Table 2

Studies presenting prevalence estimates* by independent application of several case definitions on the same population (model A)

First author, year, countryData collectionPrevalence (95% CI) (%)
Nacul,50 2011, UK609 possible cases electronically identified in databases of 29 GP practices. 70 excluded after clinical revision (explained fatigue), 135 refusals and 126 non-casesECD: 0.03 (0.02 to 0.04)
Canada: 0.10 (0.09 to 0.12)
Fukuda: 0.19 (0.17 to 0.21)
Bates,56 1993, USA995 consecutive GP visitors invited—94% screened by a questionnaire to detect major fatigue. Selected patients further evaluated by questionnaires, physical examinations and interviewsHolmes: 0.3 (0.1 to 0.9)
Oxford: 0.4 (0.1 to 1.1)
Australia: 1.1 (0.5 to 2.0)
Kawakami,57 1998, JapanAll adults (n=508) in Town A, Kofu-city, were invited to participate in this structured psychiatric diagnostic interview survey. 137 (27%) completed the studyHolmes: 0.0 (0.0 to 2.7)
Fukuda: 1.5 (0.2 to 5.2)
Oxford: 1.5 (0.2 to 5.2)
Lindal,55 2002, IcelandSurvey sent to 4000 randomly selected adult participants—63% responded. Questionnaire included questions on all items in the four case definitions. Diagnoses were set electronically based on received responses. No medical tests or examinations were undertakenHolmes 0.0 (0.0 to 1.5)
Fukuda: 2.1 (1.6 to 2.8)
Oxford: 3.7 (3.2 to 4.6)
Australia: 7.6 (6.6 to 8.7)
Wessely,54 68 1997, UK2363 patients followed in a cohort study—84% completed. Fatigued participant subjected to detailed questionnaires, interviews and laboratory testing. Separate estimates reported for inclusion/exclusion of psychiatric comorbidityHolmes: 1.2 (0.5 to 1.8)
Australia: 1.4 (0.8 to 2.0)
Oxford: 2.2 (1.4 to 3.0)
Fukuda: 2.6 (1.7 to 3.4)
  • *Prevalence estimates were calculated with the number of responders in the denominator. The choice of denominator may have large implications with regard to the subsequent prevalence estimate, particularly in studies with low response rate. Hence, depending on the actual response rate, estimates presented for each study may be biased.

  • GP, general practitioner.