Abstract
Background
Pediatric atopic dermatitis (AD) has been linked to neuropsychiatric comorbidities such as depression, anxiety, and attention-deficit/hyperactivity disorder (ADHD). However, longitudinal data are limited and the effect of AD severity on neuropsychiatric outcomes requires further characterization.
Objectives
To determine the risk of several major neuropsychiatric conditions in children with AD
Methods
We analyzed UK health records data in a population-based cohort study. Each patient <18 years old with AD was matched to up to 5 unaffected patients on age, practice, and index date. Treatments served as proxies for AD severity, which was analyzed in a time-updated manner. Outcomes were incident anxiety, depression, bipolar disorder, schizophrenia, ADHD, autism, obsessive-compulsive disorder (OCD), suicidal ideation or attempt, and completed suicide.
Results
409,431 children with AD (93.2% mild, 5.5% moderate, 1.3% severe) were compared to 1,809,029 children without AD. In Cox regression models adjusted for age, sex, socioeconomic status and other atopic comorbidities, no statistically significant relationships were observed between AD and incident anxiety (HR 1.01, 95% CI 0.99–1.03), ADHD (1.02, 0.97–1.06), autism (1.02, 0.98–1.06), bipolar disorder (1.08, 0.85–1.36), suicidal ideation/attempt (0.98, 0.95–1.01), or completed suicide (0.85, 0.64–1.14). Children with AD were less likely to develop depression (0.93, 0.91–0.95) or schizophrenia (0.72, 0.54–0.95) but more likely to develop OCD (1.26, 1.16–1.37). However, there was substantial variation by AD severity and age in both the direction and magnitude of effect for many of the neuropsychiatric conditions examined.
Conclusions
The was no substantial impact of AD on the overall risk of many neuropsychiatric conditions in children, but disease severity and age may be important modifying factors. Additional research is needed to further dissect the complex relationship between pediatric AD and neuropsychiatric comorbidities.
Keywords: atopic dermatitis, eczema, mental health, neurodevelopment, psychiatric disorders, pediatrics
Introduction
Atopic dermatitis (AD) is a common, chronic pediatric disease with detrimental effects on quality-of-life.[1–3] Growing evidence suggests that AD is associated with depression, anxiety, and suicidality in both children and adults.[4–6] Associations of AD with attention deficit hyperactivity disorder (ADHD) and autism have also been reported.[7, 8] It is postulated that the development of neuropsychiatric comorbidities may be driven by chronic itch, stigma, and sleep disturbances related to AD. Another hypothesis is that systemic inflammatory mediators in AD accelerate neuroinflammation with subsequent development of neuropsychiatric disorders.[9–11] Given these potential connections, a comprehensive assessment of the neuropsychiatric burden of AD among its most commonly affected population—children—is needed. To date, longitudinal data on neuropsychiatric outcomes in children with AD are relatively inconsistent or lacking, particularly for diagnoses such as schizophrenia, bipolar disorder, and obsessive-compulsive disorder (OCD). Moreover, the effect of AD severity on such outcomes has not been fully characterized. To address these gaps, we conducted a study to evaluate the risk of several major neuropsychiatric conditions among children with AD in a population-based cohort.
Materials and Methods
We conducted a retrospective cohort study using The Health Improvement Network (THIN), an electronic health records database of >600 general practices in the UK representative of its population and widely used for biomedical research (https://www.the-health-improvement-network.com/). Since general practitioners (GPs) are the primary contact for medical care in the UK, a majority of AD patients are managed exclusively by GPs. Diagnostic codes for many conditions including AD have been validated in THIN.[12–14] Data collected between 1994 and February 2015 were included.
All patients with AD (exposed) were identified and matched to up to 5 patients without AD (unexposed) on age (+/−3 years), practice, and encounter within +/−6 months of the index date for the AD patient (defined as latter of practice registration and diagnosis dates). Patients with AD aged <18 years at index date and their matched controls were included. AD was defined using a validated algorithm requiring ≥1 of 5 common diagnosis codes for AD and 2 AD-related therapy codes, which has a positive predictive value of 86% for a physician-confirmed diagnosis.[12] Each control was assigned a ‘diagnosis’ date based on an encounter within +/−6 months of the matched AD patient’s index date, to ensure that both groups were followed during similar time periods. Follow-up for exposed individuals began at the latest of AD diagnosis, practice registration, or Vision date (i.e. when Vision software was implemented thereby assuring data quality). For unexposed individuals, follow-up began at the latest of ‘diagnosis,’ registration, or Vision dates. Follow-up ended at the earliest of outcome occurrence, transfer out of the practice, death, or end of study period. Patients with any history of the outcome of interest at time of cohort entry were excluded from those corresponding analyses.
