Article Text

Download PDFPDF

Epidemiology of paediatric chronic fatigue syndrome in Australia
  1. Sarah Knight1,2,3,
  2. Shane Elders1,
  3. Jill Rodda1,
  4. Adrienne Harvey1,
  5. Lionel Lubitz4,
  6. Kathy Rowe5,
  7. Colette Reveley4,
  8. Sabine Hennel6,
  9. Susan Towns7,
  10. Kasia Kozlowska7,
  11. Donald N Payne8,9,
  12. Sonya Marshall-Gradisnik10,
  13. Adam Scheinberg1,2,3
  1. 1 Neurodisability and Rehabilitation, Murdoch Children’s Research Institute, Melbourne, Australia
  2. 2 Victorian Paediatric Rehabilitation Service, Royal Children’s Hospital Melbourne, Melbourne, Victoria, Australia
  3. 3 Paediatrics, The University of Melbourne, Parkville, Victoria, Australia
  4. 4 Royal Children’s Hospital Melbourne, Parkville, Victoria, Australia
  5. 5 General Medicine, Royal Children’s Hospital Melbourne, Parkville, Victoria, Australia
  6. 6 Victorian Paediatric Rehabilitation, Monash Children’s Hospital, Clayton, Victoria, Australia
  7. 7 Sydney Children’s Hospitals Network Randwick and Westmead, Westmead, New South Wales, Australia
  8. 8 Adolescent Medicine, Princess Margaret Hospital, Perth, Western Australia, Australia
  9. 9 SPACH, UWA, Perth, Western Australia, Australia
  10. 10 Griffith University—Gold Coast Campus, Southport, Queensland, Australia
  1. Correspondence to Dr Sarah Knight, Neurodisability and Rehabilitation, Murdoch Children’s Research Institute, Melbourne 3052, Australia; sarah.knight{at}


Objective To estimate the paediatrician-diagnosed incidence of chronic fatigue syndrome (CFS) in Australia, and describe demographic and clinical features, as well as approaches to diagnosis and management.

Methods The Australian Paediatric Surveillance Unit facilitates monthly national surveillance of uncommon conditions seen by paediatricians. Data from young people aged <18 years diagnosed with CFS were collected. Incidence was estimated based on new cases reported from April 2015 to April 2016.

Results A total of 164 cases of newly diagnosed CFS in young people aged 4–17 years were identified for inclusion. The estimated national incidence for children aged 4–9 years was 0.25 per 100 000 per annum. In children aged 10–17 years, the estimated incidence of paediatrician-diagnosed cases for Victoria (17.48 per 100 000) was substantially greater than other Australian states (range 1.31–5.51 per 100 000). Most cases were female and Caucasian, most commonly presenting after an infectious illness with symptoms gradual in onset. The majority were diagnosed at least 13 months after symptom onset. Symptoms, associations, investigations and management strategies were highly variable.

Conclusions Current findings suggest that, consistent with other countries, the Australian incidence of CFS in children aged <10 years is very low. In contrast, the national incidence of CFS in older children and adolescents (aged 10–17 years) is more unclear, with marked variability between geographical regions apparent. This may be due to variation in service accessibility and clinician understanding of CFS. Accordingly, national initiatives to improve equity of care for children with CFS may be required.

  • chronic fatigue syndrome
  • epidemiology

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

What is already known on this topic?

  • Paediatric chronic fatigue syndrome (CFS) causes significant disability.

  • Estimates of the incidence of paediatric CFS vary.

  • We know little about how common the condition is or how it is currently managed in Australia.

What this study adds?

  • CFS is uncommon in children aged <10 years.

  • The Australian incidence of CFS in older children and adolescents (aged 10–17 years) is unclear, with marked variability between geographical regions apparent.

