OBJECTIVES: 1) Estimate the prevalence of chronic disabling fatigue at age 13, and 2) investigate associations with early family adversity, using a population-based birth cohort study.
METHODS: The subjects were 10 001 children, of whom 5657 had sufficient data to define chronic disabling fatigue at age 13. Chronic disabling fatigue was defined as tiredness reported by mothers that had lasted for ≥3 months or ≥6 months, was disabling (school absence or prevented participation in hobbies/sport/leisure activities “quite a lot” or a “great deal”), and not due to another cause. Family adversity was defined by using 14 questions addressing housing, education, social relationships, and maternal health assessed prospectively at birth. We used multiple imputation to correct estimates of prevalence and association for loss to follow-up.
RESULTS: Overall, 1995 teenagers had been tired for the last month, of whom 117 and 53 had chronic disabling fatigue of ≥3 and ≥6 months’ duration, respectively. The estimated prevalence of chronic disabling fatigue of ≥3 and ≥6 months’ duration was 2.07% and 0.94%, respectively, but increased to 2.44% and 1.30%, respectively, after imputing missing data. Prevalence was similar in boys and girls. The risk of chronic disabling fatigue increased with family adversity score (odds ratios: 1.25 [95% confidence interval: 1.09, 1.43] and 1.23 [1.01, 1.51] for ≥3 and ≥6 months’ duration, respectively). Only 36 (30.77%) children with chronic disabling fatigue had consulted a doctor about their fatigue.
CONCLUSIONS: Chronic disabling fatigue is more common in families who experienced early family adversity and is often not reported to health care services.
- CDC —
- Centers for Disease Control and Prevention
- CFS —
- chronic fatigue syndrome
- CI —
- confidence interval
- FAI —
- Family Adversity Index
- GP —
- general practitioner
- ME —
- myalgic encephalopathy
- NICE —
- National Institutes of Health and Clinical Excellence
- OR —
- odds ratio
What’s Known on This Subject:
Chronic fatigue syndrome (CFS) is a disabling illness. Some children with chronic disabling fatigue are housebound. Some primary care physicians find making the diagnosis of CFS difficult.
What This Study Adds:
Chronic disabling fatigue is more common in 13-year-olds than previously realized with a population prevalence of ∼2.0%. Chronic disabling fatigue in teenagers is more common in those with higher levels of early family adversity.
Fatigue is common in teenagers, with a reported prevalence of 8% to 38%.1–4 The prevalence of chronic disabling fatigue is less clear: estimated prevalence varies between 0.05% and 2.34%2,5–7 depending on the definition used and the method of ascertainment. Most studies have used the Centers for Disease Control and Prevention (CDC) criteria for chronic fatigue syndrome or myalgic encephalopathy (CFS/ME), which require 6 months of disabling fatigue and 4 out of 8 possible additional symptoms.8 Others use 3 months of disabling fatigue, particularly in the United Kingdom where the National Institutes of Health and Clinical Excellence (NICE) guidelines recommend that a diagnosis of CFS/ME is given after 3 months of disabling fatigue.9 Disabling fatigue is defined as fatigue that “results in substantial reduction in previous levels of occupational, educational, social, or personal activities”8 and “has resulted in a substantial reduction in activity level”.9
Socioeconomic status is an important risk factor for CFS/ME in adults.10 If socioeconomic status is also a risk factor in children, then questionnaire or telephone based surveys may be biased, because lower socioeconomic status is also a risk factor for nonresponse. An association between socioeconomic status and CFS/ME may explain why studies using general practitioners (GPs) or physicians to report caseness11–14 have revealed lower prevalence rates than population-based studies,2,5–7 because this may be due to poor health seeking behavior as well as a lack of confidence in diagnosis.15 Early childhood adversity, such as abuse, has been shown to be a risk factor for developing adult CFS/ME in some retrospective studies16–18 and one longitudinal study.19
We used the Avon Longitudinal Study of Parents and Children (ALSPAC), a large population-based cohort study, to study both the prevalence of pediatric chronic disabling fatigue in 13 year olds and whether early family adversity (combining early childhood adversity and socioeconomic status) is a risk factor for chronic disabling fatigue at age 13 years. Because children were not examined by a physician, we use the term “chronic disabling fatigue” rather than CFS/ME to refer to disabling fatigue without another cause that was identifiable from cohort data.
