date, researchers do not fully understand what the aetiology of attention deficit hyperactivity disorder (ADHD) is. As a neurotrophin, the brain-derived neurotrophic factor (BDNF) contributes to the growth of neurons as well as to the efficiency and plasticity of the neuronal synapse, which suggests that it may be a contributing aetiological factor in ADHD. Therefore, we sought to investigate the relationship between ADHD and plasma BDNF, including whether there is an association between BDNF and the degree of severity of ADHD and with ADHD subtypes. We found significantly elevated plasma BDNF in ADHD subjects in comparison with healthy subjects, but differences among the three subtypes of ADHD did not reach statistical significance. Compared with the control group, elevated plasma BDNF levels were found in the ADHD group, which suggests it may contribute to the aetiology of ADHD.

Changes in BDNF levels have previously been reported in ADHD [8,24,[27][28][29][30][31][32][33][34] but findings have been mixed, with some reporting elevated BDNF in children with ADHD in comparison with control subjects [8,30,34], while others found no significant differences [24,27,29,31,32] and others observing significantly lower levels of BDNF in ADHD [28,33]. Methylphenidate is known to increase dopamine and norepinephrine signaling in the brain [35,36], and both these neurotransmitters contribute to regulating BDNF expression in the hippocampus [37,38]. The few studies that have been conducted in Egypt investigating BDNF levels in ADHD were carried out in small samples [39][40][41], the largest of which included 35 subjects [40]. Therefore, we sought to carry out our study with a larger number of subjects involving 60 patients with ADHD to compare plasma BDNF levels to healthy controls. We also sought to investigate whether plasma BDNF correlated with symptom severity of the disorder.

Participants
Ninety participants were recruited from the Child Psychiatry outpatient department at Okasha Institute of Psychiatry, Ain Shams University, Cairo, Egypt. We divided the participants into two groups. The patient group comprised 60 drug-naïve children who have a newly received Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV) diagnosis of ADHD, between 6 and 12 years old, with an IQ of at least 90 on the Wechsler Intelligence Scale for Children (WISC), and who had no past medical history. The control group included 30 children matched for age and sex with the case group. These were healthy children recruited from children of workers at Ain Shams University Hospital. We excluded patients who had another neurodevelopmental disorder (such as Tourette's syndrome, a long-standing tic disorder, or a pervasive developmental disorder), had an IQ of less than 90, had schizophrenia or another psychotic disorder, had a longterm comorbid medical condition, or were receiving long-term medications.

Tools
We collected sociodemographic data for all participants and used the following tools: After informing them about the details of the study, we got informed written consent for taking part in the study first by asking the child's parent's permission and then, if it was appropriate for their age, by getting agreement from the participating child. After collecting demographic data, a full physical examination was carried out to exclude any medical conditions. A full psychiatric history and examination according to DSM IV were performed using the KSADS. General intelligence was assessed using the Arabic WISC to determine eligibility for participation. Information about SES collected from one of the parents using the Fahmy and El-Sherbini scale. ADHD severity was assessed using the CPRS-RL. For plasma BDNF, a trained nurse, under standard conditions, collected a 5-cc blood sample from participants. Samples were transferred to the laboratory on the same day for centrifugation. Plasma was stored at -70°C . A trained technician carried out the analysis. The package contained 96 kits, with 6 kits used for standardization procedures and the remaining 90 kits allocated to patients (60 kits) and controls (30 kits).

Laboratory procedures
Plasma BDNF Enzyme-linked immunosorbent assay (ELISA kit) [48] It contained purified human BDNF that coated the microtiter plate. Then after the formation of solid antibodies, BDNF was added to the wells. Labeled horseradish peroxidase enzyme was then combined with the BDNF antibody resulting in the formation of an antibody-antigen-enzyme complex. Addition of tetramethylbenzine then took place after complete washing. The substrate of the tetramethylbenzine then turned blue with catalyzation of the horseradish peroxidase. Adding a stop solution resulted in termination of the reaction, and a wavelength measuring 450 nm was used to determine the change in color. The optical density of plasma samples of study subjects was compared with a standard curve to then determine the level of BDNF for each subject.

Statistical analysis
Data were analyzed using the Statistical Package for Social Science (SPSS-20). A 0.80 power of the study was done, with calculations at 95% confidence intervals and 0.05 for α error. Descriptive statistics included mean and standard deviation (SD) for numerical data, and frequency and percentage of non-numerical data. Student t test and analysis of variance (ANOVA) were used to compare the differences between means; chi-square was used to investigating the differences between categorical variables; Fisher's exact test was used for differences between categorical variables that had a count that was < 5. Pearson's ("r") and Spearman's ("r s ") analysis were used to assess correlations. Statistical significance was considered when p values were less than 0.05.

