Myalgic encephalomyelitis, chronic fatigue syndrome, or systemic exertion intolerance disease (ME/CFS/SEID) is a medical condition in which the dominant symptom is unexplained fatigue that persists for at least 1 month (6,11,17,56). The estimated prevalence of ME/CFS/SEID in the United States and England ranges from 0.007% to 2.8% (16,27,47,55,59). Diagnosis remains difficult because ME/CFS/SEID has no identified cause and requires that other illnesses and causes of fatigue be excluded (24). ME/CFS/SEID may be triggered by viral infections or environmental factors that cause immune dysfunction and inflammation (3) or autonomic and central nervous system pathologies, but the etiology remains uncertain (24). Although exercise-induced fatigue is a defining diagnostic feature, a full understanding of how much acute exertion or exercise might increase ME/CFS/SEID fatigue is unknown.
The lack of an apparent single cause of ME/CFS/SEID has led to a variety of treatments and to comparison studies of treatments (71). Exercise training increases feelings of energy and reduces feelings of fatigue in the general population (53) and has been used as an ME/CFS/SEID treatment. In 2004, a meta-analysis of five randomized controlled trials quantified the effect of exercise training on fatigue. Twelve weeks of exercise training resulted in a moderate-to-large reduction in fatigue when compared with control conditions. Exercise training combined with drug treatment (fluoxetine) was more effective than drug treatment alone, and exercise training combined with patient education was more effective than education alone (12). A 2011 meta-analysis reported that graded exercise training (five trials) and cognitive behavioral therapy (16 trials) were equally effective for treating people with CFS/ME on a range of outcomes, including functional impairment, depression, anxiety, and fatigue (8). A 2015 quantitative review of eight randomized controlled trials including 1518 CFS participants confirmed that exercise training more effectively reduced fatigue than passive control conditions such as usual care, relaxation, or flexibility training (32). This body of evidence suggests that exercise training could be efficacious for reducing fatigue states among many people with ME/CFS/SEID, although some individuals report that exercise training worsens their fatigue (63).
This literature has led to recommendations for designing exercise programs for people with ME/CFS/SEID (2,64). However, concern remains that acute physical exertion may result in adverse events for some people with ME/CFS/SEID (29,30). Indeed, an exacerbation of symptoms after acute exertion is used to diagnose ME/CFS/SEID (11,17). Fatigue is reduced after standardized (e.g., relative to each individual’s aerobic capacity) low-intensity acute exercise lasting at least 20 min in people without ME/CFS/SEID (38). No standardized acute exercise stimulus is used in diagnosis of ME/CFS/SEID, although it is noted that “minimal” activity may provoke symptoms in some people (7,11). The treatment potential of exercise training and the risk of increased fatigue after acute exercise are paradoxical, as chronic exercise adaptations are often viewed as the cumulative effect of many acute bouts of exercise on both physiological (62) and psychological (52) outcomes.
Studies have measured the effect of acute exercise on fatigue states in people with ME/CFS/SEID, but the exercise features have varied (31,34,36,42,50,58,65). The purpose of this systematic review and meta-analysis was to estimate the size of the mean effect of completing a single bout of exercise on fatigue states reported by people with ME/CFS/SEID. A second objective was to determine what patient characteristics, exercise stimulus factors, or measurement features moderate the size of the effect.
This meta-analysis was completed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement (33).
A search was conducted in Google Scholar for peer-reviewed studies written in English and published between February 1991 and May 2015. A single “advanced scholar search” was conducted for articles with the exact phrase “chronic fatigue syndrome” and “exercise.” The reference list of each article that met the inclusion criteria was manually searched for relevant studies in addition to the reference lists of relevant review articles (8,12,14,41,48,57).
The inclusion criteria used to screen studies for the analysis was the following: i) ME/CFS/SEID and healthy control participants completed a single bout of exercise; ii) participants were diagnosed with CFS/ME using either the Centers for Disease Control (CDC)/Fukuda criteria (17), the Canadian Consensus Criteria (6), or the Myalgic Encephalomyelitis Consensus Criteria (7); iii) self-reported fatigue was measured using a unipolar scale immediately before and after exercise; and iv) data usable for the analysis appeared in a peer-reviewed publication written in the English language. A flow diagram (Fig. 1) shows the process of study selection.