AD severity was estimated using treatment proxies and defined in a time-updated manner as follows: All patients with AD were considered to have mild disease by default. They were classified as having moderate AD at the first of receiving: i) a second potent topical corticosteroid within one year, or ii) a first topical calcineurin inhibitor (which is reserved in the UK for moderate AD).[15] Patients were classified as having severe AD at the first of: i) systemic immunosuppressant treatment, ii) phototherapy use, or iii) referral to dermatology (as 96% of AD patients are managed exclusively by GPs).[14] Once defined as having moderate AD, patients remained as such unless they developed severe AD; once defined as having severe AD, patients remained as such until end of study. This approach to defining AD severity has been used previously.[16]
Outcomes included ADHD, autism, anxiety, depression, bipolar disorder, schizophrenia, OCD, suicidality and completed suicide, defined by Read diagnosis codes and in the case of suicide, also death codes. We collected data on age, sex, socioeconomic status (i.e. Townsend index, a measure of material deprivation), asthma, and allergic rhinitis. Covariates were defined at cohort entry; AD severity was time-updated. Using Cox regression models, we compared time to incident outcomes between patients with AD and their matched controls, adjusted for potential confounders determined a priori including age, sex, socioeconomic status, asthma and allergic rhinitis. In secondary analyses, models were further stratified by age at cohort entry: ≤ 5, 6–11, and 12–17 years old. Sensitivity analyses were conducted to address potential biases introduced by concomitant atopic disorders, short follow-up time, ascertainment bias, and symptom-based codes for anxiety and depression. Additional analyses using maximum, rather than time-updated, disease severity were conducted to explore changes in direction of effect between mild versus moderate/severe AD. The University of Pennsylvania Institutional Review Board approved the study.
Results
409,431 children with AD (93.2% mild, 5.5% moderate, 1.3% severe) were matched to 1,809,029 patients without AD. The median age was 4 (IQR 2–9) years in the non-AD group and ranged from 4 (IQR 1–8) to 9 (4–14) years in the AD groups. Socioeconomic status was not meaningfully different across groups (Table 1). Non-AD controls were followed for a median of 5.0 (IQR 2.0–9.4) years; those with AD were followed for 5–7 years on average.
Table 1:
Baseline characteristics of children with and without AD in THIN database, 1994 to 2015
| Characteristic, N (%) | N=1,809,029 | N=381,678 | N=22,433 | N=5,320 |
|---|---|---|---|---|
| Age, median (IQR), y | 4 (2, 9) | 4 (1, 8) | 9 (4, 14) | 5 (1, 10) |
| Sex | ||||
| Female | 872279 (48.22) | 184682 (48.39) | 11054 (49.28) | 2335 (43.89) |
| Male | 936750 (51.78) | 196996 (51.61) | 11379 (50.72) | 2985 (56.11) |
| Townsend deprivation index | ||||
| 1-Lowest | 424409 (24.71) | 89820 (24.89) | 4768 (22.55) | 1251 (25.00) |
| 2-Low | 340677 (19.84) | 71979 (19.95) | 4106 (19.42) | 1069 (21.37) |
| 3-Moderate | 355559 (20.70) | 75261 (20.86) | 4551 (21.52) | 1033 (20.65) |
| 4-High | 339336 (19.76) | 70649 (19.58) | 4316 (20.41) | 900 (17.99) |
| 5-Highest | 257540 (14.99) | 53113 (14.72) | 3407 (16.11) | 750 (14.99) |
| Unknown | 91508 (5.06) | 20856 (5.46) | 1285 (5.73) | 317 (5.96) |
| Person-time, median (IQR), y | 4.99 (2.0, 9.4) | 5.22 (2.1, 9.7) | 6.02 (2.6, 10.2) | 6.89 (2.7, 12.6) |
| Allergic rhinitis | 75050 (4.15) | 23935 (6.27) | 2870 (12.79) | 521 (9.79) |
| Asthma | 169679 (9.38) | 49782 (13.04) | 6094 (27.17) | 1222 (22.97) |
| History of neuropsychiatric conditions | ||||
| Treatments for AD |
Abbreviations: AD, atopic dermatitis; IQR, interquartile range
ADHD and autism
The crude incidence rates were 0.9–1.2 per 1,000 person-years (PY) for ADHD and 1.0–1.2/1000PY for autism (Table 2).
Table 2.
Incidence rates of neuropsychiatric outcomes (per 1000 person-years) in children with and without AD in THIN database, 1994 to 2015
| Outcome | No AD N=1,809,029 |
Mild AD N=381,678 |
Moderate AD N=22,433 |
Severe AD N=5,320 |
||||
|---|---|---|---|---|---|---|---|---|
| No. of cases | Incidence rate | No. of cases | Incidence rate | No. of cases | Incidence rate | No. of cases | Incidence rate | |
| ADHD | 12272 | 1.13 (1.11–1.15) | 2704 | 1.22 (1.17–1.27) | 294 | 1.07(0.95–1.20) | 60 | 0.86 (0.67–1.11) |
| Anxiety | 52288 | 4.86 (4.82–4.90) | 10947 | 5.00 (4.91–5.09) | 2326 | 8.72 (8.37–9.08) | 469 | 6.93 (6.33–7.59) |
| Autism | 11610 | 1.06 (1.04–1.08) | 2559 | 1.15 (1.11–1.196) | 284 | 1.03 (0.91–1.15) | 73 | 1.05 (0.83–1.31) |
| Bipolar disorder | 329 | 0.03 (0.03–0.03) | 73 | 0.03 (0.03–0.04) | 21 | 0.08 (0.05–0.12) | 2 | 0.03 (0.01–0.11) |
| Depression | 43540 | 4.02 (3.99–4.06) | 8534 | 3.87 (3.79–3.95) | 2187 | 8.18 (7.84 −8.53) | 427 | 6.28 (5.71–6.90) |
| OCD | 2539 | 0.23 (0.22–0.24) | 659 | 0.29 (0.27–0.32) | 136 | 0.49 (0.41–0.58) | 17 | 0.24 (0.15–0.39) |
| Schizophrenia | 352 | 0.03 (0.03–0.04) | 43 | 0.02 (0.01–0.03) | 13 | 0.05 (0.03–0.08) | 4 | 0.06 (0.02–0.15) |
| Suicidal ideation or attempt | 21243 | 1.95 (1.92–1.97) | 4320 | 1.95 (1.89–2.01) | 807 | 2.94 (2.75– 3.15) | 163 | 2.35 (2.02–2.74) |
| Completed suicide | ||||||||
Abbreviations: AD, atopic dermatitis; ADHD, attention deficit hyperactivity disorder; OCD, obsessive compulsive disorder
n/e indicates not estimable due to insufficient number of events
The adjusted risk of new-onset ADHD was not significantly different between AD and non-AD groups (HR 1.02, 95% CI 0.97–1.06) (Figure). Stratified by disease severity, the risk of ADHD was not significantly different between mild or moderate AD versus non-AD, while severe AD was associated with significantly lower risk of ADHD (0.76, 0.58–0.98). Estimates remained similar when using maximum, rather than time-updated, disease severity (Table 3). Stratified by age, the risk of incident ADHD was elevated only among children ≤5 years old (1.09, 1.04–1.15) and 6–11 years old (1.09, 1.001–1.18) (Table 4).
