Sleep-Disordered Breathing and C-Reactive Protein: A Commentary

C-REACTIVE PROTEIN (CRP) HAS BEEN EXTENSIVELY STUDIED IN THE LAST DECADE BECAUSE OF ITS POSSIBLE ROLE AS AN EARLY MARKER OF CARDIOVASCULAR disease. More recently, sleep disordered breathing (SDB) has been investigated as a possible cause of increase in CRP. To date, no definite conclusion has been reached on the importance of SDB in low-grade systemic inflammation. The article by Punjabi and Beamer published in the January issue¹ reported a strong association between CRP levels and sleep disordered breathing in 69 men, even after adjusting for several measures of obesity. Unfortunately, this study did not shed new light on this issue.

Adiposity and CRP are so strongly related that complete adjustment may not be achieved by statistical methods. It is also troubling that adjustments were made for the 3 adiposity measures (body mass index, waist circumference, and body fat percentage): while they showed strong correlations between them, introducing the 3 variables in the same model created a multi-collinearity effect, rendering it difficult or impossible to distinguish their individual influences on CRP and thus devaluating their effect. Furthermore, the mean BMI of the sample was so high that it is likely that the majority of the subjects were overweight or obese. Therefore, adjusting for a condition that is present in nearly all the participants has little or no impact on the outcome. The same can be said about sleep duration. Although the authors did not provide the total sleep time, this variable was introduced in the multivariate regression models. A recent study has shown that even partial sleep deprivation in healthy subjects provokes an increase in CRP levels.² This is another reason why the authors found that the adjusted models have only small attenuation in the association between AHI and CRP.

Even with adjustment for known risk factors, residual confounding was likely to occur because no correction for intra-individual variability was done in major risk factors that have reversible effects on CRP, such as tobacco smoke exposure or alcohol intake. More recently, new factors have been found to have a profound impact on CRP levels.^3,4

There is not yet a lot of evidence showing that treating SDB reduces the CRP levels. Two intervention studies, one using nC-PAP and the other using surgical interview, reported a significant decrease of CRP levels after intervention.⁵ There was no information about the magnitude of weight changes in these patients. Two long-term intervention studies with CPAP^6,7 found no significant change in CRP levels, these last being associated only with BMI. A recent systematic review of 33 studies on weight loss intervention pointed out that for each 1 kilogram loss the mean change in CPR level was −0.13 mg/L.⁸ The interaction between CRP, SDB, and adiposity is complex. Using samples composed mostly of overweight or obese individuals cannot isolate the adiposity effect on the relationship between CRP and SDB, even with sophisticated statistical analyses, because weight excess is implicitly part of the sample definition.