Objective: We aimed to assess whether a symptom perception accuracy measure can be derived from routinely collected general population cohort data.
Methods: We combined information on self-reported dyspnea and physiological parameters (FEV1%pred, body weight) from the Lifelines Cohort Study (N=138,594; 59.0% female; mean age=42.3 y [SD=11.0]) to obtain a symptom perception accuracy measure. Dyspnea was operationalized via the SCL-90 SOM subscale item. Using principal component analysis of available psychosocial variables known to correlate with symptom perception, we derived three compound scores reflecting negative affect, fear of illness, and worries of contracting disease. We used multinominal regression analyses to calculate the probability of self-reported dyspnea being correctly classified based on FEV1%pred and body weight. Via multivariable logistic regression we assessed whether the dichotomized probability of correct classification is associated with derived compound scores.
Results: The symptom perception accuracy measure was non-normally distributed in control and participants with asthma/COPD. Fear of illness (OR=0.85; 95%CI=0.79-0.90 and OR=0.84; 95%CI=0.72-0.98) was negatively associated with the accuracy measure in control and asthma/COPD participants, respectively. Negative affect (OR=0.76; 95%CI=0.65-0.90) was associated negatively with the accuracy measure in asthma/COPD participants. Worries of contracting disease was associated with the measure in control participants (OR=0.88; 95%CI=0.83-0.94). Physiological parameters explain 1.6%-2.5% of the variance in self-reported dyspnea; addition of aforementioned compound scores increases this to 9.5%-16.6%.
Conclusion: We show that a symptom perception accuracy measure based on congruence between physiological parameters (FEV1%pred, body weight) and self-reported dyspnea can be developed. It is associated with known psychosocial correlates of symptom perception. The psychosocial factors explained more variance in self-reported dyspnea than physiological parameters.