Figure.
Adjusted risk of neuropsychiatric outcomes comparing children with and without AD
Table 3.
Sensitivity analysis using maximum AD disease severity
| Outcome | Adjusted hazard ratio (HR) [ref: no AD] | |||
|---|---|---|---|---|
| Overall AD | Mild AD | Moderate AD | Severe AD | |
| ADHD | 1.01 (0.97–1.05) | 1.03 (0.99–1.08) | 0.92 (0.82–1.02) | 0.68 (0.53–0.86) |
| Anxiety | 1.08 (1.06–1.10) | 1.08 (1.06–1.11) | 1.09 (1.05–1.14) | 1.03 (0.95–1.12) |
| Depression | 1.07 (1.05–1.09) | 1.05 (1.03–1.08) | 1.12 (1.07–1.16) | 1.07 (0.98–1.17) |
Abbreviations: AD, atopic dermatitis; ADHD, attention deficit hyperactivity disorder
Adjusted for age, sex, Townsend score, and history of allergic rhinitis and asthma. Disease severity reflects maximum severity during follow-up.
Table 4.
Adjusted risk of neuropsychiatric outcomes comparing children with and without AD, stratified by age at cohort entry
| Outcome | Adjusted hazard ratio (HR) [ref: no AD] | |||
|---|---|---|---|---|
| All ages | ≤ 5 years old | 6–11 years old | 12–17 years old | |
| ADHD | 1.02 (0.97–1.06) | 1.08 (1.03–1.14) | 1.09 (1.01–1.18) | 1.03 (0.87–1.23) |
| Anxiety | 1.01 (0.99–1.03) | 0.93 (0.90–0.97) | 1.02 (0.98–1.05) | 1.04 (1.01–1.08) |
| Autism | 1.02 (0.98–1.06) | 1.05 (1.002–1.11) | 1.11 (1.01–1.22) | 1.06 (0.88–1.28) |
| Bipolar disorder | 1.08 (0.85–1.36) | 0.82 (0.40–1.70) | 0.82 (0.54–1.26) | 1.38 (1.01–1.87) |
| Depression | 0.93 (0.91–0.95) | 0.65 (0.61–0.69) | 0.90 (0.87–0.93) | 1.06 (1.03–1.09) |
| OCD | 1.26 (1.16–1.37) | 1.18 (1.03–1.37) | 1.27 (1.11–1.45) | 1.31 (1.14–1.51) |
| Schizophrenia | 0.72 (0.54–0.95) | 0.44 (0.16–1.24) | 0.64 (0.38–1.07) | 0.86 (0.61–1.22) |
| Suicidal ideation or attempt | 0.98 (0.95–1.01) | 0.88 (0.83–0.93) | 0.94 (0.90–0.99) | 1.09 (1.03–1.14) |
| Completed suicide | ||||
Abbreviations: AD, atopic dermatitis; ADHD, attention deficit hyperactivity disorder; OCD, obsessive compulsive disorder
Adjusted for age, sex, Townsend score, and history of allergic rhinitis and asthma.
n/e indicates not estimable due to insufficient number of events
There was no overall association between AD and autism, but moderate AD was associated with increased risk of incident autism (1.25, 1.11–1.41) (Figure). Stratified by age, AD was associated with autism among children ≤5 years old (1.05, 1.002–1.11) and 6–11 years old (1.11, 1.01–1.22) (Table 4).
Anxiety and depression
Of all the neuropsychiatric outcomes, anxiety and depression occurred most commonly in both AD and non-AD groups with incidence rates of 4–9/1000PY (Table 2).
The adjusted risk of anxiety was slightly but statistically significantly increased among children with mild AD (HR 1.08, 1.06–1.10) and reduced among children with moderate or severe AD (0.81, 0.77–0.84 and 0.77, 0.70–0.84, respectively) (Figure). Stratified by age, the positive association between AD and anxiety was limited to children 12–17 years old at baseline, while an inverse relationship between AD and anxiety was observed among children ≤5 years old (0.93, 0.90–0.97) (Table 4).
Similar trends were observed for depression, with mild AD portending an increased risk (1.05, 1.03–1.09) while moderate and severe AD showed decreased risk (0.68, 0.65–0.71 and 0.64, 0.58–0.70, respectively) (Figure). Age-stratified analyses revealed a positive association between AD and depression among adolescents (1.06, 1.03–1.09) but an inverse association among younger children (≤5 years old: 0.65, 0.61–0.69; 6–11 years old: 0.90, 0.87–0.93) (Table 4).
