The Association between Attention-Deficit/Hyperactivity Disorder and Atopic Dermatitis: A Study among Iranian Children.

BACKGROUND: The co-occurrence of attention-deficit/hyperactivity disorder (ADHD) with atopic dermatitis (AD) has been well described in some recent association studies; however, we did not have any perspective on this relationship in our country. AIM AND
OBJECTIVE: Hence, the present study aimed to assess the prevalence of ADHD in children with AD.
MATERIALS AND METHODS: This cross-sectional study was performed on 95 consecutive children and adolescents (aged 4-18 years) who were referred to dermatology clinics at the two hospitals in Tehran during 2017 with atopic dermatitis. The evidence of atopy was assessed using the 2003 National Survey of Children's Health. The diagnosis of ADHD was based on the Conner Rating Scale. The sleep disorder was also assessed by the Pittsburg sleep quality questionnaire.
RESULTS: The prevalence of hyperactivity and attention deficit in our AD patients was 20.0% and 29.47%, respectively. Furthermore, patients with sleep problem were significantly more likely to have hyperactivity disorder (odds ratio [OR]: 2.91, 95% confidence interval [CI]: 1.04-8.16, P = 0.04). According to the results of multiple logistic regression analyses, flexor involvement was the only predictor of hyperactivity disorder in the final model. The univariate and multivariate analyses showed that having attention deficit was associated with having cheek involvement (OR = 3.63, 95% CI: 1.44-9.14, P = 0.01) and sleep problem (OR = 3.68, 95% CI: 1.45-9.33, P = 0.01).
CONCLUSION: It seems that neurocognitive disturbances due to sleep restriction in AD children may be one of the main trigger, especially for attention deficit. Copyright:

Introduction

Atopic dermatitis (AD) is a chronic condition that begins in childhood with an overall prevalence ranging from 10% to 20% with more frequency in developing countries.[1] Hence, AD has been identified as the most frequent inflammatory skin disorder in children with much less prevalence in adults, estimated in 1%–2%.[23] Although majority of the patients suffered from a mild form of the disease, the combination of AD with chronic physical and even psychological conditions even in its mild form can considerably disable the patients.[4] In this regard, some AD-related comorbidities such as diabetes mellitus, allergic disorders, and renal impairment have been well described.[56] Besides, high rates of some psychological problems such as depression, anxiety, and stress disorders have also been reported in AD.[7] Interestingly, attention-deficit/hyperactivity disorder (ADHD) is the most common behavioral disorder in children affecting 5%–8% in the entire the world.[8] Because of serious consequences of ADHD for the patients and their families such as inattention, hyperactivity, and impulsivity, the patients face with impaired learning abilities at school, sleep disorder, social isolation, and thus with reduced quality of life for the patient and his/her family members.[9] Over the last century, several studies have focused the association between AD and ADHD. Unfortunately, we did not have any perspective on this relationship in our country. Therefore, the present study aimed to assess the prevalence of ADHD in children with AD and also to test the hypothesis that the presence of AD might be a potential determinant for ADHD.

