Various studies have shown that unexplained fatigue syndromes are common conditions in both the community (Reference Walker, Katon and JemelkaWalker et al, 1993; Reference Steele, Dobbins and FukudaSteele et al, 1998; Reference Skapinakis, Lewis and MeltzerSkapinakis et al, 2000) and primary care (Reference Bates, Schmitt and BuchwaldBates et al, 1993; Reference McDonald, David and PelosiMcDonald et al, 1993; Reference Hickie, Hooker and Hadzi-PavlovicHickie et al, 1996; Reference Wessely, Chalder and HirschWessely et al, 1997). Fatigue is also common as a presenting complaint to primary care physicians (Reference Cathebras, Robbins and KirmayerCathebras et al, 1992; Reference Fuhrer and WesselyFuhrer & Wessely, 1995). The majority of the studies, however, were carried out in well-developed countries. Data from less-developed countries are relatively sparse. The present paper aims to report the cross-cultural differences in the prevalence of fatigue syndromes, using data from the World Health Organization (WHO) international study of psychological problems in general health care carried out in 14 countries (Reference Sartorius, Ustun and Costa e SilvaSartorius et al, 1993). We also examine the hypothesis that individuals from well-developed countries will be more likely to report fatigue compared with subjects from less-developed countries.
METHODS
General description of the data-set
The WHO collaborative study of psychological problems in general health care was an international prospective study carried out in 15 centres from 14 countries that examined the prevalence, 1-year outcome and public health implications of common mental disorders in primary care (Reference Sartorius, Ustun and Costa e SilvaSartorius et al, 1993). Details on the methods of the study are given elsewhere (Reference Von Korff, Üstün, Üstün and SartoriusVon Korff & Üstün, 1995). Briefly, the study used a two-phase design in which 26 969 primary care attenders aged between 18 and 64 years were approached in each participating centre and asked to complete the 12-item general health questionnaire (GHQ—12; Reference Goldberg and WilliamsGoldberg & Williams, 1988). A total of 25 916 subjects agreed to participate (96% response rate, range across centres from 91% to 100%). Patients were selected for the second phase assessment using a stratified random sampling procedure according to site-specific GHQ—12 thresholds (100% of the subjects scoring above the 80th percentile, 35% of those scoring between the 60th and 80th percentile and 10% of those scoring below the 60th percentile). The second phase assessment included the Composite International Diagnostic Interview (CIDI; Reference Wittchen, Robins and CottlerWittchen et al, 1991) modified for use in primary care and involved 5438 out of 8698 eligible subjects (62% response rate, range from 43% to 99%). Data collection took place between May 1991 and April 1992.
In a previous paper we reported the overall prevalence of unexplained fatigue syndromes and their association with psychiatric disorders (Reference Skapinakis, Lewis and MavreasSkapinakis et al, 2003), whereas in this paper we focus on the differences between centres.
Measures
The CIDI is a fully structured interview developed for use in cross-cultural psychiatric epidemiology studies (Reference Wittchen, Robins and CottlerWittchen et al, 1991). The modified version used in the present study includes only the sections that assess mental symptoms common in primary care, notably the sections on somatisation, anxiety, depression and hypochondriasis, as well as a new section on neurasthenia. The primary care version rates both current (1-month) and life-time symptomatology. Non-English-speaking participating centres translated and back-translated the interview. Training and procedures for assuring reliability are described elsewhere (Reference Von Korff, Üstün, Üstün and SartoriusVon Korff & Üstün, 1995). The interviewer—observer reliability coefficient for the primary care version of CIDI was found to be 0.92 overall, ranging between 0.81 and 1.00 for individual sections.
Measurement of fatigue: substantial unexplained fatigue
Fatigue was assessed using the neurasthenia section of the primary care version of the CIDI. Three screening questions were put to all subjects: Q1 ‘In the past month have you felt tired all the time?’; Q2 ‘Do you get easily tired while performing everyday tasks?’; and Q3 ‘Does even minimal physical effort cause exhaustion?’ Then the interviewer asked a specific sequence of questions to determine the clinical importance and possible cause of the symptom. Fatigue was considered ‘medically explained’ if a doctor had given the patient a definite diagnosis or if there had been any abnormalities reported on examination or further investigation. Subjects with medically unexplained fatigue were also asked a fourth question that assessed the severity of fatigue: Q4 ‘Is it difficult to recover from these periods of fatigue or exhaustion when you rest?’
