Do Sleepy and Tired Go Together? Rasch Analysis of the Relationships between Sleepiness, Fatigue and Nonrestorative Sleep Complaints in a Nonclinical Population Sample

Objective: The lack of distinction in the clinical use of terms like fatigue and sleepiness is an important issue. While both fatigue and sleepiness can potentially be associated with nonrestorative sleep (NRS) complaints, their relationships are still poorly described. We propose to use Rasch analysisbased methods to study the interrelations of fatigue, sleepiness and NRS. Methods: 150 subjects (mean age = 39.3 years, range = 18–65) from a community sample underwent a structured computer-assisted web interview. We assessed demographic data, sleep habits, and subjective fatigue with the Fatigue Severity Scale (FSS), global and situational sleepiness with the Epworth Sleepiness (ESS) and the Stanford Sleepiness Scales, respectively, and affective symptoms with the Hospital Anxiety and Depression Scale. Dimensionality, measurement invariance and common person equating were investigated to study the FSS, ESS and their relations to NRS. Results: NRS was linked to shorter habitual sleep duration and to higher scores on psychometric scales. Both sleepReceived: September 23, 2009 Accepted: January 29, 2010 Published online: March 25, 2010 Dr. Daniel Neu, MD Sleep Laboratory and Unit for Chronobiology U78, Department of Psychiatry Brugmann University Hospital, Université Libre de Bruxelles Building Hh, Arthur Van Gehuchten Square, BE–1020 Brussels (Belgium) Tel. +32 2 477 25 54, Fax +32 2 477 21 62, E-Mail daniel.neu @ chu-brugmann.be © 2010 S. Karger AG, Basel 0251–5350/10/0351–0001$26.00/0 Accessible online at: www.karger.com/ned D ow nl oa de d by : 54 .1 91 .4 0. 80 9 /1 6/ 20 17 7 :5 0: 27 P M Neu /Mairesse /Hoffmann /Valsamis / Verbanck /Linkowski /Le Bon Neuroepidemiology 2010;35:1–11 2 ten reported in general population samples and both fatigue and sleepiness can present heavy burdens for public health care [4–7] . Sleepiness is generally described as a trigger signal for a spontaneous onset of sleep. It is first of all a physiological phenomenon depending on previous sleep and occurring at regular intervals following a circadian rhythm [8, 9] . But in pathological conditions, excessive daytime sleepiness can be irrepressible when associated with narcolepsy or sleep apnea-hypopnea syndrome for instance. Excessive daytime sleepiness is also linked to other sleep disorders such as periodic limb movement disorder, idiopathic hypersomnia or sleep deprivation [5] . On the other hand, fatigue is generally described as a condition in which maintaining of a motor or mental energy level gets more difficult with duration of exercise. Fatigue needs rest not sleep to recover from. Chronic severe daytime fatigue is, for instance, the core symptom in clinical conditions like chronic fatigue syndrome. Nevertheless, even in presumably exclusive fatigue conditions such as chronic fatigue syndrome significantly overlapping subjective sleepiness has recently been described [10] . Although classical psychometric assessment tools like the Epworth Sleepiness (ESS) and the Fatigue Severity Scales (FSS) are available for the assessment of subjective fatigue and sleepiness complaints, the definition and semiological distinction of both concepts often remain difficult for both patients and clinicians [5, 11] . Moreover, although both fatigue and sleepiness can be associated with complaints of NRS [1] , their relationships to each other and to sleep remain insufficiently understood. These are important issues as the etiopathogenesis of related clinical conditions and the implications for therapeutic orientation can be very different [5, 12] . Potentially overlapping descriptive features can also lead to imprecise diagnosis and subsequently to inadequate or insufficient treatment attitudes [5, 6, 12] . Shen et al. [13] pointed towards the coexistence of both phenomena and to the fact that both can potentially be related to sleep deprivation. However, only a few previous studies reported associations between sleepiness and fatigue [11, 14–16] . As a contribution to the disentangling of fatigue and sleepiness, Bailes et al. [11] proposed two empirical scales measuring either fatigue or sleepiness. The authors found significant correlations between the ESS and the FSS ranging from 0.16 to 0.42. Subsequently, they developed the Empirical Fatigue and Sleepiness Scales based on a discriminant validity rationale by retaining items not significantly correlated with items of the opposite construct. The authors reported the ability of those scales to identify ‘sleepiness which is not fatigue’ [11] . Furthermore, a Rasch methodology-based analysis has recently been applied separately to qualify the measurement properties and stable hierarchical item structure of the ESS and the FSS in Parkinson’s disease [17, 18] . However, due to the absence of combined explorations, it is presently unclear how fatigue and sleepiness relate to each other and to NRS. Hence, in order to assess if individuals perceive subjective sleepiness and fatigue in a similar way and if they are related differently to NRS, we propose to use Rasch-based common person equating [19] . Common person equating is conventionally used to test the equivalence among examinee populations and evaluate different tests that measure the same construct, administered to a common group of persons [20] . In the present study, a somewhat different approach is used regarding common person equating. We hypothesize that examinees confounding sleepiness with fatigue or vice versa will have similar Rasch-derived person estimates on the ESS and the FSS or put differently, examinees will perceive fatigue and sleepiness as undifferentiated concepts. Additionally, by linking persons with or without NRS with similar fatigue and sleepiness levels, different response probabilities to specific ESS and FSS items can be investigated (i.e. differential item functioning or DIF) which eventually allows for a substantive examination of the perceptive associations between NRS, sleepiness and fatigue.