A total of seven studies contributing 47 total effects (median = 3 effects per study) met the inclusion criteria. The studies examined 159 people with ME/CFS/SEID, with a median of 13 people per study. Additional study features are shown in Table 1. Data were presented in text for 18 effects and in figures for 29 effects. Lead authors who presented data in figures were contacted by e-mail with requests for precise data before the final analyses were completed. The lead author of one study (34) provided data, and the final meta-analysis included five effects that were derived from figures. The first and the second authors independently derived all effects, and a two-way (effects × raters) intraclass correlation (ICC) for absolute agreement was calculated to test the reliability of effect size calculations. There was high initial agreement between authors (ICC [2, 46] = 0.97, 95% confidence interval [CI] = 0.95–0.98), and the remaining discrepancies due to mathematical or data extraction errors were resolved.
Effect size calculation
The Hedges g effect sizes were calculated by subtracting the change in the control group (postexercise fatigue mean minus preexercise fatigue mean) from the change in the ME/CFS/SEID group. This difference was divided by the preexercise pooled standard deviation (20). Positive effect sizes were calculated when larger postexercise fatigue increases occurred in the CFS/ME/SEID group, and negative effect sizes indicated larger postexercise fatigue increases for the control group. Effect sizes were corrected for small sample bias to yield the Hedges d (20,35).
Moderator selection and coding
Several factors were hypothesized to influence the size and/or direction of the effect of acute exercise on fatigue. These factors, or moderators, were chosen and coded based on a priori rationale and the data consistently reported in the included studies. The seven moderators selected were categorized as diagnostic characteristics (ME/CFS diagnosis criteria and fibromyalgia comorbidity), exercise stimulus factors (intensity and type), or measurement features (construct, format, and measurement time after exercise).
Although some organizations have worked to create international consensus criteria for diagnosing ME (7) and SEID (11), others suggest the varied diagnostic criteria reflect the continued debate and uncertainty over causes of ME/CFS (24). To reflect this debate, studies were coded based on the CDC (17), Oxford (56), or Canadian (6) criteria. Because fibromyalgia and ME/CFS are often comorbid (7), studies were also coded based on the inclusion of at least one person diagnosed with fibromyalgia using the American College of Rheumatology criteria (74).
Exercise stimulus factors
Light and moderate, but not vigorous, intensity exercise has previously been reported to reduce fatigue (38). Exercise intensity coded as light, moderate, or vigorous according to a position stand from the American College of Sports Medicine (19). The exercise type used in each study was coded as continuous-increasing, continuous-steady, or discontinuous.
The specific construct measured was coded (overall fatigue, mental fatigue, and physical fatigue) because previous research has reported different effects of acute exercise on mental and physical fatigue (37). Visual analog scales (VAS) are more precise and potentially more sensitive to change than categorical questionnaires; therefore, the measurement format was coded, as was measurement time (during to immediately after exercise, 5 min to ≤1 h after, and ≥4 h after exercise cessation).
The mean effect size (Δ) and the 95% CI were computed overall using a random-effects model (35). The I2 statistic was used to test the heterogeneity of mean effects (22), and funnel plots with Egger’s test were used to examine risk of publication bias (13).
IBM SPSS version 20 was used with macros (73) to calculate the overall mean effect size and to conduct a moderator analysis using multiple linear regression analysis with maximum likelihood estimation. The use of more than one scale to measure fatigue or repeated time points yielded effects that were nested within studies (median of three effects per study) that might systematically differ from one another. Hence, a multilevel model with robust maximum likelihood estimation in Mplus 7.3.1 (46) was used to adjust for between-study variance and correlated effects within studies according to standard procedures (10,25). Parameters and their errors were estimated with clustering on study using the Huber–White sandwich estimator to calculate standard errors that are robust to heteroscedasticity and correlated effects (15,70,72). Although simulations indicate that regression coefficients in multilevel models can be estimated without substantial bias in as few as 10 studies (39), maximum likelihood estimates of standard errors will be too small (39). Here, Monte Carlo integration was used to provide robust maximum likelihood estimates (46). The conditional model (which included the intercept and the moderators) was compared with the unconditional intercept-only model using a likelihood ratio test and the adjusted Bayesian information criterion (BIC) (25). Two- and three-way interactions of the exercise stimulus (i.e., intensity and type) and the time when fatigue was measured were also tested the same way in a single multiple linear regression model.