The ALSPAC is a population-based longitudinal birth cohort of children who had an expected date of delivery between April 1991 and December 1992 and whose mothers were resident in the Avon region of southwest England at the time of recruitment.20 Over 14 000 expectant mothers were recruited into the study, and the children have been followed up regularly since birth with postal questionnaires for both children and their parents, clinical assessments, and the collection of biological samples (www.alspac.bris.ac.uk). Overall, 7114 mothers returned inventories when their child was 13 years old. The primary source of data collection has been via self-completion questionnaire.
When the study children were aged 13 years, their mothers or main carers were sent a questionnaire that included questions on whether their teenager had been feeling tired or lacking in energy over the last month (yes, no); how long the tiredness had lasted (<3 months, 3 to 5 months, 6 months to 5 years, ≥5 years); how many days their teenager had missed school because of tiredness; and whether the tiredness/lack of energy had stopped the teenager taking part in hobbies, sport, or leisure activities (not at all; only a little; quite a lot; a great deal). The questionnaires asked whether the teenager snored (never, sometimes, often); whether the mother thought the fatigue was due to the teenager playing too much sport; and whether the teenager took regular medication (free text question: “Please indicate below any medicines your child has used in the last 12 months”). At the same time, teenagers completed the Short Moods and Feelings Questionnaire21 to screen for self-reported depressive symptoms that might indicate low mood.
Mothers were sent postal questionnaires about themselves at 4 time points during pregnancy and then at specific time points after the birth of the child. The Family Adversity Index (FAI) was derived from responses to questions about early childhood adversity and socioeconomic status, which the mother completed during pregnancy about the following family based factors: age of mother, housing adequacy, basic amenities within the home, housing defects/infestations, mother’s educational qualifications, financial difficulties, partner relationship status, partner affection, partner aggression (physical–emotional cruelty), family size, family problems such as child in care or not with natural mother, or child on at risk register, social network including emotional and practical/financial support, maternal psychopathology including depression, anxiety, or suicide attempts, substance abuse including drugs or alcohol, and crime including being in trouble with police or actual convictions (see Appendix for more details). Binary Family Adversity factors were generated from the multiple questions for each FAI factor. Four factors with low prevalence were excluded because they caused perfect prediction problems in analyses. Percentages were then calculated based on these binary factors. Teenagers were grouped according to whether they had 0, 1, 2, 3, or 4 or more family adversity factors (FAI score).
Definition of Chronic Disabling Fatigue
Based on the responses to the questionnaires sent at age 13, we identified teenagers reported to have had fatigue lasting more than 3 months, or more than 6 months, that was associated with absence from full-time school or that had stopped them from taking part in activities “quite a lot” or “a great deal.” We defined these teenagers as having chronic disabling fatigue, after excluding those who (1) had a maternal report that their fatigue was caused by playing too much sport; (2) had probable comorbid depression defined as a Short Moods and Feelings Questionnaire score more than 10; (3) snored often; or (4) had other illnesses that could cause fatigue (for example, cancer or juvenile idiopathic arthritis), based on self-reported medication use. Teenagers with missing values for snoring or medication use were assumed not to snore often or not to use medication. Teenagers were assigned a missing value for chronic disabling fatigue if the question for depression or school attendance had not been answered. We did not require an answer to the initial screening question for fatigue if the response to the subsequent question on timing of fatigue demonstrated the teenager to have had fatigue for more than 3 months.