Using maximum, instead of time-updated, disease severity, the inverse associations of moderate/severe AD with anxiety and depression disappeared; rather, moderate AD became associated with increased risk of anxiety (1.09, 1.05–1.14) and depression (1.12, 1.07–1.16) and severe AD was not associated with either (Table 3).
Suicidality
Suicidal ideation or attempt occurred at a rate of 2–3/1000PY across all groups (Table 2). The risk of suicidal ideation/attempt was slightly greater among children with mild AD (1.04, 1.01–1.08) but lower among those with moderate or severe AD (0.76, 0.70–0.81 and 0.74, 0.63–0.86, respectively) (Figure). In age-stratified analyses, younger AD-affected children were less likely to have suicidal ideation/attempt compared to their non-AD peers (≤5 years old: 0.88, 0.83–0.93; 6–11 years old: 0.94, 0.90–0.99), while 12- to 17-year-olds were more likely to have suicidal ideation/attempt (1.09, 1.03–1.14) (Table 4). Completed suicide was an extremely rare event, with incidence rates of 0.01–0.04/1000PY across the groups (Table 2). Risk of completed suicide did not differ by AD status (Figure).
Bipolar disorder, OCD, and schizophrenia
Bipolar disorder, OCD, and schizophrenia occurred with crude incidence rates of 0.02–0.49/1000PY (Table 2).
There was no association between AD and bipolar disorder overall (Figure). Age-stratified models showed increased risk of bipolar disorder (1.38, 1.01–1.87) among 12–17-year-old children with AD versus those without AD (Table 4).
The adjusted risk of OCD was significantly elevated among children with AD versus unaffected children (1.26, 1.16–1.37), primarily those with mild (1.30, 1.20–1.42) or moderate (1.20, 1.01–1.43) AD (Figure). AD was associated with OCD across all ages, ranging from 18% greater risk in the youngest to 31% greater in the oldest groups (Table 4).
Schizophrenia risk was significantly lower among children with moderate AD (0.50, 0.28–0.89) but not those with mild or severe AD, relative to their unaffected peers (Figure). No variations by age were observed (Table 4).
Sensitivity analyses
Analyses restricted to patients with at least yearly GP follow-up, ≥5 years of follow-up, or ≥1 year of observation prior to cohort entry were similar to the primary findings (Table S1). Exclusion of patients with asthma or allergic rhinitis also led to similar results.
Discussion
In summary, we observed mostly null associations between overall AD and multiple neuropsychiatric conditions in a pediatric cohort. There was also substantial variation by disease severity and age in both the direction and magnitude of associations between AD and ADHD, autism, anxiety, depression, suicidality, OCD and bipolar disorder.
Previous cross-sectional studies have reported greater rates of depression and anxiety in children with AD but few pediatric cohort studies have been conducted.[4, 5, 17] Two Taiwanese cohort studies found greater risk of depressive and anxiety disorders among adolescents with AD.[18, 19] Birth cohort studies in the UK and Netherlands have similarly associated AD with internalizing, including depressive and anxious, symptoms.[20, 21] In contrast, we observed no overall association between AD and anxiety and a small inverse association between AD and depression. However, some subgroups—children with mild AD and adolescents—were slightly more likely to develop anxiety or depression compared to peers without AD.
Most strikingly, our primary analyses suggest a potentially inverse relationship between moderate/severe AD and anxiety and depression. Although few studies have examined AD severity effects on these outcomes, a greater risk of depressive and internalizing symptoms was associated with more severe AD in one recent study using caregiver-reported questionnaires.[20] One possible explanation for the differences in our findings may be due to differences in assessment, as we relied on medical records to capture diagnoses and estimate AD severity. Another possibility is that treatments for severe AD could mitigate neuropsychiatric symptoms.[22] As Th1/Th2 pathophysiology may play a role in neuropsychiatric disorders, it is also possible that Th2 skewing in AD bears a protective effect against conditions such as major depression.[23] Reverse causation is another potential explanation for the inverse associations observed. However, the baseline prevalence of these neuropsychiatric conditions was not consistently higher among children with more severe AD. We examined only incident outcomes, which were recorded prospectively by GPs, and sensitivity analyses restricted to patients with ≥1 year follow-up before baseline showed similar results. We also posited that inverse associations could be attributed to a “depletion of susceptibles” effect whereby anxiety or depression may arise soon after AD onset, when skin disease is generally mild, in turn leading to seemingly lower incidence (i.e., fewer newly diagnosed cases) with longer-standing AD, when AD may become more severe. Our sensitivity analyses using maximum AD severity showed increased risk of anxiety and depression with moderate AD and no significant association with severe AD, indicating that earlier onset of these outcomes during periods of mild AD could account for findings in the primary analyses. This suggests that the inverse relationships observed between moderate/severe AD and anxiety and depression could be an artifact of the analytic approach. However, it remains notable that severe AD was not associated with anxiety or depression in the sensitivity analysis using maximum AD severity.
Our study also found contradictory evidence on suicidality. AD has previously been linked to greater odds of suicidal ideation/attempt.[5, 17, 24, 25] However, data on completed suicide are sparser and inconsistent.[26–28] While one Canadian case-control study found a 22% greater risk of completed suicide among adolescents and adults with persistent AD, a population-based Danish study found no increased risk among adults with AD despite greater rates of depression and anxiety.[26, 28] Few studies to date have evaluated preadolescent populations or the impact of AD severity.[26, 29–31] In our study, the risk of suicidal ideation/attempt was minimally elevated among children with mild AD and inversely associated with moderate/severe AD, similar to our findings for anxiety and depression. As the analytic approach and potential misclassification of suicidality in young children could have contributed to the inverse associations observed between moderate/severe AD and suicidality, further investigation is warranted to better characterize those relationships.