Materials and Methods

This cross-sectional study was performed on 95 consecutive children and adolescents (aged 4–18 years) with atopic dermatitis who were referred to dermatology clinics at the two hospitals in Tehran, during 2017. Because of difficulty in definitive diagnosis of ADHD in younger children, we excluded those who were aged <4 years from the study. Other exclusion criteria were a history of psychiatric disorders with organic causes such as brain trauma or intracranial infections, mental retardation-related disability for understanding the test or participating in the interview. On admission to the clinics and by interviewing, the history of atopy in patients or among their first relatives was elicited. Then, all participants were physically and dermatologically examined. The evidences of eczema were assessed using the 2003 National Survey of Children's Health. In this regard, the prevalence of eczema was assessed by the question of “during the past 12 months, have you been told by a doctor or other health professional that the child had eczema or any kind of skin allergy?”[10] The diagnosis of ADHD was based on the Conner Rating Scale. This instrument was designed to assess ADHD and its most common comorbid problems in patients aged 6–18 years. This tool consisted of 27 questions divided into four subscales including: (1) oppositional problems, (2) cognitive problems, (3) hyperactivity and (4) an ADHD index that scored based on 4-point scale as 0 for “not at all,” 1 for “just a little,” 2 for “pretty much” and 3 for “very much.” The internal consistency of the questionnaire was assessed to be in the range of 0.55–0.89 with the Cronbach's alpha ranged 0.85–0.92.[11] The sleep disorder was also assessed by the Pittsburg sleep quality questionnaire (PSQI). This questionnaire contained 19 self-rated questions and five questions filled by the bed partner. The 19-self rated items were combined to form seven “component” scores, each had a range of 0 (for no difficulty) to 3 (for severe difficulty). The seven component scores were then added to yield one global score ranged from 0 (for no difficulty) to 21 (for severe difficulty). In our population, the corrected item-total correlations ranged from 0.30 to 0.75 for the seven component scores of the PSQI.[12]

Statistical methods

Statistical analysis was performed using the statistical software SPSS 16.0.0. (SPSS Inc. Chicago, IL, USA). Values of P < 0.05 were considered statistically significant.

Normality assumption was assessed using the Shapiro–Wilk test. Results were presented as median with total and interquartile ranges (IQRs) for quantitative variables and were summarized by absolute frequencies and percentages for qualitative variables. Quantitative variables were compared with t-test or Mann–Whitney U-test, whenever appropriate.

First, univariate logistic regression models were used to examine whether there were relationships between the role of AD and its components and hyperactivity status. Furthermore, these models were applied to investigate the relationships between the role of AD and its components and attention deficit. The odds ratio (OR) was calculated and presented with a 95% confidence interval (CI). Any variable having a P < 0.25 was selected as a candidate for the multiple logistic regression analysis. In the process of variable selection, independent variables were removed from the model if they were nonsignificant and not a confounder. The cut off point for significance was 0.10 alpha level and confounding as a change in any remaining parameter estimate >20% as compared to the full model.

After this process of deleting, refitting, and verifying, the model included significant independent variables and confounders. At this point, any variable not selected for the original multiple model was added back one at a time, with significant independent variables and confounders retained earlier. Any variable that was significant at the 0.1 level was kept in the model.

In addition, multiple linear regression analysis was applied to determine the parameters most predictive of the ADHD score. A step-wise forward regression algorithm was used to select variables entering in the final standard least square model. All variables that were significant in univariate analysis and biologically plausible to affect ADHD score were chosen for examining in this algorithm. The logarithmic transformation of ADHD score was used to improve the fit of the model.

Results

Of the 95 patients with definitive diagnosis of AD, 47 cases (49.47%) were male and 48 (50.53%) were female with median age of 9 years (range: 4–18 years). Family history of AD was revealed in 72 (75.79%) of AD patients. Regarding the different features of AD, itching and dry skin were the most reported findings (100% and 92.63%, respectively). In this study, flexor involvement was observed in 54 patients (56.84%), cheek involvement in 34 patients (35.79%), allergic rhinitis in 23 patients (24.21%), and asthma in seven patients (7.37%). Furthermore, 31 patients (32.63%) suffered from sleep problem. Based on different components of atopy, the median ADHD score was 31 (IQR: 29–48 and range: 26–78). According to the criteria for the diagnosis of ADHD, hyperactivity and attention deficit were found in 19 (20.00%) and 28 (29.47%) of AD patients, respectively.

Univariate logistic regression analyses indicated that AD patients with flexor involvement were 5.33 times more likely to develop hyperactivity disorder (95% CI: 1.44–19.82, P = 0.01) [Table 1]. Furthermore, patients with sleep problem were significantly more likely to have hyperactivity disorder (OR: 2.91, 95% CI: 1.04–8.16, P = 0.04) [Table 1]. However, no significant association was found between other features of atopy and hyperactivity status [Table 1].