Fatigue increased after exercise in 40% (19 of 47) of the effects for control groups and 91% (43 of 47) of the effects for ME/CFS/SEID. The distribution of effect sizes is shown in a forest plot (Fig. 2A), and the weighted effect size of each study is shown in Figure 2B. Ninety-one percent (43 of 47) of the effects were positive, which indicates that postexercise fatigue increases were larger for people with ME/CFS/SEID than control participants. The mean effect size was Δ = 0.82 (95% CI = 0.66–0.97), and heterogeneity was moderately high, I2 = 67% (95% CI = 61%–72%). In the multilevel model (χ22 = 190.1, BIC = 191.6), the mean delta was 0.73 (95% CI = 0.24–1.23) with nonsignificant variance between studies (0.37, SE = 0.28, z = 1.29, P = 0.198). However, the between-study variance accounted for 55% of the total variance (ICC = 0.55). Because statistical power to detect variance that size was low here (0.54 at an alpha of 0.05)(25), the planned moderator analysis was conducted using multilevel modeling to adjust standard errors for between-study variance.
A funnel plot (Fig. 3) of individual fatigue effects was asymmetrical. A visual examination of the funnel plot suggests that effects with larger standard errors were more likely to be zero or negative, showing a decrease or no change in fatigue for ME/CFS/SEID groups (outside the low 95% CI). Effects with smaller standard errors were more often positive (outside the high 95% CI) and should be interpreted as showing a larger increase in fatigue for ME/CFS/SEID than control groups. The estimate of the mean effect size also does not increase in precision as standard error decreases. Egger’s test for bias was statistically significant, F1,46 = 12.12, P = 0.001. However, funnel plot asymmetry and potential bias identified via Egger’s test can yield a false-positive indication of bias when the true effect is actually larger in effects with small sample size because of unique characteristics of the sample population (60). A sensitivity analysis that excluded the four smallest and the five largest outlying effects yielded a similar mean effect of exercise on fatigue, Δ = 0.81 (95% CI = 0.69–0.94), and heterogeneity remained moderately high, I2 = 43% (95% CI = 31%–53%). The multilevel (χ22 = 132.1, BIC = 133.1) mean Δ was 0.76 (95% CI = 0.54–0.97) with nonsignificant variance between studies (0.013, SE = 0.015, z = 0.86, P = 0.392).
Univariate moderator results are shown in Table 2. Exercise intensity significantly moderated the effect, QB(2,44) = 13.29, P = 0.001, with only the 95% CI for light-intensity exercise including zero. Exercise type also significantly moderated the effect, QB(2,44) = 21.48, P < 0.001. Discontinuous (handgrip) exercise resulted in a negative effect size estimate, indicating larger increases in postexercise fatigue for control than ME/CFS/SEID groups. Continuous-increasing and continuous-steady exercise both increased postexercise fatigue, and continuous-steady exercise led to the largest increases in fatigue for people with ME/CFS/SEID relative to control groups. The time that fatigue was measured also moderated the effect, QB(2,46) = 12.24, P = 0.002. The effect size was smaller when fatigue was measured during or ≤l h after exercise compared with measurements made ≥4 h after exercise. The construct measured, QB(2,44) = 10.85, P = 0.004, and the measurement format, QB(1,45) = 10.48, P = 0.001, also moderated the effect size (see Table 2). Fatigue was increased in ME/CFS/SEID regardless of the fatigue construct or scale format, but effects were significantly larger when measures specified fatigue as mental or physical and when VAS were used.
Multiple linear regression analysis indicated that only intensity (B = 0.291, z = 2.23, P = 0.026) and the time when fatigue was measured (B = 0.657, z = 4.88, P < 0.001) were independently related to the increase in fatigue in ME/CFS/SEID relative to control groups. Effects for fatigue were greater after moderate-intensity exercise and when fatigue was measured 4 h or more after exercise ended. The type of exercise (B = −0.007, z = 0.067, P = 0.95), the fatigue construct (B = 0.153, z = 1.07, P = 0.29), and the measurement format (B = 0.120, z = 0.598, P = 0.55) were not significantly related to effect size. After multilevel adjustment of standard errors, only the time when fatigue was measured was related to fatigue effect size (B = 0.683, z = 6.59, P < 0.001). Model fit [χ2(7) = 151.9, BIC = 156.9] was improved [Δχ2(5) = −43.6, P < 0.001] compared with the intercept-only model [χ2(2) = 190.1, BIC = 191.6]. This effect remained in the sensitivity analysis that removed outlying effects (B = 0.580, z = 7.86, P < 0.001). Two- and three-way interactions of the exercise stimulus (i.e., intensity and type) and the time when fatigue was measured were not significant (P values ≥ 0.186).