ADHD has received much attention in AD research, with cross-sectional studies finding greater odds of ADHD in patients with AD compared to those without.[4, 32–37] Some case-control studies have similarly identified greater odds of atopic illness among children with ADHD compared to non-ADHD controls.[38, 39] In a few pediatric cohort studies, children with AD had a 16–300% increased risk of ADHD compared to the general population.[7, 8, 40, 41] Yet, other studies found no association.[42, 43] In one birth cohort, children with early-onset AD were at greater risk for early-onset ADHD but not later-onset ADHD,[44] while another study observed no association between childhood AD and ADHD medication prescription.[45] Our study found no overall association between AD and ADHD, and severe AD was inversely associated with ADHD even in sensitivity analyses. However, age also modified the AD-ADHD association. It is possible that some of the age-related effects could be attributed to differences in baseline incidence by age, i.e., ADHD more commonly diagnosed at younger ages. External factors, such as sleep or antihistamine use, may additionally mediate the AD-ADHD relationship.[46, 47]
AD and autism have been previously linked by mostly cross-sectional studies.[4, 48] While two Taiwanese cohort studies also identified greater risk of autism in children with AD compared to those without AD, the estimates ranged widely from 10% to 900% higher.[7, 8] In another case-control study of children with autism versus without, there was no association with pre-existing AD.[49] While we did not observe an overall association between AD and autism in our cohort, subgroups of younger children and children with moderate AD had slightly greater risk of autism compared to non-AD controls, suggesting a potentially complex relationship between AD and autism that may be influenced by both age and disease severity.
Data on schizophrenia, bipolar disorder, and OCD risk in children with AD are limited. In one Danish cohort study of adolescents, having any atopic disorder increased the risk of subsequent schizophrenia diagnosis by 45% but this was attributed primarily to asthma.[50] In another UK cohort, AD at age 10 was associated with 30–40% greater odds of psychotic experiences at age 13.[51] Our study, in contrast, found a mostly null to perhaps inverse effect of AD on schizophrenia. With respect to bipolar disorder, one cohort study of Taiwanese adolescents found a greater risk of bipolar disorder in children with any atopy.[18] Our study of AD found no overall association between AD and bipolar disorder, except among adolescents. Finally, a small cross-sectional study reported higher rates of AD in children with OCD compared to those without OCD.[52] Our present longitudinal study confirms an elevated risk of OCD in children with AD across all age groups. Although the mechanisms driving this association require further investigation, one possibility is that chronic AD or itch elicits repetitive or compulsive behaviors, such as scratching or skin picking, which may then be misdiagnosed as OCD. It is also possible that compulsive handwashing in the setting of OCD may lead to the development or exacerbation of AD.
We note several potential limitations. First, misclassification of AD severity is possible as treatments and dermatology referrals served as proxies. Albeit a common approach in epidemiologic studies when direct severity measures are unavailable,[16, 53] it is difficult to disentangle the influence of treatments versus disease severity on outcomes. For example, it is possible that inverse associations between severe AD and neuropsychiatric conditions are due to treatments, i.e., systemic therapies more effectively reduce systemic inflammation thereby decreasing risk of mood disorders or ADHD. Indeed, recent pharmacologic trials have demonstrated that treatment of AD can improve symptoms of depression and anxiety.[22] The time-updated nature of AD severity may also lead to misclassification among children whose disease subsequently improves at older ages; we would expect such misclassification to bias toward the null, and this may account for some of the null or inverse associations between moderate/severe AD and ADHD, anxiety, depression, and suicidality. Practice variations across GPs may also influence treatment prescriptions and specialty referrals; AD and non-AD patients were thus matched on practice to try to control for this. Second, outcome misclassification is possible, but we would expect it to be nondifferential with respect to the exposure, as GPs were unaware of our hypotheses. Nevertheless, some outcomes such as depression or suicidality are challenging to diagnose in young children or in primary care settings; caution is thus warranted when interpreting those findings. Some neuropsychiatric outcomes, e.g. schizophrenia, are also uncommon in children, so even seemingly large differences in relative risk may translate to only minor differences in absolute risk. Although we also stratified the analyses by age into early childhood, mid-childhood and adolescence, it is possible that more granular age stratification would lead to different estimates. Third, ascertainment bias is possible if patients with AD are more likely to see their GP and thus receive neuropsychiatric diagnoses; however, analyses restricted to patients seen at least annually yielded similar results. Finally, reverse causation remains possible as discussed above.
In conclusion, our findings suggest a complex landscape of neuropsychiatric burdens among children with AD. They also call for additional prospective research—studies which measure AD severity and treatment independently and assess neuropsychiatric outcomes via both symptom report and physician diagnosis—to further dissect the nature of relationships, and modifying factors thereof, between pediatric AD and neuropsychiatric comorbidities.
Supplementary Material
Funding sources:
This study was supported by a contract from Pfizer, Inc. paid to the Trustees of the University of Pennsylvania (JMG). The academic authors designed and executed the study with input from the sponsor. The academic authors had final approval over the manuscript.