We calculated the prevalence of chronic disabling fatigue among those teenagers for whom sufficient data were available to define its presence or absence. Ignoring those teenagers with missing values for chronic disabling fatigue could result in bias and/or inflation of SEs. If missingness is dependent only on observed data (referred to as missing at random),22 then multiple imputation can be used to correct this bias and produce more efficient estimates.
We generated 50 imputed data sets, and combined estimates by using Rubin’s rules.22 We multiply imputed the variables for fatigue and chronic disabling fatigue separately because they were derived by using the same questionnaire information, and if a teenager had a missing value for 1 fatigue variable then in almost all cases they were also missing values for the other fatigue variables. Imputations were carried out separately in boys and girls because female gender is a risk factor for chronic disabling fatigue in adults.23 Each variable was imputed based both on the FAI score and on the individual FAI factors, using univariate chained equations method. We performed imputation diagnostics by calculating, for each set of imputed data sets, the percentage difference between the prevalence based on the observed data and the prevalence based on the imputed data in the first 10 imputed data sets. We examined the association of FAI score (as a continuous explanatory variable) with fatigue and chronic disabling fatigue by using logistic regression. All analyses were done by using Stata version 11.
Ethical approval for the study was obtained from the ALSPAC Law and Ethics Committee and the Local Research Ethics Committees.
Of 13 988 children alive at 1 year, 10 001 children had complete data on gender and the FAI and were included in analyses. Of these, 5721 mothers or main carers provided usable information on the presence or absence of fatigue at age 13 years. Figure 1 reveals how answers to different questions on fatigue were used to define chronic disabling fatigue at age 13. Of the 5721 who provided useable information on fatigue, 2059 (35.99%) had been fatigued for the past month. We were unable to define chronic disabling fatigue in 64 respondents because of missing information on fatigue duration (26), whether the fatigue was disabling (6) or related to depression (32). Table 1 summarizes the characteristics of those teenagers whose mothers did and did not provide usable information. Mothers who did not complete questionnaires at 13 years of age were more likely to be from homes with evidence of early family adversity.
Of the 5657 13 year olds with sufficient information to define chronic disabling fatigue, 1995 were tired or lacking in energy in the last month. Of these, the following did not have chronic disabling fatigue: 1284 who were tired for <3 months; 516 who had fatigue but were not disabled by it; 18 whose mothers thought they were fatigued due to playing too much sport; 27 who snored frequently (to exclude sleep apnea); and 33 who had probable depression. There were no teenagers on medication for long-term illnesses that could have caused the fatigue. There were 296 children with missing data on snoring, and it was assumed they did not snore frequently.
Of 117 13 year olds with chronic disabling fatigue of at least 3 months’ duration, 53 (45.30%) had been affected for more than 6 months and 7 for more than 5 years. The numbers of girls were 58 (49.57%) and 26 (49.06%) for fatigue of ≥3 and ≥6 months’ duration, respectively. Only 11 of the 117 had not missed any school but had been prevented from taking part in hobbies, sport, or leisure activities either “quite a lot” or “a great deal” because of their fatigue: 5 (4.27%) were unable to attend school at all and would be defined as severely affected by using the NICE guideline definition.9 One of these housebound teenagers had been unwell for <6 months, whereas 4 had been unwell for between 6 months and 5 years. Only 36 (30.77%) had been to see their doctor because of fatigue.
Prevalence of Disabling Fatigue
Table 2 reveals estimates of the prevalence of disabling fatigue of at least 3 and 6 months’ duration. In children with complete data, the estimated prevalence of chronic disabling fatigue ≥3 months was 2.07% (95% confidence interval [CI]: 1.70, 2.44); this increased to 2.21% (1.82, 2.60) and 2.44% (1.91, 2.97) after allowing for loss to follow-up by using multiple imputation based on the FAI score and individual FAI factors, respectively. The estimated prevalence of disabling fatigue ≥6 months was 0.94% (95% CI: 0.69, 1.19) based on children with complete data and was 1.00% (0.76, 1.25) and 1.30% (0.84,1.76) after allowing for loss to follow-up by using multiple imputation based on the FAI score and individual FAI factors, respectively. Prevalence was similar in boys and girls.