Chronic fatigue syndrome (CFS) is a condition of unknown aetiology characterised by severe, persistent, unexplained fatigue. This fatigue may be accompanied by other symptoms including postexertional malaise, pain, cognitive deficits and disrupted sleep.1 In children and adolescents, CFS can lead to significant functional impairment, including a substantial reduction in school attendance.2

As the aetiology of CFS is unknown and there is no diagnostic test for CFS,3 it is diagnosed clinically through the use of case definitions.4 Several case definitions and terminologies for CFS have been used internationally in clinical practice and research since the first case definition was published in 1988.5

Epidemiological estimates of prevalence for CFS vary significantly. A meta-analysis of adult studies found estimates of prevalence varying between 0.2% and 6.41%, with a pooled prevalence of 3.48% for self-reported cases, and 0.76% for clinically assessed cases.6 There is less literature specifically examining the epidemiology of paediatric CFS, with varying prevalence estimates ranging from 0% to 2.91%.7–18 Four published studies, from the Netherlands, Norway and the UK, report variable incidence rates of paediatric CFS with estimates ranging from 12 (0.012%) to 500 (0.5%) per 100 000.7 8 19 20 Some of the variation in incidence estimates may be explained by variation in diagnostic methods, case ascertainment and age range captured.

The primary aim of this study was to provide an estimate of the paediatrician-diagnosed incidence of CFS in Australian young people, based on prospective national reporting by paediatricians. Secondary aims were to describe demographics, symptomatology, diagnosis and management of CFS, based on reported cases.

Patients and methods

The Australian Paediatric Surveillance Unit (APSU) facilitated national surveillance of CFS in young people aged under 18 years from April 2015 to April 2016. The APSU undertakes national surveillance of rare paediatric conditions in Australia.21 Each month clinicians on the APSU contact database are emailed a report card listing conditions currently being studied by the APSU. Clinicians are asked to report children newly diagnosed. In April 2015–April 2016, the total number of paediatricians reporting to the APSU was 1467 and the overall monthly report card return rate was 93%. According to the Australia Government Department of Health, there were 2059 paediatricians employed in Australia in 201522; therefore, approximately 71% of Australian paediatricians reported to the APSU at the time of the study.

Paediatricians received a protocol which included CFS case definition (online supplement 1). Reportable cases were defined as young people aged under 18 years, who were newly diagnosed with CFS in the past month. The upper age limit was defined by the APSU. CFS was defined according to the Centers for Disease Control and Prevention Case Definition (modified for a paediatric population under the auspices of the Royal Australasian College of Physicians).23 For every newly reported case, paediatricians completed a 20-item questionnaire which included: date of birth, sex, postcode, ethnicity, presenting symptoms including duration, type of onset, severity of and triggers for symptoms, comorbid conditions, family history, investigations undertaken and advised management.

Data analysis

Data were analysed using Stata V.14.1 (StataCorp, 2016). Population data from the Australian Bureau of Statistics were used to calculate estimates of incidence of CFS. Values were calculated as the mathematical average of the 2015 and 2016 population estimates to coincide with the current study data collection period. Poisson regression was used to determine estimated incidences and were presented as rates per 100 000 persons per annum, with 95% CIs. National level incidence estimates for all children aged 4–9 years, all adolescents aged 10–17 years, female adolescents aged 10–17 years and male adolescents aged 10–17 years, were calculated. Age ranges were determined based on previous epidemiological studies in the paediatric CFS literature to allow for comparison of results. State-specific and gender-specific incidences were not calculated for the 4–9 years due to the low number of reports. In two cases, paediatricians failed to report the sex of the young person. In order to prevent under-reporting of male and female incidences, the female incidence assumed these two individuals were female, while the male incidence assumed they were male.

The characteristics of CFS were presented as frequencies and percentages of the total sample of reported cases. Missing data from the questionnaire were categorised in the results as being ‘not specified’. Chi-square analyses were used to compare demographics and clinical characteristics of Victorian patients with those in other states and territories. Due to multiple comparison, p<0.01 was considered to be statistically significant.


Incidence estimates

Between April 2015 and April 2016, 191 cases were reported to the APSU. Questionnaires were returned for 184 (96%) cases, and 164 were confirmed as CFS cases. Of the remaining cases for which questionnaires were returned, 10 were duplicates and 10 were classified as errors (administrative errors or cases outside the case definition or age range). Ages ranged from 4 to 17 years for the 164 young people with confirmed CFS. Five confirmed cases were children aged 4–9 years, while 159 were adolescents aged 10–17 years.