Conflicts of interest:
J.W. has received research and fellowship funding from Pfizer, Inc. (paid to the University of Pennsylvania and Johns Hopkins University) and research funding from the National Eczema Association (paid to Johns Hopkins University). M.S. has received fellowship funding from Pfizer, Inc. (paid to the University of Pennsylvania). K.A. has received research funding from Pfizer, Inc. and L’Oréal (paid to UCSF), and receives consulting fees for serving on the academic steering committee for TARGET RWE. A.R.L. is an employee of Pfizer, Inc. J.M.G. has served as a consultant for Abcentra, Abbvie, BMS, Boehringer Ingelheim, GSK, Lilly (DMC), Janssen Biologics, Novartis Corp, UCB (DSMB), Neuroderm (DSMB), Trevi, and Mindera Dx., receiving honoraria; and receives research grants (to the Trustees of the University of Pennsylvania) from Boehringer Ingelheim, and Pfizer Inc.; and received payment for continuing medical education work related to psoriasis that was supported indirectly by pharmaceutical sponsors; is a co-patent holder of resiquimod for treatment of cutaneous T cell lymphoma; serves as a Deputy Editor for the Journal of Investigative Dermatology receiving honoraria from the Society for Investigative Dermatology; is Chief Medical Editor for Healio Psoriatic Disease (receiving honoraria); and is a member of the Board of Directors for the International Psoriasis Council, receiving no honoraria. D.B.S. has no disclosures to report.
Data availability statement:
The data that support the findings of this study are available from Cegedim Health Data, which oversees The Health Improvement Network. Restrictions apply to the availability of these data, which were used under license for this study.
References
- [1].Odhiambo JA, Williams HC, Clayton TO, Robertson CF, Asher MI. Global variations in prevalence of eczema symptoms in children from ISAAC Phase Three. The Journal of allergy and clinical immunology. 2009;124; 1251–1258.e1223. [DOI] [PubMed] [Google Scholar]
- [2].Silverberg JI, Hanifin JM. Adult eczema prevalence and associations with asthma and other health and demographic factors: a US population-based study. The Journal of allergy and clinical immunology. 2013;132; 1132–1138. [DOI] [PubMed] [Google Scholar]
- [3].Drucker AM, Wang AR, Li WQ, Sevetson E, Block JK, Qureshi AA. The Burden of Atopic Dermatitis: Summary of a Report for the National Eczema Association. The Journal of investigative dermatology. 2017;137; 26–30. [DOI] [PubMed] [Google Scholar]
- [4].Yaghmaie P, Koudelka CW, Simpson EL. Mental health comorbidity in patients with atopic dermatitis. The Journal of allergy and clinical immunology. 2013;131; 428–433. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Ronnstad ATM, Halling-Overgaard AS, Hamann CR, Skov L, Egeberg A, Thyssen JP. Association of atopic dermatitis with depression, anxiety, and suicidal ideation in children and adults: A systematic review and meta-analysis. Journal of the American Academy of Dermatology. 2018;79; 448–456.e430. [DOI] [PubMed] [Google Scholar]
- [6].Schonmann Y, Mansfield KE, Hayes JF, Abuabara K, Roberts A, Smeeth L, et al. Atopic Eczema in Adulthood and Risk of Depression and Anxiety: A Population-Based Cohort Study. The journal of allergy and clinical immunology In practice. 2020;8; 248–257.e216. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Liao TC, Lien YT, Wang S, Huang SL, Chen CY. Comorbidity of Atopic Disorders with Autism Spectrum Disorder and Attention Deficit/Hyperactivity Disorder. The Journal of pediatrics. 2016;171; 248–255. [DOI] [PubMed] [Google Scholar]
- [8].Lee CY, Chen MH, Jeng MJ, Hsu JW, Tsai SJ, Bai YM, et al. Longitudinal association between early atopic dermatitis and subsequent attention-deficit or autistic disorder: A population-based case-control study. Medicine. 2016;95; e5005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Farzanfar D, Dowlati Y, French LE, Lowes MA, Alavi A. Inflammation: A Contributor to Depressive Comorbidity in Inflammatory Skin Disease. Skin Pharmacol Physiol. 2018;31; 246–251. [DOI] [PubMed] [Google Scholar]
- [10].Yarlagadda A, Alfson E, Clayton AH. The blood brain barrier and the role of cytokines in neuropsychiatry. Psychiatry (Edgmont). 2009;6; 18–22. [PMC free article] [PubMed] [Google Scholar]
- [11].Park G, Lee SH, Oh DS, Kim YU. Melatonin inhibits neuronal dysfunction-associated with neuroinflammation by atopic psychological stress in NC/Nga atopic-like mouse models. Journal of pineal research. 2017;63. [DOI] [PubMed] [Google Scholar]
- [12].Abuabara K, Magyari AM, Hoffstad O, Jabbar-Lopez ZK, Smeeth L, Williams HC, et al. Development and Validation of an Algorithm to Accurately Identify Atopic Eczema Patients in Primary Care Electronic Health Records from the UK. The Journal of investigative dermatology. 2017;137; 1655–1662. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Lewis JD, Schinnar R, Bilker WB, Wang X, Strom BL. Validation studies of the health improvement network (THIN) database for pharmacoepidemiology research. Pharmacoepidemiology and drug safety. 2007;16; 393–401. [DOI] [PubMed] [Google Scholar]
- [14].Emerson RM, Williams HC, Allen BR. Severity distribution of atopic dermatitis in the community and its relationship to secondary referral. The British journal of dermatology. 1998;139; 73–76. [DOI] [PubMed] [Google Scholar]