Subjects with medically unexplained fatigue (at least one positive answer to questions Q1-Q3) who answered positively to question Q4 were classified as cases of substantial unexplained fatigue. In comparison with the ICD—10 definition of neurasthenia (World Health Organization, 1998), our definition of unexplained fatigue differs in that: it does not include the multiple somatic symptoms criterion (Criterion B in ICD—10); it refers to 1-month duration instead of 3 (Criterion D in ICD—10); and, it does not exclude other comorbid psychiatric disorders (Criterion E in ICD—10). Therefore, it is a much broader definition compared with ICD—10 neurasthenia.
Fatigue as the main reason for consultation
The above definition of fatigue is independent of whether the subjects complained of fatigue to their primary care physician. For comparison, we also report the prevalence of fatigue as a presenting complaint. Subjects were asked to report the three main reasons for their consultation, choosing from a list of symptoms. Those who reported weakness or lethargy (the only items in the list related to fatigue) as one of their main reasons for consultation were considered as having fatigue as a presenting complaint.
Measurement of morbidity
Psychiatric morbidity was assessed with the CIDI. Diagnostic algorithms were developed to give diagnoses according to the ICD—10 criteria. For the purposes of the present study, subjects were classified as cases of psychiatric morbidity if they had any of the following current ICD—10 disorders: depressive disorders (including dysthymia); generalised anxiety disorder; agoraphobia; panic disorder; somatisation disorder; and hypochondriasis.
Chronic physical morbidity was assessed by asking patients whether they were suffering from a list of common chronic medical conditions. Patients were classified as cases of chronic physical morbidity if they were suffering from at least one chronic physical disorder.
Classification of primary care centres
Centres were entered into the analysis as dummy variables. However, for the economic development hypothesis, centres were classified into three categories according to the gross national income (GNI) per head in 2000 as follows: high-income countries, with more than US$10 000 GNI per head (Athens, Berlin, Groningen, Mainz, Manchester, Nagasaki, Paris, Seattle, Verona); middle-income countries, with less than US$10 000 but more than US$1000 (Ankara, Rio de Janeiro, Santiago); and low-income countries, with less than US$1000 (Bangalore, Ibadan, Shanghai). Data for the GNI per head were derived from the World Bank databases available on the internet (World Bank, 2000).
Analysis
All data analyses were conducted using Stata version 6.0 (Stata Corporation, 1999). The weighted prevalence with 95% confidence intervals of fatigue syndromes was estimated using the SVYPROP command. This command allows for sampling weights and is suitable for the analysis of the two-phase design of the study (Reference Dunn, Pickles and TansellaDunn et al, 1999). The association of GNI per head with fatigue was analysed by means of a series of logistic regression models (separately for each fatigue syndrome) using the SVYLOGIT command in Stata. We used fatigue case status (Yes—No) as the dependent variable, and the classification of centres according to GNI as the independent variable, adjusting for socio-demographic variables, psychiatric morbidity and physical morbidity. Odds ratios (with 95% CIs) of fatigue were calculated for each category of the GNI variable. Psychiatric and physical morbidity were entered into the models as binary variables. In all the analyses we used sampling weights.
RESULTS
Description of the sample
The sample at the baseline assessment (n=5438) was predominantly female (62%), 58% of the participants were older than 35 years, 62.1% were married, 57.4% had basic education and 58.7% were employed. As expected, socio-demographic characteristics differed significantly between centres, reflecting the diverse cultures that are represented in this data-set.
Prevalence of fatigue syndromes
Table 1 shows the prevalence of substantial unexplained fatigue and fatigue as a presenting complaint across centres.