Table 1

Univariate logistic regression predicting hyperactivity disorder

Criteria	Patients with hyperactivity (n=19), n (%)	Patients without hyperactivity (n=76), n (%)	OR (95% CI)	P
Male gender	13 (68.42)	34 (44.74)	2.68 (0.92-7.79)	0.07
Age (year)
Median	10	9	0.96 (0.85-1.09)	0.52
IQR	5-12	6.25-13.75
Range	4-18	4-18
Dry skin	17 (89.47)	71 (93.42)	0.60 (0.11-3.53)	0.56
Flexor involvement	16 (84.21)	38 (50.00)	5.33 (1.44-19.8)	0.01
Cheek involvement	9 (47.37)	25 (32.89)	1.84 (0.66-5.09)	0.24
Family history*	14 (73.68)	58 (76.32)	0.87 (0.28-2.74)	0.81
Sleep problem	10 (52.63)	21 (27.63)	2.91 (1.04-8.16)	0.04
Allergic rhinitis	6 (31.58)	17 (22.37)	1.60 (0.53-4.85)	0.40
Asthma	3 (15.79)	4 (5.26)	3.38 (0.69-16.58)	0.13

*Family history of atopic dermatitis. The values are expressed as n (%) unless otherwise noted. IQR: Interquartile range, OR: Odds ratio, CI: Confidence interval

According to the results of multiple logistic regression analyses, flexor involvement was the only predictor of hyperactivity disorder in the final model.

The univariate analyses showed that having attention deficit was associated with having cheek involvement (OR = 3.63, 95% CI: 1.44–9.14, P = 0.01) and sleep problem (OR = 3.68, 95% CI: 1.45–9.33, P = 0.01) [Table 2].

Table 2

Univariate logistic regression predicting attention deficit

Criteria	Patients with attention deficit (n=28), n (%)	Patients without attention deficit (n=67), n (%)	OR (95% CI)	P
Male gender	17 (60.71)	30 (44.78)	1.91 (0.78-4.68)	0.16
Age (year)
Median	10	9	1.03 (0.93-1.14)	0.60
IQR	8-12	6-13
Range	4-18	4-18
Dry skin	26 (92.86)	62 (92.54)	1.05 (0.19-5.75)	0.96
Flexor involvement	20 (71.43)	34 (50.75)	2.43 (0.94-6.27)	0.07
Cheek involvement	16 (57.14)	18 (26.87)	3.63 (1.44-9.14)	0.01
Family history*	23 (82.14)	49 (73.13)	1.69 (0.56-5.12)	0.35
Sleep problem	15 (53.57)	16 (23.88)	3.68 (1.45-9.33)	0.01
Allergic rhinitis	9 (32.14)	14 (20.90)	1.79 (0.67-4.82)	0.25
Asthma	4 (14.29)	3 (4.48)	3.56 (0.74-17.07)	0.11

*Family history of atopic dermatitis. The values are expressed as n (%) unless otherwise noted. IQR: Interquartile range, OR: Odds ratio, CI: Confidence interval

Cheek involvement and sleep problem remained significant predictors for developing attention deficit after adjusting for the gender and asthma status of the patients [Table 3].

Table 3

Multiple logistic regression predicting attention deficit (final model)

Criteria	Adjusted OR (95% CI)
Male gender	2.03 (0.75-5.51)
Cheek involvement	3.87 (1.43-10.50)
Sleep problem	3.51 (1.30-9.50)
Asthma	3.25 (0.64-16.43)

OR: Odds ratio, CI: Confidence interval

The median ADHD score was significantly higher in children with cheek involvement than those without this sign (37.5 [IQR: 30.75–54; range: 26–75] versus 31 [IQR: 29–35.5; range: 26–78], P = 0.04). Furthermore, the median ADHD score was significantly higher in patients with sleep disorder compared to those without sleep problem (43 [IQR: 30–58; range: 26–78] versus 31 [IQR: 29–34.75; range: 26–64], P = 0.002).