- [15].NICE. Tacrolimus and pimecrolimus for atopic eczema. Technology appraisal guidance [TA82]. National Institute for Health and Care Excellence, 2014. [Google Scholar]
- [16].Silverwood RJ, Forbes HJ, Abuabara K, Ascott A, Schmidt M, Schmidt SAJ, et al. Severe and predominantly active atopic eczema in adulthood and long term risk of cardiovascular disease: population based cohort study. BMJ (Clinical research ed). 2018;361; k1786. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Patel KR, Immaneni S, Singam V, Rastogi S, Silverberg JI. Association between atopic dermatitis, depression, and suicidal ideation: A systematic review and meta-analysis. Journal of the American Academy of Dermatology. 2019;80; 402–410. [DOI] [PubMed] [Google Scholar]
- [18].Wei HT, Lan WH, Hsu JW, Huang KL, Su TP, Li CT, et al. Risk of developing major depression and bipolar disorder among adolescents with atopic diseases: A nationwide longitudinal study in Taiwan. J Affect Disord. 2016;203; 221–226. [DOI] [PubMed] [Google Scholar]
- [19].Cheng CM, Hsu JW, Huang KL, Bai YM, Su TP, Li CT, et al. Risk of developing major depressive disorder and anxiety disorders among adolescents and adults with atopic dermatitis: a nationwide longitudinal study. J Affect Disord. 2015;178; 60–65. [DOI] [PubMed] [Google Scholar]
- [20].Kern C, Wan J, LeWinn KZ, Ramirez FD, Lee Y, McCulloch CE, et al. Association of Atopic Dermatitis and Mental Health Outcomes Across Childhood: A Longitudinal Cohort Study. JAMA dermatology. 2021;157; 1200–1208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Hu C, Nijsten T, Pasmans S, de Jongste JC, Jansen PW, Duijts L. Associations of eczema phenotypes with emotional and behavioural problems from birth until school age. The Generation R Study. The British journal of dermatology. 2020;183; 311–320. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Cork MJ, Eckert L, Simpson EL, Armstrong A, Barbarot S, Puig L, et al. Dupilumab improves patient-reported symptoms of atopic dermatitis, symptoms of anxiety and depression, and health-related quality of life in moderate-to-severe atopic dermatitis: analysis of pooled data from the randomized trials SOLO 1 and SOLO 2. The Journal of dermatological treatment. 2020;31; 606–614. [DOI] [PubMed] [Google Scholar]
- [23].Drevets WC, Wittenberg GM, Bullmore ET, Manji HK. Immune targets for therapeutic development in depression: towards precision medicine. Nature reviews Drug discovery. 2022;21; 224–244. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].Sandhu JK, Wu KK, Bui TL, Armstrong AW. Association Between Atopic Dermatitis and Suicidality: A Systematic Review and Meta-analysis. JAMA dermatology. 2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Pompili M, Bonanni L, Gualtieri F, Trovini G, Persechino S, Baldessarini RJ. Suicidal risks with psoriasis and atopic dermatitis: Systematic review and meta-analysis. J Psychosom Res. 2021;141; 110347. [DOI] [PubMed] [Google Scholar]
- [26].Thyssen JP, Hamann CR, Linneberg A, Dantoft TM, Skov L, Gislason GH, et al. Atopic dermatitis is associated with anxiety, depression, and suicidal ideation, but not with psychiatric hospitalization or suicide. Allergy. 2018;73; 214–220. [DOI] [PubMed] [Google Scholar]
- [27].Singhal A, Ross J, Seminog O, Hawton K, Goldacre MJ. Risk of self-harm and suicide in people with specific psychiatric and physical disorders: comparisons between disorders using English national record linkage. Journal of the Royal Society of Medicine. 2014;107; 194–204. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [28].Drucker AM, Thiruchelvam D, Redelmeier DA. Eczema and subsequent suicide: a matched case-control study. BMJ open. 2018;8; e023776. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Kimata H Prevalence of suicidal ideation in patients with atopic dermatitis. Suicide Life Threat Behav. 2006;36; 120–124. [DOI] [PubMed] [Google Scholar]
- [30].Noh HM, Cho JJ, Park YS, Kim JH. The relationship between suicidal behaviors and atopic dermatitis in Korean adolescents. J Health Psychol. 2016;21; 2183–2194. [DOI] [PubMed] [Google Scholar]
- [31].Lee S, Shin A. Association of atopic dermatitis with depressive symptoms and suicidal behaviors among adolescents in Korea: the 2013 Korean Youth Risk Behavior Survey. BMC Psychiatry. 2017;17; 3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [32].Schmitt J, Romanos M, Schmitt NM, Meurer M, Kirch W. Atopic eczema and attention-deficit/hyperactivity disorder in a population-based sample of children and adolescents. Jama. 2009;301; 724–726. [DOI] [PubMed] [Google Scholar]
- [33].Schmitt J, Buske-Kirschbaum A, Roessner V. Is atopic disease a risk factor for attention-deficit/hyperactivity disorder? A systematic review. Allergy. 2010;65; 1506–1524. [DOI] [PubMed] [Google Scholar]
- [34].Schans JV, Çiçek R, de Vries TW, Hak E, Hoekstra PJ. Association of atopic diseases and attention-deficit/hyperactivity disorder: A systematic review and meta-analyses. Neurosci Biobehav Rev. 2017;74; 139–148. [DOI] [PubMed] [Google Scholar]