The estimated national incidence of paediatrician-diagnosed cases of CFS was 0.25 per 100 000 per annum (95% CI 0.08 to 0.59) for children aged 4–9 years, and 6.38 per 100 000 per annum (95% CI 5.43 to 7.45) for adolescents aged 10–17 years. Estimated national incidence for adolescent males was 3.68 per 100 000 per annum (95% CI 2.70 to 4.89), and 9.39 per 100 000 per annum (95% CI 7.75 to 11.28) for adolescent females. There was mild variation in the estimated incidence rates of paediatrician-diagnosed cases of CFS for adolescents according to state or territory of Australia, but incidence rate of CFS in adolescents for the state of Victoria (17.48 per 100 000) was greater when compared with other Australia states (range 1.31–5.51 per 100 000) (table 1). During the study period, the rates of reporting of CFS were disproportionally higher in some states and unexpectedly low in others compared with non-CFS conditions being reported to the APSU (online supplement 2).

Table 1

Estimated total, female and male incidence rates (95% CI) of paediatrician-diagnosed cases of CFS according to region in adolescents aged 10–17 years


Average age of included cases was 14.9 years (95% CI 14.6 to 15.2) (table 2). Females constituted 68.9% of cases (female-to-male ratio=2.3:1). Majority of cases (95.1%) were Caucasian. Patient demographics did not differ between regions; therefore, national data are reported.

Table 2

Demographics for included cases


There were no differences in symptomatology between Victorian patients and those in other states; therefore, national data reported (table 3). The most common pattern for onset of symptoms was gradual (56% cases), while 37% were sudden. Fifty-two per cent of diagnoses were made at least 13 months after symptom onset, while 29% were diagnosed after 24 months. For approximately two-thirds of patients, an infectious illness was reported as the trigger for symptom onset. On average, individuals had 9.9 symptoms (95% CI 9.2 to 10.4) over the course of their illness. The seven most common symptoms, being fatigue, postexertional malaise, sleep disturbance/unrefreshing sleep, headache, attention/concentration difficulties, light-headedness/dizziness and musculoskeletal pain, were present in more than two-thirds of patients. Symptom severity was reported as moderate for 63.4% of cases.

Table 3

Medical characteristics of included cases

The most commonly associated medical condition was postural orthostatic tachycardia syndrome (POTS), observed in 26.2% of patients (table 4). Joint hypermobility, fibromyalgia/chronic widespread pain and migraines were present in over 10% of cases. No medical comorbidities were reported for 41% of patients and most patients (59.8%) did not have any reported comorbid psychiatric conditions. Paediatricians reported that anxiety and depression (26% and 13% of cases, respectively) were the most common psychiatric comorbidities. In terms of self-reported family history, patients were reported to have a first-degree relative with anxiety disorder (18%), depression (17%) or CFS (13%), respectively. There was no relevant family history for 50% of patients.

Table 4

Comorbidities and family history of included cases


Clinicians requested a variety of investigations (online supplement 3). As there were no major observed differences in requested investigations according to state, total results are presented. Eight tests were ordered for >50% of cases: full blood count and differential urea, electrolytes and creatinine, thyroid function tests, erythrocyte sedimentation rate, liver function tests, Epstein-Barr virus (EBV) serology, C reactive protein and coeliac screen. The investigations that had the highest proportion of abnormal results were EBV serology (34%), allergy tests (33%), cytomegalovirus serology (25%), serum vitamin D (24%) and antinuclear antibodies (ANA) (23%).