Prevalence of substantial unexplained fatigue and fatigue as a presenting complaint in international primary care (n=5438)
| Unweighted number of cases/Total number interviewed Weighted prevalence % (95% CI) | ||
|---|---|---|
| Centre1 (Response rate)2 | Substantial unexplained fatigue3 | Fatigue as a presenting complaint4 |
| Manchester (71%) | 115/428 15.05 (10.85-20.49) | 2/428 0.19 (0-0.76) |
| Santiago (46%) | 61/274 12.86 (8.41-19.18) | 3/274 2.17 (0.47-9.52) |
| Berlin (43%) | 70/400 12.53 (8.77-17.6) | 19/400 3.13 (1.81-5.36) |
| Groningen (69%) | 78/340 11.75 (8.48-16.05) | 26/340 6.90 (3.71-12.50) |
| Paris (62%) | 69/405 10.89 (7.74-15.11) | 59/405 10.35 (7.35-14.38) |
| Mainz (37%) | 56/400 9.91 (7.27-13.38) | 4/400 1.01 (0.31-3.22) |
| Nagasaki (74%) | 37/336 8.05 (5.08-12.54) | 62/336 10.87 (7.74-15.06) |
| Athens (43%) | 22/196 8.00 (4.29-14.45) | 4/196 3.79 (1.14-11.85) |
| Ankara (97%) | 47/400 7.99 (5.11-12.28) | 46/400 10.64 (6.94-15.97) |
| Rio de Janeiro (46%) | 50/393 6.71 (4.30-10.33) | 14/393 3.83 (1.69-8.45) |
| Bangalore (83%) | 30/398 3.83 (2.46-5.91) | 95/398 17.12 (13.18-21.94) |
| Shanghai (99%) | 44/576 3.63 (2.34-5.59) | 47/576 5.98 (3.83-9.20) |
| Seattle (61%) | 26/373 3.37 (1.90-5.90) | 3/373 0.25 (0-0.78) |
| Verona (55%) | 18/250 2.34 (1.41-3.86) | 7/250 1.84 (0.54-6.06) |
| Ibadan (88%) | 11/269 2.26 (1.17-4.33) | 47/269 16.09 (11.03-22.88) |
| All centres (63%) | 734/5438 7.99 (7.13-8.85) | 438/5438 6.27 (5.47-7.18) |
| χ2=6.64, d.f.=14, P < 0.001 | χ2=10.51, d.f.=14, P < 0.001 | |
Prevalence of substantial unexplained fatigue differed fifteen-fold across primary care centres (P<0.001). Centres with high prevalence of substantial unexplained fatigue (>10%) were Manchester, Santiago, Berlin, Groningen, Paris and Mainz, and centres with low prevalence (<4%) were Bangalore, Shanghai, Seattle, Verona and Ibadan.
Fatigue as a presenting complaint (either weakness or lethargy) was reported by 6.27% of the subjects (95% CI 5.47-7.18) with wide variation across centres. Centres with high prevalence of fatigue as a presenting complaint (>10%) were Bangalore, Ibadan, Nagasaki, Ankara and Paris, whereas centres with low prevalence (<2%) were Manchester, Seattle, Mainz and Verona.
In the whole data-set, very few subjects with substantial unexplained fatigue presented with fatigue (11%; 95% CI 8.21-14.57).
Association with level of economic development
In the logistic regression analysis, unexplained fatigue was associated positively with GNI per head after adjustment for all other socio-demographic variables, psychiatric morbidity and chronic physical morbidity. High-income countries had an odds ratio of 2.62 (95% CI 1.67-4.11) compared with low-income countries (Table 2).
Association of gross national income (GNI) per head and substantial unexplained fatigue or fatigue as a presenting complaint in international primary care
| Unweighted number of cases/Total number interviewed Odds ratios (95% CI)1 | ||
|---|---|---|
| GNI per head2 | Substantial unexplained fatigue3 | Fatigue as presenting complaint4 |
| Low income | 85/1243 | 189/1243 |
| 1.00 (reference) | 1.00 (reference) | |
| Middle income | 158/1067 | 63/1067 |
| 1.67 (1.04-2.66) | 0.39 (0.23-0.68) | |
| High income | 491/3128 | 186/3128 |
| 2.62 (1.67-4.11) | 0.38 (0.25-0.58) | |
In contrast, fatigue as a presenting complaint was associated negatively with GNI per head, and subjects from higher-income countries had an odds ratio of 0.38 (95% CI 0.25-0.58) compared with subjects from lower income countries.