The primary finding of this meta-analysis is that a single session of exercise resulted in larger fatigue increases for people with ME/CFS/SEID than control groups. The overall effect size is consistent with previous reports of increased fatigue after exertion in ME/CFS/SEID (6,7,11,17). Multiple regression analysis of moderators indicated that intensity of exercise and the time when fatigue was measured were independent influences on variation in fatigue effects. However, the heterogeneity of the fatigue effects was mainly explained by variation between studies. After a multilevel adjustment of standard errors for between-study variation, only the time when fatigue was measured influenced effect size. The increase in fatigue among people with ME/CFS/SEID was greater when fatigue was measured more than 4 h (and up to 96 h) after exercise ended than when fatigue was measured during or within 40 min after the exercise session ceased. A previous meta-analysis of 16 studies, none of which included people diagnosed with ME/CFS/SEID, also reported that postexercise fatigue effects were heterogeneous, but that fatigue was reduced in people with elevated fatigue when at least 20 min of exercise was performed at a light or moderate intensity (38). Thus, findings here and other meta-analytic reviews have found that the fatigue effect size is influenced by exercise intensity, although fatigue is still generally increased postexercise in people with ME/CFS/SEID.
Exercise intensity significantly moderated the size of the fatigue effect as moderate- or vigorous-intensity exercise increased postexercise fatigue, but fatigue was not significantly increased in two studies of light-intensity exercise (36,50). Light-intensity exercise has previously been shown to reduce fatigue in people with elevated fatigue but not diagnosed with ME/CFS/SEID (38), although others have stated that any exertion may increase fatigue in ME/CFS/SEID (11). In addition, fatigue may not be caused by moderate- to vigorous-intensity exercise itself but could theoretically be triggered by related factors, including postexercise muscle injury, inflammation, or psychological stress caused by performing vigorous-intensity exercise (41). Because only two studies contributing five effects (11% of the total) examined light-intensity exercise, confidence in this result is minimal, and future studies should examine the effect of light-intensity on fatigue in people with ME/CFS/SEID.
The finding that fatigue is elevated from baseline for several days after exercise is not unexpected given that postexertional malaise sometimes lasting greater than 24 h is part of the diagnostic criteria for ME/CFS/SEID (6,7,11,17). This result remained significant when a sensitivity analysis was conducted, which included only studies that measured fatigue multiple times after exercise. Postexertional malaise is considered a separate feature of ME/CFS/SEID that is related to fatigue but also includes a cluster of other symptoms (e.g., pain, mood disturbance) that are worsened after exertion (26). Although one prior study suggested that symptoms categorizing postexertional malaise do not peak until 5 d after exercise is performed (75), an unexpected result here was a larger fatigue effect size when measurements were conducted for more than 4 h (and up to 96 h) postexercise. There are several reasons this could have occurred. Delayed onset muscle soreness peaks ~48 h after exercise is performed and is associated with pain and inflammation of the muscle (9). Although the exact mechanism of delayed onset muscle soreness is unknown, neutrophils may have a role (9). Neutrophils may also be regulated by the inflammatory cytokine interleukin 6 (IL-6) (28), and IL-6 has also been suggested as a fatigue mediator in both exercise and illness (66). In one study, the increase from baseline in several pro-inflammatory cytokines (IL-6 and IL-1ß) was larger at 8 h postexercise for ME/CFS/SEID than control participants, and VAS ratings of mental and physical fatigue were also greater (69). The cessation of regular daily exercise also results in fatigue increases that peak 48 h after the last exercise session (43), although there is no clear neurobiological explanation for this effect. Most (21 of 25, 84%) of the fatigue ratings made ≥4 h after exercise were conducted in a laboratory setting along with other physiological measures, which might limit the influence of extraneous factors on fatigue ratings.
A potential limitation of the literature meta-analyzed here was the imprecise prescription of exercise intensity. One study used an absolute exercise prescription, i.e., a 1-mph walk (50). Prescribing unstandardized exercise, rather than relativizing to fitness level, may lead to an over or underestimation of intensity depending on the fitness level of the participant. Another study assigned exercise intensity based on age-predicted HRmax (34). Precise exercise prescription is important because intensity moderates fatigue (38). There is a possibility of substantial error when prescribing exercise intensity by age-predicted HRmax because age-predicted HRmax is only moderately correlated (r = 0.73) with actual HRmax (61). Three other studies examined the effects of progressive, maximal exercise tests on fatigue (31,42,58). Such tests are used to assess cardiorespiratory fitness but do not represent the normal activity patterns of people with CFS/ME/SEID (67). This limitation may have influenced the precision of the exercise intensity moderator analysis. Because only two studies prescribed exercise relativized to the fitness level of each participant (36,65), additional experiments are needed that document the effect of light- and moderate-intensity exercise in ME/CFS/SEID. In addition, trials that have reported fatigue reductions after chronic exercise training have prescribed exercise based on cardiorespiratory fitness tests (18,68,71). Future ME/CFS/SEID acute exercise studies should also conduct graded exercise tests to determine V˙O2peak (45) and then prescribe chosen exercise intensity as a percentage of the peak (19).
