4.3 Strengths and limitations
This study has some strengths that support the validity of our findings. First, we observed the association of eczema phenotypes with allergic multimorbidity in the largest cohort, (N = 6,852) to date. Further, all allergic diseases, which are considered in the atopic march, were included in the analysis. The overall prevalence of allergic diseases that we observed are in line with other reports.14,15,36-39 However, eczema prevalence was slightly higher than commonly reported.15,17,38,39This could be, partly, due to the self-/guardian-reported factors utilized for verification of eczema presence. Prevalence of eczema phenotypes resembled the distribution of phenotypes in the cohorts that were used to define them initially, with the most prevalent phenotypes (besides “Never eczema”) being “Very early onset – remitting” and “Early onset – remitting”.7 Further, our investigation covers a wide age-range (0 – 17 years) including infancy, childhood, and adolescence, while many other studies assess outcomes at a younger age. Regression models were adjusted comprehensively for confounders, known to influence allergic disease development. Lastly, due to the population-based study design, our findings are derived from real-life data and do not rely on selective sampling, leading to results representative for the general population.
However, some limitations need to be addressed. Firstly, the data used in this study was derived from self-/guardian-reported questionnaires. Additionally, the calculation of age of onset displayed some imprecision, as it was based on the age brackets which had been predefined with differing durations, resulting in estimates within those age brackets. Further, recall bias at baseline may have been present, since subjects had to report presence, treatment and medication of allergic diseases since birth. Lastly, due to the Lifelines cohort exclusively consisting of participants based in the Northern Netherlands, our results are subject to a regional bias. However, as described above, our data is in line with findings from comparable cohorts.