Sociodemographic features
A total of 90 participants took part and were divided into a case group with 60 participants with ADHD and a control group of 30 matched healthy controls. Fortyfive subjects (75.0%) of the ADHD group were males, and 15 were females (25.0%), while 20 subjects (66.7%) of the control group were males, and 10 (33.3%) were females. The mean age of the ADHD group was 8.18 (SD ± 1.61) years, and the mean age of the control group was 8.37 (SD ± 1.79) years. The mean age of onset of symptoms in children with ADHD was 5.27 (SD ± 0.71) years, and the mean duration of illness was 3.08 (SD ± 1.62) years (Table 1).
Intelligence quotient (IQ) of the case and control groups As measured by the WISC, mean total IQ of children in the ADHD group was 97.07 (SD ± 6.01), mean verbal IQ was 99.13 (SD ± 6.37), and mean performance IQ was 95.68 (SD + 8.41). For the control group, the mean total intelligence quotient (IQ) of the children was 96.93 (SD + 4.67) as measured by the Wechsler Intelligence Scale for Children (WISC), the mean verbal IQ was 97.27 (SD ± 5.95), and the mean performance IQ was 97.50 (SD ± 7.70). No significant differences between cases and controls were found regarding total, verbal, or performance IQ (p > 0.05).

Diagnostic subtypes of the ADHD group and severity of symptoms assessed by Conners' scale
In terms of the distribution of diagnostic subtypes of the ADHD group, five (8.3%) subjects had inattentive type, 15 (25%) had hyperactive type, and 40 (66.7%) had the combined type.
As for the severity of ADHD symptoms, our assessment in the case group using Conners' Parent Rating Scale (long version) revealed significant problems in a number of its domains. Results are as follows: oppositional (mean 67.87, SD = 10.80), cognitive impairment (mean = 71.73, SD = 9.77), and hyperactivity (mean = 75.6, SD = 10.46) as shown in Table 2.

BDNF in healthy versus control subjects
Comparison between children with ADHD and healthy children regarding levels of plasma BDNF (ng/ml) revealed that children with ADHD exhibit higher levels of plasma BDNF (27.47 ± 18.37) than in healthy controls (12.30 ± 7.17). This difference of levels of plasma BDNF was statistically significant (Table 3).
On comparing the BDNF levels between each subtype and controls, significant differences were found between the hyperactive and combined subtypes of ADHD and the control group but not between the inattentive type and controls (Table 4).

Correlation BDNF with ADHD severity
The correlation between plasma BDNF level and severity of ADHD symptoms in the ADHD group as assessed by the Conners' Parent Rating Scale showed a significant negative correlation between plasma BDNF levels and emotional lability. There were no statistically significant correlations with the remaining domains of Conner's Parent Rating Scale (Table 5).

BDNF levels in ADHD versus control groups
Our study found that children with ADHD exhibited significantly higher levels of plasma BDNF than healthy controls. This finding is in line with Shim et al. [8] who found elevated levels of plasma BDNF in subjects with ADHD than in healthy subjects. Similar to our study, they measured plasma BDNF, with the groups matched by body mass index (BMI) and age. Similarly, the Egyptian study by Reda et al. [39] found that plasma BDNF in a group of children with ADHD was higher than in a control group of children, with the difference being statistically significant. Shim et al. [8] hypothesized that differences in BDNF levels might be due to BDNF upregulating in children with ADHD as a compensatory response for the abnormalities occurring in brain pathways involving serotonin and dopamine [8]. This is in line with Miyzaki et al. [49] who also observed significant increases of BDNF in conditions such as intellectual disability and autism spectrum disorders, in comparison with healthy participants. Meanwhile, contrary to our findings, another Egyptian study conducted by Sayyah [41] reported that plasma BDNF in children with ADHD children was not significantly different from controls. The difference in our findings might be due to the fewer number of participants in their study, as it included 21 ADHD subjects and 20 controls. In addition, a number of ADHD cases in their sample were already receiving treatment (unlike our study where ADHD were drug-naive), which may have influenced their outcomes [41].
On the other hand, Sargin et al. [32] and Scallestia et al. [24] found no statistically significant difference in serum BDNF levels between ADHD and control groups. The difference with our study may be due to our study measuring plasma BDNF, whereas these studies measured serum BDNF, which is about 20-fold more concentrated. It has been debated whether plasma or serum measurements are more reliable in reflecting BDNF levels in the brain, with some suggesting that it is plasma because of the minimal effect of stored platelet BDNF [8], while others support the validity of serum samples as they may reflect the role of either the illness or its treatment, during which there is an accumulation of platelet BDNF [2].
Different sampling methods and different laboratories may contribute to different results. Karege et al. [26] reported that there were variations in the level BDNF that was taken from plasma, serum or whole blood samples, according to the laboratory where the analysis was carried out. Therefore, the type of tubes in which samples are collected or the type of ELISA kit used may result to BDNF readings being different between studies.
Contrary to our results, the study in Iran by Saadat et al. [33] found significantly lower plasma BDNF levels ADHD patients compared with controls. The difference between our results and Saadat et al. may be due to the lower age range of their study which was 3.5-10 years. Different BDNF levels at different ages have been reported, with BDNF levels decreasing as age increases [49].