It remains uncertain why exercise training may effectively reduce fatigue in people with ME/CFS/SEID (8,12,32,71), although the results of this meta-analysis support a worsening of fatigue after a single session of exercise. Differences between the respective bodies of literature include the study design, the exercise stimulus chosen, and the fatigue measures used. None of the studies in the meta-analysis included sedentary control conditions to account for confounding variables associated with the passage of time. For example, fatigue could change because of the completion of experimental cognitive tasks (37) or circadian variation (54), and it is uncertain if changes in fatigue reported in the analysis here are due to exercise itself or extraneous factors. Conversely, exercise training trials have often compared exercise to a sedentary treatment, such as drug (68) or relaxation (32). Future acute exercise studies should use control groups or conditions, and it is recommended that repeated-measures designs be used because of the possibility that ME/CFS/SEID is not a homogeneous condition (1). No prior acute exercise study has used a within-subjects design with random assignment to exercise or quiet rest. It is unknown how the fatigue responses to exercise may differ as a function of time within the same person. Although prior studies have shown that acute, high-intensity exercise increases fatigue (51), it is possible that these individuals would have reported even greater fatigue had they not exercised.
Second, when chronic exercise has reduced fatigue in people with ME/CFS/SEID (68,71), graded exercise training is used where training begins at 40% of HRmax (18), which is categorized as light-intensity exercise (19). In the present meta-analysis, only two small studies (accounting for <15% of the patients) tested the effect of light-intensity exercise on fatigue. Only one study included in the analysis examined a form of resistance exercise (36), and no studies were located that examined acute weight lifting exercise. Single sessions of weight lifting have previously been shown to reduce fatigue (21). In people who are not diagnosed with ME/CFS/SEID, both acute aerobic and resistance exercises are equally efficacious for reducing fatigue (38). Acute weight lifting exercise could be more effective for people with ME/CFS/SEID because the only included study examining discontinuous exercise found statistically nonsignificant decreases in postexercise fatigue. A single bout of weight lifting may also confer additional mental health benefits, including reductions in anxiety and depression symptoms, improved sleep quality, enhanced cognition, and increased self-esteem (49). It is recommended that future experiments determine whether a single bout of weight lifting exercise is efficacious or harmful for people diagnosed with ME/CFS/SEID and that larger studies examine the effect of light-intensity exercise (aerobic or resistance) on fatigue in this population. Future studies should also investigate the benefit of other forms of discontinuous exercise, as one study found larger fatigue reductions immediately after intermittent cycling sprints than during continuous exercise (44).
Third, the paradox (in the published literature) between acute and chronic exercise in ME/CFS/SEID may be related to fatigue measurement. In the moderator analysis, VAS resulted in larger fatigue effect sizes than categorical scales, although this did not remain significant after adjustment for correlated effects within studies. Previous exercise training trials in ME/CFS/SEID have used categorical scales (68,71) rather than VAS, and some categorical scales may also measure multiple symptoms besides fatigue alone (23). In addition, baseline fatigue may also affect the effect of acute exercise on fatigue (38). This meta-analysis was not able to consider baseline fatigue as a moderator because some studies used measures without established normative values, which makes it difficult to directly compare fatigue measures. Future studies should use fatigue measures with normative values so such comparisons may be conducted.
Finally, related neurobiological precedent exists for an acute–chronic paradox. The initial treatment of anxiety disorders with selective serotonin reuptake inhibitors (SSRI) can produce an increase in anxiety symptoms and the processing of anxiety-related stimuli (4), but over time, the SSRI are effective for the treatment of anxiety disorders (40). The available evidence from rodents supports a model to explain the paradox; in particular, that acute SSRI administration acts directly on the bed nucleus of the stria terminalis and central amygdala to augment the expression of cued fear conditioning, whereas chronic SSRI treatment yields the inhibition of fear learning by modulating glutamatergic actions in the amygdala (5). More human and animal data are needed before a plausible, defensible model can help to explain why acute exercise increases fatigue symptoms and chronic exercise reduces fatigue symptoms in people with ME/CFS/SEID.