Most commonly used services were teacher and school-based services (42%) and specialised CFS services (40%) (online supplemen 4). Psychology (27%) and physiotherapy services (20%) were generally included. Paediatricians reported that alternative and complementary medicine services were not often used by patients. Victorian patients were significantly more likely to be seen in a specialist CFS service compared with other states (Victoria: 53%, other states/territories: 16%, χ2=20.67, p<0.001), while patients in other states and territories were significantly more likely to be seen in adolescent medicine (Victoria: 5%, other states/territories: 36%, χ2=25.14, p<0.001), use physiotherapy (Victoria: 8%, other states/territories: 43%, χ2=27.35, p<0.001), occupational therapy (Victoria: 6%, other states/territories: 38%, χ2=27.36, p<0.001) or rheumatology (Victoria: 7%, other states/territories: 32%, χ2=16.9, p<0.001). There were no other significant state-based differences in services used by patients.

Sleep hygiene was a treatment strategy recommended to almost all patients (95%), with graded exercise therapy, a modified school programme/home tutoring, pacing/balancing activities or symptom management with medication/supplementation frequently recommended. There were no significant differences in treatment strategies used for Victorian patients compared with those in other states, with the exception of cognitive behavioural therapy, which was reported to be less likely to be recommended in Victoria than in other states (Victoria: 19%, other states: 71%, χ2=44.50, p<0.001).


This study is the first to provide estimates of paediatrician-reported incidence of CFS in the Australian paediatric population, and the second worldwide to collect prospective incidence data.8 The estimated incidence in children aged 4–9 years was 0.25 per 100 000 per annum (95% CI 0.08 to 0.59), while the estimated incidence in adolescents aged 10–17 years was 6.38 per 100 000 per annum (95% CI 5.43 to 7.45). The reported incidence for the state of Victoria (17.48 per 100 000) was substantially higher compared with other Australian states and territories (1.31–5.51 per 100 000).

There has been very little data available to quantify the clinical impression that CFS is rarely diagnosed in children aged under 10 years. Our results support this by demonstrating that CFS is uncommonly diagnosed in Australian children aged under 10 years. The only other study to document the incidence of CFS in younger children also reported this age range to have the lowest incidence.20

Australian estimates of incidence of CFS in the current study is lower than estimates from other countries in the current literature. The estimated Victorian incidence is similar to that in other countries,8 20 suggesting incidence rates in Victoria could be more reflective of the true incidence of CFS. The underlying reasons for the geographic variation of diagnostic rates in Australia are unclear, but likely multiple. While the contribution of true geographical or biological factors in the observed incidence rates between regions cannot be ruled out based on the results of the current study, we believe that the differences between Victoria and the other states is most likely attributable to local phenomena affecting case ascertainment. The geographical reporting pattern of CFS in this study is not consistent with the overall pattern of APSU reporting rates across states, suggesting that results cannot be explained by the current study methodology. Contextual factors causing variation in diagnosis could relate to differences in access to specialists and/or diagnostic practices/culture (eg, lack of recognition or misconceptions about CFS). Further research is required to understand the influence of service provision on the epidemiology of this condition. For example, from the current study we cannot determine whether the incidence is greater in Victoria due to increased access to CFS specialist services or whether there is greater access to services due to increased need. Furthermore, it is unclear whether Victorian paediatricians are more likely to make a diagnosis of CFS as there is a dedicated service to refer to where access requires a formal diagnosis of CFS or whether greater access to specialist services in certain areas influences physician awareness and education of CFS, resulting in increased diagnostic rates.

Our results regarding gender ratio, ethnicity and symptom frequency are comparable to that in the international literature.8 15 24–30 Every study that reports ethnicity data, including our own, occurred in high-income countries with Caucasian majority populations. It is possible that the high proportion of Caucasian cases of CFS found in the literature reflects this ethnic group’s majority in the population, rather than an increased susceptibility compared with other ethnicities.

