Table 2.
Summary of findings
| Allergy diseases in children with ADHD | |||||
|---|---|---|---|---|---|
| Population: Children with ADHD Setting: Germany, Korea, Taiwan, Thailand, UK Intervention: none (observation of the difference in risk of allergy diseases) Comparison: Children without ADHD Outcome: Allergic diseases | |||||
| Allergy diseases | Anticipated absolute effects* (95% CI) | Relative effect (95% CI) | № of participants (studies) | Quality of the evidence (GRADE) | |
| Control group risk (Assumed risk) | ADHD group risk (Corresponding risk) | ||||
| Asthma | 125 per 1000 | 205 per 1000 (183 to 228) | OR 1.80 (1.57 to 2.07) | 59,646 (5 studies) | ⨁⨁◯◯ LOW a,h |
| a) Nationwide studies | a) 71 per 1000 | a) 130 per 1000 (113 to 149) | a) OR 1.96 (1.67 to 2.30) | a) 51,033 (2 studies) | a) ⨁◯◯◯ VERY LOW b,h |
| b) Institutional-based studies | b) 241 per 1000 | b) 347 per 1000 (320 to 374) | b) OR 1.67 (1.48 to 1.88) | b) 8613 (3 studies) | b) ⨁◯◯◯ VERY LOW c.h |
| Allergic rhinitis | 153 per 1000 | 222 per 1000 (169 to 286) | OR 1.59 (1.13 to 2.23) | 59,646 (5 studies) a) |
⨁◯◯◯ VERY LOW a,d,e,h |
| a) Nationwide studies | a) 134 per 1000 | a) 205 per 1000 (129 to 312) | a) OR 1.67 (0.96 to 2.93) | b) 51,033 (2 studies) | a) ⨁◯◯◯ VERY LOW b,d,e,h |
| b) Institutional-based studies | b) 325 per 1000 | b) 416 per 1000 (325 to 511) | b) OR 1.48 (1.00 to 2.17) | c) 8613 (3 studies) | b) ⨁◯◯◯ VERY LOW c,d,e,h |
| Atopic dermatitis | 100 per 1000 | 137 per 1000 (108 to 173) | OR 1.43 (1.09 to 1.88) | 59,646 (5 studies) | ⨁◯◯◯ VERY LOW a,d,h |
| a) Nationwide studies | a) 94 per 1000 | a) 154 per 1000 (124 to 188) | a) OR 1.74 (1.36 to 2.22) | a) 51,033 (2 studies) | a) ⨁◯◯◯ VERY LOW b,d,h |
| b) Institutional-based studies | b) 100 per 1000 | b) 114 per 1000 (100 to 130) | b) OR 1.16 (1.00 to 1.35) | b) 8613 (3 studies) | b) ⨁◯◯◯ VERY LOW c,h |
| Allergic Conjunctivitis | 203 per 1000 | 301 per 1000 (210 to 413) | OR 1.69 (1.04 to 2.76) | 41,908 (3 studies) | ⨁◯◯◯ VERY LOW c,d,e,h |
| a) Nationwide studies | a) 203 per 1000 | a) 347 per 1000 (333 to 360) | a) OR 2.08 (1.96 to 2.21) | a) 37,715 (1 study) | a) ⨁⨁◯◯ LOW f |
| b) Institutional-based studies | b) 175 per 1000 | b) 224 per 1000 (144 to 334) | b) OR 1.36 (0.79 to 2.36) | b) 4193 (2 studies) | b) ⨁◯◯◯ VERY LOW g,d,e,h |
| Food allergy Institutional-based studies |
75 per 1000 | 84 per 1000 (67 to 106) | OR 1.13 (0.88 to 1.47) | 8613 (3 studies) | ⨁◯◯◯ VERY LOW c,e,h |
*The risk in the ADHD group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the exposure (and its 95% CI)
CI confidence interval, OR odds ratio
GRADE Working Group grades of evidence (level of evidence of grading for observational studies)
Very low: Observational studies with uncertainty about the directness of results or unsystematic observations
Low: Observational studies with no threats to validity
Moderate: Observational studies with no threats to validity and evidence of a dose-response or exposure-response gradient
High: Observational studies with no threats to validity yielding very large effects
a Four of the studies had limitation on the selection of participants and blinding of outcomes assessments by their study designs but one study had unclear risk of bias in the measurement of exposure and one study indicated high risk of bias on the outcome reporting which lowered the quality of the observational evidence
b The proportion of information was from two studies indicated with limitation on selection of participants and blinding by their study designs but the unclear risk of selective reporting which lowered the quality of the observational evidence
c The proportion of information was from two studies indicated limitation on selection of participants and blinding of outcome assessments by their study designs but one study had high risk of outcome reporting, which lowered the quality of the observational evidence
d There is an indication of significant inconsistency (I2 > 80%)
e Information were from high heterogeneity and small sample size with a wide confidence interval
f The information is based on one study, which had limitation on the selection of participants, blinding of outcomes assessments and selective outcome reporting by the study design
g The information from two studies that had limitation on selection of participants, measurement of exposure, blinding of outcome assessment by their study designs but one with high risk of bias on selective outcome reporting which lowered the quality of the observational evidence
h The possibility of publication bias is not disregarded but it was not considered to downgrade the quality of the observational evidence