Association of FAI With Fatigue
Table 3 reveals the prevalence of family adversity factors in teenagers with and without fatigue in the last month and with and without chronic disabling fatigue of at least 3 or 6 months’ duration. The strongest association was seen between maternal psychopathology (high level of anxiety, depression, or attempted suicide) in pregnancy and fatigue or chronic disabling fatigue at 13. The association with early family adversity was stronger for chronic disabling fatigue (odds ratios [ORs] per FAI: 1.25 [95% CI: 1.09, 1.43] and 1.23 [1.01, 1.51] for duration 3 and 6 months, respectively) than for fatigue lasting 1 month (OR per FAI: 1.10 [95% CI: 1.05, 1.15]). Results based on multiple imputation were similar (ORs: 1.24 [1.08, 1.43] and 1.23 [1.00, 1.51], respectively).
Figure 2 reveals that there was an approximately linear increase in the log odds of developing chronic disabling fatigue lasting 3 months and 6 months with increasing number of family adversity factors. The odds of developing chronic disabling fatigue were 2.54 (95% CI: 1.35, 4.78) and 2.56 (1.01, 6.45) times greater for duration 3 and 6 months, respectively, in children from families with 4 or more factors compared with those from families with no factors.
In this population-based cohort study, the estimated prevalence of chronic disabling fatigue of duration more than 3 months was 2.21% (95% CI: 1.82, 2.60) after allowing for loss to follow-up by using multiple imputation based on an FAI score, rising to 2.44% (1.91, 2.97) based on imputation by using individual FAI factors. Corresponding estimates of the prevalence of fatigue of duration more than 6 months were 1.00% (0.76, 1.25) and 1.30% (0.84, 1.76), respectively. Only 30.7% of children with chronic disabling fatigue of more than 3 months’ duration had been to see their doctor because of their fatigue, whereas 4.3% were unable to attend any school. Prevalence was similar in boys and girls. The odds of developing chronic disabling fatigue lasting more than 3 months or more than 6 months, at age 13 years, increased with increasing family adversity and were 2.5 and 2.6 times greater, respectively, in children from families with 4 or more FAI factors compared with those from families with no factors.
Strengths and Limitations
This is, to our knowledge, the largest population-based study investigating chronic disabling fatigue in teenagers. We used both 3 months and 6 months’ fatigue duration to define those with chronic disabling fatigue, enabling us to estimate prevalence by using durations that are comparable with standard definitions of CFS/ME in the United States and the United Kingdom.8,9 Because mothers answered multiple questions on symptoms and medication, and children answered questions on mood disorders, we were able to exclude teenagers with treated medical conditions and probable mood disorders. We used multiple imputation to correct for potential bias in prevalence estimates because of loss to follow-up. Bias correction appeared greatest when imputation was based on individual FAI factors rather than the FAI score, which is likely attributable to improved identification of subgroups with higher prevalence in whom dropout was more likely. However, between-imputation variance was greater when imputations were based on individual factors because of the small number of children exposed to some FAI factors. This increased variability was reflected in wider CIs for imputations based on individual factors.
The main limitation of this study is that definitions of chronic disabling fatigue were based on parental reports of symptoms. The CDC8 criteria require that 4 of 8 additional symptoms are present, whereas the NICE9 criteria require 1 additional symptom. Each set of criteria require physician diagnosis to exclude other conditions that cause fatigue. Nonetheless, the data available to us are likely to have allowed us to exclude such conditions: for example, children with cancer or arthritis causing fatigue would have been on medication and would have been excluded unless parents did not list regular medication used in the free text answer. We excluded all teenagers with probable depression, though some teenagers with depression might not have been identifiable by using cohort data. On the other hand, depression could be secondary to chronic disabling fatigue,24 leading to exclusion of children with such fatigue and hence underestimation of prevalence.