The systematic review and meta-analysis limited to seven studies, including 159 people with ME/CFS/SEID, found that a single bout of exercise moderately increased postexercise feelings of fatigue in people diagnosed with ME/CFS/SEID more than control participants. Although more studies are needed to verify the accuracy of the effects estimated here, the effect size distributions observed here were each moderately heterogeneous and moderated by exercise intensity and the time after exercise that fatigue was measured. On the basis of the published acute exercise literature meta-analyzed here, it is too soon to conclude that all doses of exercise cause fatigue. Future experiments should be conducted to further elucidate the potential efficacy or harm of a range of doses (intensity and duration) and modalities of acute exercise for people with ME/CFS/SEID. Additional studies should also consider patient characteristics (e.g., symptom severity and illness length) that were not reported in studies analyzed here but may influence the effect of exercise on fatigue. The current meta-analysis also could not consider diagnostic criteria as a moderator because all studies used the CDC criteria (17), and this should be a consideration in future research. Examining the full spectrum of exercise doses may also lead to additional mechanistic insight on the paradox between the harm of a single bout of exercise and the reported benefit of exercise training on fatigue in people with ME/CFS/SEID.
The authors thank David Kupshik, Kaitlin Lang, Thomas Austin Nuckols, Daniel Schulz, and Taylor Vaughn for assistance with article retrieval. This work was supported by NIH-NCCIH T32 AT002688 (BDL).
The authors declare no conflicts of interest. The results of the present study do not constitute endorsement by the American College of Sports Medicine.
1. Afari N, Buchwald D. Chronic fatigue syndrome: a review. Am J Psychiatry
2. Bailey SP. Tired of being tired: exercise as a treatment for chronic fatigue syndrome. Disabil Rehabil
3. Bansal AS, Bradley AS, Bishop KN, Kiani-Alikhan S, Ford B. Chronic fatigue syndrome, the immune system and viral infection. Brain Behav Immun
4. Browning M, Reid C, Cowen PJ, Goodwin GM, Harmer CJ. A single dose of citalopram increases fear recognition in healthy subjects. J Psychopharmacol
5. Burghardt NS, Bauer EP. Acute
and chronic effects of selective serotonin reuptake inhibitor treatment on fear conditioning: implications for underlying fear circuits. Neuroscience
6. Carruthers BM, Jain AK, De Meirleir KL, et al. Myalgic encephalomyelitis
/chronic fatigue syndrome: clinical working case definition, diagnostic and treatment protocols. J Chronic Fatigue Syndr
7. Carruthers BM, van de Sande MI, De Meirleir KL, et al. Myalgic encephalomyelitis
: international Consensus Criteria. J Intern Med
8. Castell BD, Kazantzis N, Moss-Morris R. Cognitive behavioral therapy and graded exercise for chronic fatigue syndrome: a meta-analysis. Clin Psychol: Sci Pr
9. Cheung K, Hume PA, Maxwell L. Delayed onset muscle soreness: treatment strategies and performance factors. Sports Med
10. Cheung MA. A model for integrating fixed-, random-, and mixed-effects meta-analyses into structural equation modeling. Psychol Methods
11. Clayton EW, Alegria M, Bateman L, et al. Beyond myalgic encephalomyelitis
/chronic fatigue syndrome-Redefining an illness. JAMA
12. Edmonds M, McGuire H, Price JR. Exercise therapy for chronic fatigue syndrome. Cochrane Database Syst Rev
13. Egger M, Smith GD, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ
14. Evering RM, van Weering MG, Groothuis-Oudshoorn KC, Vollenbroek-Hutten MM. Daily physical activity of patients with the chronic fatigue syndrome: a systematic review. Clin Rehabil
15. Froot KA. Consistent covariance matrix estimation with cross-sectional dependence and heteroskedasticity in financial data. J Financ Quant Anal
16. Fukuda K, Dobbins JG, Wilson LJ, Dunn RA, Wilcox K, Smallwood D. An epidemiologic study of fatigue with relevance for the chronic fatigue syndrome. J Psychiatric Res
17. Fukuda K, Straus SE, Hickie I, et al. The chronic fatigue syndrome: a comprehensive approach to its definition and study. International Chronic Fatigue Syndrome Study Group. Ann Intern Med
18. Fulcher KY, White PD. Chronic fatigue syndrome: a description of graded exercise treatment. Physiotherapy
19. Garber CE, Blissmer B, Deschenes MR, et al. American College of Sports Medicine Position Stand: quantity and quality of exercise for developing and maintaining cardiorespiratory, musculoskeletal, and neuromotor fitness in apparently healthy adults: guidance for prescribing exercise. Med Sci Sports Exerc
20. Hedges LV, Olkin I. Statistical Methods for Meta-Analysis
. New York: Academic Press; 1985. p. 369.