The median ADHD score was significantly higher in patients with flexor involvement in comparison with the patients without flexor involvement (33 [IQR: 30–55.25; range: 26–78] versus 30 [IQR: 28–34; range: 26–61], P = 0.002).

According to the results of step-wise forward regression, flexor involvement, sleep disorder, allergic rhinitis, and asthma status were significantly associated with log-transformed ADHD score (adjusted R2 = 0.19, F = 6.66, P < 0.0001) [Table 4].

Table 4

Summary of multiple linear regression analyses to identify factors associated with attention-deficit/hyperactivity disorder score

Predictors	Estimate	SE	t	P
Intercept	3.40	0.05	70.59	<0.0001
Flexor involvement	0.14	0.06	2.43	0.02
Sleep problem	0.18	0.06	2.98	0.004
Asthma	0.24	0.11	2.25	0.03
Allergic rhinitis	0.12	0.07	1.86	0.07

SE: Standard error

Discussion

Over the last century, several studies have focused the association between AD and ADHD. In this regard, AD has been identified as a potential risk factor for ADHD.[13] According to some recent reports by population-based studies, AD may increase the likelihood of ADHD up to 1.5 times;[14] however, some other studies could not demonstrate this association.[15] It seems that some community-based features may be responsible for the paradoxical findings in the association between AD and ADHD.

The pathogenesis of AD has been exclusively assessed that is rooted in two phenomena including immune disturbances and skin barrier defects which related to genetic background, geographical variability, and some environmental factors.[1617] Because of its chronic nature as well as debilitating comorbidities, AD can adversely affect infants' quality of life and burdensome socioeconomic impact.[1819]

There are now some evidences on the relationship between AD and its related components and ADHD.[131420] In line with the recent evidences, we hypothesized that different components of AD might predict the presence of both characteristics of ADHD including hyperactivity and attention deficit. In this regard, we could show that sleeping disorder and history of cheek involvement related to atopy could strongly predict attention deficit in the affected patients and flexor involvement could predict hyperactivity in the AD patients. Similar results about ADHD and AD and association with sleep problem have been found in previous studies.[2122] In Lee et al. survey, although the presence of AD significantly increased the risk of developing ADHD, contrary to our study, allergic rhinitis, allergic conjunctivitis, and asthma were associated with the greatest risk of developing ADHD.[23] Our study showed flexor involvement, sleep disorder, allergic rhinitis, and asthma could increase ADHD score.

Regarding the pathogenesis of ADHD in AD children, some recent studies attributed the components of ADHD directly to the pathways related to atopy and some others presented early antihistamine exposure as the main etiology for ADHD.[24] As a whole, it seems that sleeping disorder following the development of atopy may play a critical role in the appearance of ADHD. It has been recently shown that ADHD is frequently coincident with sleep disorders, especially with circadian-dysrhythmia condition.[25] In this regard, current guidelines strongly recommend assessment of sleep disturbance during the evaluation of ADHD, especially before the initiation of pharmacotherapy. In our study, the overall prevalence of sleep disorders in our children was 32.63% that was completely in the range globally reported between 25% and 55%.[26272829] Of the pathophysiological viewpoints, inadequate sleeping due to complications of atopy may lead to neurocognitive disturbances so that sleep restriction may lead to deficit attention and cognitive function in children.[3031] Certainly, these problems can flare-up in those who suffered underlying comorbidities such as atopy.

Conclusion

AD especially its-related sleep disturbance can be the main determinant for developing ADHD in children. It seems that neurocognitive disturbances due to sleep restriction and cheek involvement in AD children may be the main trigger for ADHD especially attention deficit. Flexor involvement in AD patients may be the principle trigger for hyperactivity disorder. According to paradoxical findings on the link between AD components and increased odds for ADHD, further studies on this association with the approach to ethical and genetics characteristics should be conducted.

Financial support and sponsorship

This study was supported financially by the Tehran University of Medical Sciences, Tehran, Iran.

Conflicts of interest

There are no conflicts of interest.