There were limited parental data on whether the fatigue was disabling, and we only included teenagers whose parents reported that the fatigue had either stopped them going to school or stopped them taking part in activities “quite a lot” or “a great deal.” We may have excluded teenagers whose parents were not aware of the impact of fatigue or included teenagers where parents have overestimated the symptoms of fatigue or its impact.
As in any observational analysis, our results might be explained by confounding factors that influence early life family adversity and are prognostic for chronic fatigue in adolescence. The influence of such factors would have to occur during or before pregnancy, and we were not able to identify such factors in the data available to us. It is likely that families who experienced early family life adversity also experienced more adversity later in life. The role of factors measured during mid childhood or adolescence in explaining (mediating) the role of early life adversity is of interest and will be addressed in future work. Further studies are needed to see if our results can be replicated before any associations described can be assumed to be causal.
It is not clear whether particular questions within the FAI were more relevant than others to the development of chronic disabling fatigue later in life. Because the number of positive responses to individual FAI questions was relatively small, our study has insufficient power for a comprehensive exploration of this issue. Many of the topics addressed in the FAI were sensitive (for example, partner aggression) and so they may have been underreported. It is possible that mothers experiencing early family adversity might be more likely to report lack of energy or fatigue in their children so that the maternal answers at 13 were biased, but we know of no evidence of a link between family adversity and such a reporting bias.
Comparison With the Literature
To our knowledge, this is the first study to reveal an association of early life family adversity with chronic disabling fatigue in teenagers. Our results are consistent with studies in adults, which describe an inverse association between CFS/ME and socioeconomic status.23 Our results are also consistent with recent data on adult CFS/ME that describe an association between adult CFS/ME and early childhood adversity. Childhood trauma (physical, emotional, and sexual abuse, or physical or emotional neglect) identified retrospectively by questionnaire has been reported to be associated with a 5.6-fold increased risk of CFS/ME.17 That study suggested that this association could be mediated through neuro-endocrine dysfunction because only those with CFS/ME and exposure to childhood trauma had decreased mean early morning cortisol levels.17 Alternatively, the association we describe could be due to environmental effects exerted on the fetus during pregnancy, which would be consistent with the large body of evidence revealing that fetal environmental is associated with disease in adults.25
Our results for the prevalence of fatigue lasting at least 1 month (36%) are consistent with previous reports that describe the prevalence as 8% to 38% in the United States, United Kingdom, and Holland.2–4,26,27 However, our estimates of the prevalence of chronic disabling fatigue are higher than those of CFS/ME reported from previous questionnaire- or telephone-based surveys, in which the estimated prevalence of fatigue lasting 6 months (using the CDC definition) was between 0.05% and 0.57%.2,5,7,11,12 As noted above, the additional symptoms required by the CDC definition may not have been present in all the teenagers identified in our study as having disabling fatigue of 6 months duration. Our results are consistent with 1 previous study where the estimated prevalence disabling fatigue of estimates lasting 3 and 6 months was 2.34% and 1.29% respectively, using a twin cohort.28
In our study, only 30.7% of teenagers with chronic disabling fatigue had been to see their GP because of their fatigue. This may imply that prevalence estimates from studies relying on medical surveillance may be too low.13,14 It has been reported that only 48% of GPs feel confident in making a diagnosis of CFS/ME in adults.15
We were surprised that the estimated prevalence of chronic disabling fatigue was similar in girls and boys. A high girl to boy ratio for both fatigue lasting 1 month and CFS/ME has been reported in previous studies.2,7,12,26,28 It is possible the gender ratio changes as teenagers get older: the prevalence of CFS/ME appears to increase with age in girl but not boy teenagers.3,4
Chronic disabling fatigue in teenagers appears to be more common than described in reports based on medical surveillance, which may be because only a minority of affected teenagers are diagnosed or offered treatment or could alternatively be because not all teenagers with chronic disabling fatigue meet standard definitions of CFS/ME. Pediatricians should consider the possibility of chronic disabling fatigue in children who are not attending school full-time, particularly in those from disadvantaged backgrounds because teenagers from families that experience early family adversity are more vulnerable and yet potentially less likely to access the help they need. Awareness needs to be raised to ensure that families of affected teenagers access medical care and that pediatricians and those looking after children are trained in the identification and management of chronic disabling fatigue. Future research should examine the type of fatigue experienced by teenagers and its different phenotypes and investigate important potential etiologic factors that might explain the association of fatigue with family adversity.