BDNF and ADHD subtypes
Regarding BDNF levels among different subtypes in the ADHD group, our study showed that children with attention deficit-predominant type had a lower plasma level of BDNF than children with combined type and hyperactivity-predominant type ADHD, but none of these were statistically significant differences. Similar to our finding, Sargin et al. [32] reported no significant link between the different subtypes of ADHD and serum BDNF. Saadat et al. [33] also found no significant differences in plasma BDNF levels between ADHD subtypes. On the other hand, Sayyah [41] found significantly lower plasma BDNF levels in the inattentive compared with the hyperactive type of ADHD. However, they found no significant difference between the hyperactive and combined subtypes, nor between the inattentive and combined types.
In the current study, on comparing plasma BDNF in different diagnostic categories of ADHD and controls, differences between each of the hyperactive and combined subtypes of ADHD when compared with the  control group were found which were significant, whereas there was no significant difference between the inattentive group of children and the control group. Contrary to this finding, Sayyah [41] found a significantly reduced plasma BDNF in the inattentive type of ADHD compared with the controls. Similarly, Saadat et al. [33] found that the level of plasma BDNF in the various categories of ADHD (hyperactive/impulsive, inattentive and combined) were of a significant difference in comparison to controls. The difference between our study and other studies regarding different BDNF levels between different subtypes of ADHD may be due to our sample having a small number of subjects with the inattentive type that represented only 8.3%.

BDNF levels and gender
Our study found no significant differences in the levels of BDNF between males and females in either group. Findings from other studies that assessed the possible difference of BDNF level between sexes also found no differences [8,41].

Correlation of BDNF and ADHD symptoms
Regarding BDNF levels and severity of ADHD symptoms, a significant negative correlation between emotional lability and BDNF was found. This might point to the fact that BDNF is highly expressed in the limbic system and the cerebral cortex, making it a mandatory factor in the regulation of affective and reward process [50]. On the other hand, research has shown that downregulation of BDNF is associated with increase in anxiety-like symptoms [51]. The exact explanation of the negative correlation between BDNF and emotional lability is not clearly understood owing to the difficulty in defining the term of emotional lability and its heterogenicity, and its amplification of an underlying complex psychopathology [52]. Otherwise, no significant correlation between BDNF levels and the scores of the remaining domains of CPRS were found. Similarly, Sargin et al. [32] found that serum BDNF and the severity of impairment due to ADHD did not significantly correlate. Similarly, Saadat et al. [33] found no significant association when correlating the level of BDNF with ADHD severity, also using the CPRS. In contrast to this finding, Shim et al. [8] observed a significant positive correlation between the severity of inattention symptoms in ADHD and the level of BDNF in the plasma. However, the process of how BDNF can affect ADHD severity is still not understood. Reda et al. [39] reported that BDNF correlated negatively with hyperactivity/impulsivity and positively with anxiety/shyness. The difference between the results of the current study and the previously discussed study is probably due to the use of different scales for assessing the severity of symptoms of ADHD. Shim and colleagues used the K-ARS (Korean

Limitations
Our entire case sample was selected from the Child Psychiatry Clinic, which represents cases with more severe symptoms; hence, it is difficult to generalize our results to the general population. Our study also did not investigate the effect of diet, exposure to smoking, physical activity, or any other environmental factors that might affect the BDNF levels.
In addition, our sample included a small proportion of subjects with the predominantly inattentive subtype of ADHD. Also, we did not correlate IQ data for attention and concentration with BDNF levels.

Conclusions
Children with ADHD exhibit higher levels of plasma BDNF levels in comparison with age-and sex-matched healthy controls. There is no obvious correlation between the severity of symptoms in children with ADHD and their plasma BDNF level. Our study showed that children with attention deficit-predominant type had a lower plasma level of BDNF than children with combined type and hyperactivity-predominant type ADHD yet no significant difference between subtypes of ADHD regarding BDNF level was detected. This raises the question of the possibility that drugs acting on enhancement of BDNF might play an important role in the improvement of patients with ADHD not only on symptomatic level but on the neurodevelopmental level. However, further studies are needed with larger sample sizes. However, we must take into consideration that our sample was not representative of the inattentive diagnostic subtype.

Recommendations
As BDNF plays a curial rule in neurogenesis and synaptic transmutation, further studies with larger sample sizes are needed to explore different domains concerning ADHD and BDNF as cognitive functions and changes to BDNF in relation to medication, as well as other psychiatric disorders as ASD and mood disorders in children. Future research should also follow up drug-naïve ADHD after different drug treatments.