Consistent with current results, patients most commonly have gradual onset of symptoms, onset linked to infectious illness and moderate to high severity.2 8 9 29 30 As in this study, existing literature has demonstrated associations between CFS and POTS, joint hypermobility, fibromyalgia, anxiety and depression.31–35 Although the rates of a family history of depression, anxiety and CFS in the current study need to be interpreted cautiously as they are based on self-report from the family or young person to the paediatrician, other research has suggested a likely association with depression, anxiety and CFS in first-degree relatives.36 37

The protracted time between symptom onset and diagnosis observed in our study has also been observed in two Australian studies.30 38 This delay is substantial, with 54% of patients not receiving a diagnosis, and therefore likely not having access to appropriate treatment, for over a year. Investigations recommended by the Australian clinical practice guidelines23 yielded very few abnormal results. This is consistent with the current practice of using diagnostic tests to rule out other medical conditions prior to diagnosis of CFS. Tests that most commonly yielded abnormal results, such as EBV serology and the ANA test, are not recommended by the guidelines, yet were requested for many patients. It is possible that because the existing CFS guidelines are based largely on adult care, these tests were not included as they are more relevant to a paediatric population.

Overall, the demographic and clinical characteristics, and approaches to diagnosis and management, do not vary greatly between Victoria and other states. Victorian patients were more likely to be referred to a specialist CFS service, while patients elsewhere were more likely to be seen in adolescent medicine and referred to individual allied health professionals and rheumatology. This difference likely reflects regional differences in service models and approaches to CFS management.

Strengths and limitations

This study used for case ascertainment, an established national surveillance system, which is independent of the research team. This reduces the potential for observer bias. All diagnoses were made by medical practitioners following a consultation and physical examination, which is likely to improve the validity of each diagnosis.

Although this is the most complete national dataset currently available for paediatric CFS in Australia, the study relies on self-reporting by paediatricians and some paediatricians may not report cases to the APSU. Paediatricians who are not confident in diagnosing CFS, do not believe in the existence of CFS, or the importance a diagnosis and providing management, may not report cases even though the child or adolescent meets the criteria. Accordingly, our results may underestimate the true incidence of paediatric CFS in Australia. It was not possible to include general practitioners in the data collection which may also result in an underestimate of incidence. It is possible that milder cases of CFS are not referred to paediatricians, and so were not identified in our study. The low proportion of severe and very severe cases may be influenced by selection bias, as these young people may be too sick to attend a consultation for diagnosis and so not captured by our study.

The age ranges used in the current study were informed by previous epidemiological research in paediatric CFS and were confined by the surveillance methodology used. Future research studies should consider using narrower age bands to further understand the incidence of CFS throughout childhood, particularly during the adolescent period.


This study suggests that there is substantial variation in diagnostic rates of CFS between Victoria and other states, by Australian paediatricians. The variation in diagnostic rates of CFS by paediatricians, plus protracted time to diagnosis, highlights the need for a national, coordinated approach to improving healthcare for children and adolescents with CFS. This should include prospective evaluation of health outcomes for adolescents with CFS and supporting an evidence-informed approach to management. Review and update of the national clinical practice guideline for CFS, with a systematic implementation strategy, will be crucial to ensuring children and adolescents with CFS receive access to more consistent care at national level.


The authors would like to thank Dr David Burgner, Murdoch Children’s Research Institute, who provided advice in the interpretation of the study results. The authors would also like to thank the staff at the APSU Amy Phu and Marie Deverell for facilitation of reporting and management of the APSU database.



  • Contributors SK and AS conceptualised and designed the study, coordinated and supervised data collection and analysis, drafted the initial manuscript and reviewed and revised the manuscript. KR, CR, ST, KK, DNP, LL and SH designed the data collection instrument and reviewed and revised the manuscript. SE and JR carried out the initial analyses, drafted the initial manuscript and reviewed and revised the manuscript. AH conceptualised and designed the study, and critically reviewed the manuscript for important intellectual content. All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

  • Funding This study was funded by the Mason Foundation and the Murdoch Children’s Research Institute Government Infrastructure Grant.

  • Competing interests None declared.

  • Ethics approval This study received ethics approval from The Sydney Children’s Hospital Network Human Research Ethics Committee (LNR/14/SCHN/399).

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement Any unpublished data from this study are available on request by emailing the corresponding author.

  • Patient consent for publication Not required.

Linked Articles

  • Atoms
    Nick Brown