The dilemma of who is actually OSA-free/healthy

Recently, our team published a systematic review assessing the association between craniofacial features in children and adolescents with pediatric obstructive sleep apnea (OSA).¹ We received Wang and colleagues’ letter,² describing a solid concern regarding a possible bias in the eligibility criteria adopted in our review. We appreciate their interest in our review and praise such discussion on this controversial topic. We are happy to address their opinions.

The authors described that polysomnography (PSG)-negative children are not always healthy controls, and we agree. Our review only considered healthy controls as those with a negative diagnosis for OSA through a PSG consideration. We did not include studies that only reported PSG results without an associated medical specialist assessment. We followed the American Academy of Sleep Medicine International Classification of Sleep Disorders to define the presence of OSA in children over two years old, defined by overnight PSG. Considering the PSG is the gold standard test, we followed this criterion also to define a negative diagnosis of pediatric OSA.

Adopting the PSG-based diagnosis might have limitations. We made our best effort to acknowledge these limitations in our Discussion section. We provided a supplementary table with data from all retrieved studies that presented healthy children without OSA symptoms and without PSG exam as a control group (see supplemental material, Table S5).¹

As we described in the supplementary table, articles without a negative PSG group often reported an increased range of differences when comparing control and OSA groups, such as a narrow palate and increased facial height. The differences observed among a non-PSG healthy control group and a negative-PSG healthy control group could also be associated with other factors rather than solely sleep features reported by a PSG.³

It is challenging, if not questionable, to have children without sleep-disordered breathing signs and symtoms go through PSG to confirm a negative presumption. Biases that can be incorporated include parents with excessive concerns about minor mouth-breathing events or parents with a history of OSA. Nevertheless, it is essential to acknowledge that any definition adopted as a “healthy” control group within published studies might be a limitation factor. The option of only considering healthy controls as participants without a negative-PSG diagnosis may also have drawbacks, including that cases with OSA that were not fully diagnosed would be considered 100% healthy. Apparent “healthy” controls are not always OSA-free individuals.

In summary, the definition of a control group for future pediatric OSA studies within clinical settings must be better defined to avoid potential bias. Current status may contribute to the diversity of conclusions in the associated recent systematic reviews.

DISCLOSURE STATEMENT

All authors have read and approved this letter. Institution where work was performed: University of Alberta–Canada. Nathalia Fernandes Fagundes was sponsored by CAPES Foundation (Brazilian Federal Agency for Support and Evaluation of Graduate Education (88881.128202/2016-01). The authors report no conflicts of interest.

ABBREVIATIONS

OSA

obstructive sleep apnea

PSG

polysomnography