DISCUSSION
In this multi-national study in primary care, we found that unexplained fatigue was present in all cultures but the prevalence varied widely. Subjects from higher-income countries were more likely to report fatigue in response to direct questions compared with subjects from lower-income countries. In contrast, subjects in higher-income countries were less likely to complain to their doctors of fatigue than those in poorer countries.
Limitations and strengths
These results should be interpreted in the context of the following limitations. First, this is a study carried out in primary care and therefore the results cannot be generalised to the general population. Second, response rates for the second phase baseline assessment were below 50% for five centres. Therefore, a systematic bias in either direction cannot be ruled out, even though participation was not related significantly to age, gender or screening GHQ—12 score. Third, although CIDI has been developed for use in cross-cultural epidemiological research, this does not prove its cultural validity. Fourth, medical causes of fatigue were excluded in a crude way by asking patients if a doctor had given them a definite diagnosis for their symptom or if there had been any abnormalities reported on examination or further investigation.
Despite these limitations our study had the advantage of investigating fatigue in a large multicultural sample using the same methodology. We are not aware of any other studies that used such a culturally diverse sample. The present research was therefore able to study fatigue independently of the confounding effect of the sociocultural context.
Prevalence of fatigue syndromes across centres
The prevalence of substantial unexplained fatigue differed significantly across centres, with an average prevalence of 7.99% (95% CI 7.13-8.85) in primary care but with a range between 2.26 and 15.05 in different countries.
The prevalence estimates reported from studies carried out in Western countries are generally consistent with the estimates provided here for similar countries. For example, Buchwald et al (Reference Buchwald, Sullivan and Komaroff1987) found a prevalence of unexplained chronic fatigue of 21% in an American primary care centre. In Australia, 24% of the primary care attenders reported substantial unexplained fatigue (Reference Hickie, Hooker and Hadzi-PavlovicHickie et al, 1996), whereas in Canada 14% reported this (Reference Cathebras, Robbins and KirmayerCathebras et al, 1992). In UK primary care the prevalence of unexplained fatigue has been found to be approximately 10% (Reference David, Pelosi and McDonaldDavid et al, 1990; Reference Wessely, Chalder and HirschWessely et al, 1997). By contrast, studies from the more-developed countries that used more culturally heterogeneous samples reported lower prevalence rates compared with the rates mentioned previously, for example 2% in a multiracial sample in San Francisco (Reference Steele, Dobbins and FukudaSteele et al, 1998) and 6.4% in a study of Chinese—Americans in Los Angeles (Reference Zheng, Lin and TakeuchiZheng et al, 1997). In the present study we also observed lower rates of fatigue in Asian countries.
Differences in the prevalence
In order to explain the reported differences in the prevalence of unexplained fatigue we examined whether economic factors at the aggregate level influence rates of fatigue. The classification of centres according to the GNI per head for the country of location showed that subjects coming from middle- or high-income countries were more likely to report substantial unexplained fatigue compared with subjects from low-income countries. Therefore, there is an indication that economic development might influence the reporting of unexplained fatigue. It is worth noting that this pattern was not found when similar analyses were carried out for other unexplained somatic symptoms such as pain, headache, dizziness, excessive flatulence and palpitations (data on file). These symptoms, unlike fatigue, tended to be less prevalent in higher-income countries. Therefore, this finding cannot be attributed to a general tendency for unexplained functional symptoms to be reported more commonly in well-developed countries.
This is a difficult finding to interpret because economic development might be associated with many other (confounding) variables such as organisation of primary health care or local diagnostic preferences. To find out whether a similar pattern is observed for primary care patients, who presented to primary care physicians with complaints of weakness or lethargy (‘presenting fatigue’), we carried out a similar analysis with the presenting complaint of fatigue as the dependent variable. This showed a very different picture; subjects from high- or middle-income countries were less likely to present with fatigue compared with those from low-income countries.