- [35].Strom MA, Fishbein AB, Paller AS, Silverberg JI. Association between atopic dermatitis and attention deficit hyperactivity disorder in U.S. children and adults. The British journal of dermatology. 2016;175; 920–929. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [36].Lin YT, Chen YC, Gau SS, Yeh TH, Fan HY, Hwang YY, et al. Associations between allergic diseases and attention deficit hyperactivity/oppositional defiant disorders in children. Pediatr Res. 2016;80; 480–485. [DOI] [PubMed] [Google Scholar]
- [37].Shyu CS, Lin HK, Lin CH, Fu LS. Prevalence of attention-deficit/hyperactivity disorder in patients with pediatric allergic disorders: a nationwide, population-based study. J Microbiol Immunol Infect. 2012;45; 237–242. [DOI] [PubMed] [Google Scholar]
- [38].Tsai JD, Chang SN, Mou CH, Sung FC, Lue KH. Association between atopic diseases and attention-deficit/hyperactivity disorder in childhood: a population-based case-control study. Ann Epidemiol. 2013;23; 185–188. [DOI] [PubMed] [Google Scholar]
- [39].van der Schans J, Pleiter JC, de Vries TW, Schuiling-Veninga CC, Bos JH, Hoekstra PJ, et al. Association between medication prescription for atopic diseases and attention-deficit/hyperactivity disorder. Annals of allergy, asthma & immunology : official publication of the American College of Allergy, Asthma, & Immunology. 2016;117; 186–191. [DOI] [PubMed] [Google Scholar]
- [40].Riis JL, Vestergaard C, Deleuran MS, Olsen M. Childhood atopic dermatitis and risk of attention deficit/hyperactivity disorder: A cohort study. The Journal of allergy and clinical immunology. 2016;138; 608–610. [DOI] [PubMed] [Google Scholar]
- [41].Qu X, Lee LC, Ladd-Acosta C, Hong X, Ji Y, Kalb LG, et al. Association between atopic diseases and neurodevelopmental disabilities in a longitudinal birth cohort. Autism Res. 2022;15; 740–750. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [42].Schmitt J, Apfelbacher C, Chen CM, Romanos M, Sausenthaler S, Koletzko S, et al. Infant-onset eczema in relation to mental health problems at age 10 years: results from a prospective birth cohort study (German Infant Nutrition Intervention plus). The Journal of allergy and clinical immunology. 2010;125; 404–410. [DOI] [PubMed] [Google Scholar]
- [43].Galéra C, Cortese S, Orri M, Collet O, van der Waerden J, Melchior M, et al. Medical conditions and Attention-Deficit/Hyperactivity Disorder symptoms from early childhood to adolescence. Mol Psychiatry. 2021. [DOI] [PubMed] [Google Scholar]
- [44].Genuneit J, Braig S, Brandt S, Wabitsch M, Florath I, Brenner H, et al. Infant atopic eczema and subsequent attention-deficit/hyperactivity disorder--a prospective birth cohort study. Pediatric allergy and immunology : official publication of the European Society of Pediatric Allergy and Immunology. 2014;25; 51–56. [DOI] [PubMed] [Google Scholar]
- [45].Johansson EK, Ballardini N, Kull I, Bergström A, Wahlgren CF. Association between preschool eczema and medication for attention-deficit/hyperactivity disorder in school age. Pediatric allergy and immunology : official publication of the European Society of Pediatric Allergy and Immunology. 2017;28; 44–50. [DOI] [PubMed] [Google Scholar]
- [46].Romanos M, Gerlach M, Warnke A, Schmitt J. Association of attention-deficit/hyperactivity disorder and atopic eczema modified by sleep disturbance in a large population-based sample. Journal of epidemiology and community health. 2010;64; 269–273. [DOI] [PubMed] [Google Scholar]
- [47].Fuhrmann S, Tesch F, Romanos M, Abraham S, Schmitt J. ADHD in school-age children is related to infant exposure to systemic H1-antihistamines. Allergy. 2020;75; 2956–2957. [DOI] [PubMed] [Google Scholar]
- [48].Billeci L, Tonacci A, Tartarisco G, Ruta L, Pioggia G, Gangemi S. Association Between Atopic Dermatitis and Autism Spectrum Disorders: A Systematic Review. American journal of clinical dermatology. 2015;16; 371–388. [DOI] [PubMed] [Google Scholar]
- [49].Zerbo O, Leong A, Barcellos L, Bernal P, Fireman B, Croen LA. Immune mediated conditions in autism spectrum disorders. Brain, behavior, and immunity. 2015;46; 232–236. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [50].Pedersen MS, Benros ME, Agerbo E, Borglum AD, Mortensen PB. Schizophrenia in patients with atopic disorders with particular emphasis on asthma: a Danish population-based study. Schizophr Res. 2012;138; 58–62. [DOI] [PubMed] [Google Scholar]
- [51].Khandaker GM, Zammit S, Lewis G, Jones PB. A population-based study of atopic disorders and inflammatory markers in childhood before psychotic experiences in adolescence. Schizophr Res. 2014;152; 139–145. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [52].Yuce M, Guner SN, Karabekiroglu K, Baykal S, Kilic M, Sancak R, et al. Association of Tourette syndrome and obsessive-compulsive disorder with allergic diseases in children and adolescents: a preliminary study. Eur Rev Med Pharmacol Sci. 2014;18; 303–310. [PubMed] [Google Scholar]
- [53].Lowe KE, Mansfield KE, Delmestri A, Smeeth L, Roberts A, Abuabara K, et al. Atopic eczema and fracture risk in adults: A population-based cohort study. The Journal of allergy and clinical immunology. 2020;145; 563–571.e568. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from Cegedim Health Data, which oversees The Health Improvement Network. Restrictions apply to the availability of these data, which were used under license for this study.