APPENDIX The ALSPAC FAI in Pregnancy
We thank all the families who took part in this study, the midwives for their help in recruiting them, and the whole ALSPAC team, which includes interviewers, computer and laboratory technicians, clerical workers, research scientists, volunteers, managers, receptionists, and nurses.
- Accepted March 12, 2012.
- Address correspondence to Esther Crawley, BA(Hons), BM BCh, PhD, FRCPCH, Centre for Child and Adolescent Health, School of Social and Community Medicine, University of Bristol, Oakfield House, Oakfield Grove, Clifton, Bristol BS8 2BN, United Kingdom. E-mail:
The concept of this article belongs to Dr Crawley, who also worked on the analyses, interpreted the data, and drafted the article. Ms Hughes did the imputation analyses and contributed to the analyses of prevalence; Dr Northstone retrieved the data and contributed to the identification of appropriate variables; Dr Tilling supervised the imputation analyses; Dr Emond provided help with the interpretation of the results; Dr Sterne designed the analyses and contributed substantially to the revision of this article; and all authors contributed to revisions of this article and have approved the final version to be published.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: Dr Crawley is funded by a National Institute for Health Research Clinician Scientist Fellowship (award number NIHRCS/01/2008/006). Ms Hughes is funded by Medical Research Council grant G0900724. Dr Northstone receives funding from a variety of sources including the Wellcome Trust and the Medical Research Council. Drs Tilling and Sterne are funded by the Higher Education Funding Council for England. Dr Emond is funded by North Bristol National Health Service trust and receives funding from the Higher Education Funding Council for England. The UK Medical Research Council (grant 74882), the Wellcome Trust (grant 076467), and the University of Bristol provided core support for Avon Longitudinal Study of Parents and Children.
- Viner R,
- Hotopf M
- Rimes KA,
- Goodman R,
- Hotopf M,
- Wessely S,
- Meltzer H,
- Chalder T
- Ghandour RM,
- Overpeck MD,
- Huang ZJ,
- Kogan MD,
- Scheidt PC
- ter Wolbeek M,
- van Doornen LJ,
- Kavelaars A,
- Heijnen CJ
- Jordan KM,
- Ayers PM,
- Jahn SC,
- et al
- Chalder T,
- Goodman R,
- Wessely S,
- Hotopf M,
- Meltzer H
- ↵National Institute for Health and Clinical Excellence. Chronic fatigue syndrome/myalgic encephalomyelitis (or encephalopathy); diagnosis and management. Report no. CG53. National Institute for Health and Clinical Excellence; 2007
- Dobbins JG,
- Randall B,
- Reyes M,
- Steele L,
- Livens EA,
- Reeves WC
- Haines LC,
- Saidi G,
- Cooke RW
- Nijhof SL,
- Maijer K,
- Bleijenberg G,
- Uiterwaal CS,
- Kimpen JL,
- van de Putte EM
- Bowen J,
- Pheby D,
- Charlett A,
- McNulty C
- Clark C,
- Goodwin L,
- Stansfeld SA,
- Hotopf M,
- White PD
- Rubin DB
- Bould H,
- Lewis G,
- Emond A,
- Crawley E
- Farmer A,
- Fowler T,
- Scourfield J,
- Thapar A
- Copyright © 2012 by the American Academy of Pediatrics