Past research in developed countries, in both the community and primary care, has shown that subjects with fatigue usually attribute their symptoms to psychosocial causes. For example, in a community study in the UK almost half of the subjects attributed fatigue to psychosocial causes such as work, family and lifestyle (Reference Pawlikowska, Chalder and HirschPawlikowska et al, 1994) and similar findings have been reported in primary care (Reference David, Pelosi and McDonaldDavid et al, 1990). If psychosocial explanations are prevalent then it seems reasonable that fatigue will not be a presenting complaint in primary care in developed countries. In that case, fatigue is more of a social ‘metaphor’ rather than a legitimate or useful medical complaint (Reference Lee and WongLee & Wong, 1995). By contrast, in less-developed countries a somatic presentation might ensure an appropriate medical examination. This reminds us of the process of ‘facultative somatisation’ described by Goldberg & Bridges (Reference Goldberg and Bridges1988) where patients present with somatic symptoms as a ‘ticket of admission’ to the primary care clinic. Simon et al (Reference Simon, Von Korff and Piccinelli1999) have reported a similar finding in their study of the relation between somatic symptoms and depression. Analysing the same dataset used in the present study, they found that a somatic presentation of depression was more common at centres where patients lacked an ongoing relationship with a primary care physician. All the centres from low-income countries and most of the centres from middle-income countries were of this type.
Given the strong association of psychiatric disorders with fatigue (Reference Skapinakis, Lewis and MeltzerSkapinakis et al, 2000), we think that this finding might have important clinical implications. In less-developed countries, the complaint of fatigue might be an indicator of hidden psychiatric morbidity. By contrast, in more-developed countries, although syndromes of fatigue are common, they should not be always considered as evidence of unmet need as they might represent a common expression of psychosocial distress.
Clinical Implications and Limitations
CLINICAL IMPLICATIONS
-
▪ Prevalence of unexplained fatigue in primary care differs widely across countries.
-
▪ Patients from well-developed countries are more likely to report fatigue in response to direct questions but are less likely to present with fatigue to physicians compared with patients from less-developed countries.
-
▪ Presenting fatigue in less-developed countries might be an important indicator of hidden psychiatric need.
LIMITATIONS
-
▪ The study was carried out in primary care and results cannot be generalised to the general population.
-
▪ Participation rates for the baseline assessment were relatively low for some centres.
-
▪ Medical causes of fatigue were excluded in a crude way by asking patients if a doctor had given them a definite diagnosis for their symptom or if there had been any abnormalities reported on examination or further investigation.
Acknowledgements
The data reported in this article were collected as part of a World Health Organization's Psychological Problems in General Health Care project. Participating investigators include: O. Ozturk and M. Rezaki (Ankara, Turkey); C. Stefanis and V. Mavreas (Athens, Greece); S. M. Channabasavana and T. G. Sriram (Bangalore, India); H. Helmchen and M. Linden (Berlin, Germany); W. van der Brink and B. Tiemens (Groningen, The Netherlands); M. Olatawura and O. Gureye (Ibadan, Nigeria); O. Benkert and W. Maier (Mainz, Germany); R. Gater and S. Kisely (Manchester, UK); Y. Nakane and S. Michitsuji (Nagasaki, Japan); Y. Lecrubier and P. Boyer (Paris, France); J. Costa e Silva and L. Villano (Rio de Janeiro, Brazil); R. Florenzano and J. Acuna (Santiago, Chile); G. E. Simon and M. von Korff (Seattle, USA); Y. He-Quin and X. Shi Fu (Shanghai, China); and M. Tansella and C. Bellantuono (Verona, Italy). The study advisory group includes J. Costa e Silva, D. P. Goldberg, Y. Lecrubier, M. von Korff and H.-U. Wittchen. Coordinating staff at World Health Organization headquarters include N. Sartorius and T. B. Ustun. This paper was started while P.S. was studying for a PhD degree at the University of Wales College of Medicine. P.S. was funded throughout his studies by the Alexander S. Onassis Public Benefit Foundation.
eLetters
No eLetters have been published for this article.