21. Herring MP, O’Connor PJ. The effect of acute
resistance exercise on feelings of energy
and fatigue. J Sports Sci
22. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analysis. BMJ
23. Hjollund NH, Andersen JH, Bech P. Assessment of fatigue in chronic disease: a bibliographic study of fatigue measurement scales. Health Qual Life Outcomes
24. Holgate ST, Komaroff AL, Mangan D, Wessely S. Chronic fatigue syndrome: understanding a complex illness. Nat Rev Neurosci
25. Hox J. Multilevel Analysis: Techniques and Applications
. New York: Routledge; 2010. p. 392.
26. Jason LA, Evans M, So S, Scott J, Brown A. Problems in defining post-exertional malaise
. J Prev Interv Community
27. Jason LA, Richman JA, Rademaker AW, et al. A community-based study of chronic fatigue syndrome. Arch Intern Med
28. Kaplanski G, Marin V, Montero-Julian F, Mantovani A, Farnarier C. IL-6: a regulator of the transition from neutrophil to monocyte recruitment during inflammation. Trends Immunol
29. Kindlon T. Reporting of harms associated with graded exercise therapy and cognitive behavioural therapy in myalgic encephalomyelitis
/chronic fatigue syndrome. Bull IACFS/ME
30. Kindlon T, Goudsmit EM. Graded exercise for chronic fatigue syndrome: too soon to dismiss reports of adverse reactions. J Rehabil Med
. 2010;42(2):184–6; author reply.
31. LaManca JJ, Sisto SA, Zhou X, et al. Immunological response in chronic fatigue syndrome following a graded exercise test to exhaustion. J Clin Immunol
32. Larun L, Brurberg KG, Odgaard-Jensen J, Price JR. Exercise therapy for chronic fatigue syndrome. Cochrane Database Syst Rev
33. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. Ann Intern Med
34. Light AR, Bateman L, Jo D, et al. Gene expression alterations at baseline and following moderate exercise in patients with Chronic Fatigue Syndrome and Fibromyalgia Syndrome. J Intern Med
35. Lipsey MW, Wilson DB. Practical Meta-Analysis
. Thousand Oaks (CA): SAGE Publications; 2001. p. 264.
36. Lloyd A, Gandevia S, Brockman A, Hales J, Wakefield D. Cytokine production and fatigue in patients with chronic fatigue syndrome and healthy control subjects in response to exercise. Clin Infect Dis
. 1994;18(1 Suppl):S142–6.
37. Loy BD, O’Connor PJ. The effect of histamine on changes in mental energy
and fatigue after a single bout of exercise. Physiol Behav
38. Loy BD, O’Connor PJ, Dishman RK. The effect of a single bout of exercise on energy
and fatigue states: a systematic review and meta-analysis. Fatigue
39. Maas CJM, Hox JJ. Sufficient sample sizes for multilevel modeling. Methodology
40. Mayo-Wilson E, Dias S, Mavranezouli I, et al. Psychological and pharmacological interventions for social anxiety disorder in adults: a systematic review and network meta-analysis. Lancet Psychiatry
41. McCully KK, Sisto SA, Natelson BH. Use of exercise for treatment of chronic fatigue syndrome. Sports Med
42. Meyer JD, Light AR, Shukla SK, et al. Post-exertion malaise
in chronic fatigue syndrome: symptoms and gene expression. Fatigue
43. Mondin GW, Morgan WP, Piering PN, et al. Psychological consequences of exercise deprivation in habitual exercisers. Med Sci Sports Exerc
44. Monroe DC, Gist NH, Freese EC, O’Connor PJ, McCully KK, Dishman RK. Effects of sprint interval cycling on fatigue, energy
, and cerebral oxygenation. Med Sci Sports Exerc
45. Mullis R, Campbell IT, Wearden AJ, Morriss RK, Pearson DJ. Prediction of peak oxygen uptake in chronic fatigue syndrome. Br J Sports Med
46. Muthen LK, Muthen BO. Mplus Statistical Analysis with Latent Variables User’s Guide
. Muthen & Muthen: Los Angeles; 1998–2012. p. 752.
47. Nacul LC, Lacerda EM, Pheby D, et al. Prevalence of myalgic encephalomyelitis
/chronic fatigue syndrome (ME/CFS) in three regions of England: a repeated cross-sectional study in primary care. BMC Med
48. Nijs J, Paul L, Wallman K. Chronic fatigue syndrome: an approach combining self-management with graded exercise to avoid exacerbations. J Rehabil Med
49. O’Connor PJ, Herring MP, Caravalho A. Mental health benefits of strength training in adults. Am J Lifestyle Med
50. Peterson PK, Sirr SA, Grammith FC, et al. Effects of mild exercise on cytokines and cerebral blood flow in chronic fatigue syndrome patients. Clin Diagn Lab Immunol
51. Pronk NP, Crouse SF, Rohack JJ. Maximal exercise and acute mood
response in women. Physiol Behav
52. Puetz TW, Flowers SS, O’Connor PJ. A randomized controlled trial of the effect of aerobic exercise training on feelings of energy
and fatigue in sedentary young adults with persistent fatigue. Psychother Psychosom
53. Puetz TW, O’Connor PJ, Dishman RK. Effects of chronic exercise on feelings of energy
and fatigue: a quantitative synthesis. Psychol Bull
54. Rahman K, Burton A, Galbraith S, Lloyd A, Vollmer-Conna U. Sleep-wake behavior in chronic fatigue syndrome. Sleep
55. Reyes M, Nisenbaum R, Hoaglin DC, et al. Prevalence and incidence of chronic fatigue syndrome in Wichita, Kansas. Arch Intern Med
56. Sharpe MC, Archard LC, Banatvala JE, et al. Chronic fatigue syndrome: guidelines for research. J Roy Soc Med
57. Shephard RJ. Chronic fatigue syndrome: an update. Sports Med
58. Sisto SA, LaManca J, Cordero DL, et al. Metabolic and cardiovascular effects of a progressive exercise test in patients with chronic fatigue syndrome. Am J Med
59. Steele L, Dobbins JG, Fukuda K, et al. The epidemiology of chronic fatigue in San Francisco. Am J Med
60. Sterne JAC, Gavaghan D, Egger M. Publication and related bias in meta-analysis: power of statistical tests and prevalence in the literature. J Clin Epidemiol
61. Tanaka H, Monahan KD, Seals DR. Age-predicted maximal heart rate revisited. J Am Coll Cardiol
62. Thompson PD, Crouse SF, Goodpaster B, Kelley D, Moyna N, Pescatello LS. The acute
versus the chronic response to exercise. Med Sci Sports Exerc
63. Twisk FN, Maes M. A review on cognitive behavioral therapy (CBT) and graded exercise therapy (GET) in myalgic encephalomyelitis
(ME)/chronic fatigue syndrome (CFS): CBT/GET is not only ineffective and not evidence-based, but also potentially harmful for many patients with ME/CFS. Neuroendocrinol Lett
64. Van Cauwenbergh D, De Kooning M, Ickmans K, Nijs J. How to exercise people with chronic fatigue syndrome: evidence-based practice guidelines. Eur J Clin Invest
65. Van Oosterwijck J, Nijs J, Meeus M, et al. Pain inhibition and postexertional malaise
in myalgic encephalomyelitis
/chronic fatigue syndrome: an experimental study. J Intern Med
66. Vargas NT, Marino F. A neuroinflammatory model for acute
fatigue during exercise. Sports Med
67. Vercoulen JHMM, Bazelmans E, Swanink CMA, et al. Physical activity in chronic fatigue syndrome: assessment and its role in fatigue. J Psychiatr Res
68. Wearden AJ, Morriss RK, Mullis R, et al. Randomised, double-blind, placebo-controlled treatment trial of fluoxetine and graded exercise for chronic fatigue syndrome. Br J Psychiat
69. White AT, Light AR, Hughen RW, et al. Severity of symptom flare after moderate exercise is linked to cytokine activity in chronic fatigue syndrome. Psychophysiology
70. White H. A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica
71. White PD, Goldsmith KA, Johnson AL, et al. Comparison of adaptive pacing therapy, cognitive behaviour therapy, graded exercise therapy, and specialist medical care for chronic fatigue syndrome (PACE): a randomised trial. Lancet
72. Williams RL. A note on robust variance estimation for cluster-correlated data. Biometrics
74. Wolfe F, Smythe HA, Yunus MB, et al. The American College of Rheumatology 1990 Criteria for the Classification of Fibromyalgia. Report of the Multicenter Criteria Committee. Arthritis Rheum
75. Yoshiuchi K, Cook DB, Ohashi K, et al. A real-time assessment of the effect of exercise in chronic fatigue syndrome. Physiol Behav