The effectiveness of activity pacing interventions for people with chronic fatigue syndrome: a systematic review and meta-analysis

Abstract Purpose To investigate whether activity pacing interventions (alone or in conjunction with other evidence-based interventions) improve fatigue, physical function, psychological distress, depression, and anxiety in people with chronic fatigue syndrome (CFS). Materials and methods Seven databases were searched until 13 August 2022 for randomised controlled trials that included activity pacing interventions for CFS and a validated measure of fatigue. Secondary outcomes were physical function, psychological distress, depression, and anxiety. Two reviewers independently screened studies by title, abstract and full text. Methodological quality was evaluated using the PEDro scale. Random-effects meta-analyses were performed in R. Results 6390 articles were screened, with 14 included. Good overall study quality was supported by PEDro scale ratings. Activity pacing interventions were effective (Hedges’ g (95% CI)) at reducing fatigue (–0.52 (–0.73 to −0.32)), psychological distress (–0.37 (–0.51 to −0.24)) and depression (–0.29 (–0.49 to −0.09)) and improving physical function (mean difference 7.18 (3.17–11.18)) when compared to no treatment/usual care. The extent of improvement was greater for interventions that encouraged graded escalation of physical activities and cognitive activities. Conclusion Activity pacing interventions are effective in reducing fatigue and psychological distress and improving physical function in CFS, particularly when people are encouraged to gradually increase activities. Registration PROSPERO CRD42016036087. IMPLICATIONS FOR REHABILITATION A key feature of chronic fatigue syndrome (CFS) is a prolonged post-exertional exacerbation of symptoms following physical activities or cognitive activities. Activity pacing is a common strategy often embedded in multi-component management programs for CFS. Activity pacing interventions are effective in reducing fatigue and psychological distress and improving physical function in CFS, particularly when patients are encouraged to gradually increase their activities. Healthcare professionals embedding activity pacing as part of treatment should work collaboratively with patients to ensure successful, individualised self-management strategies.


Introduction
Chronic fatigue syndrome (CFS), also commonly referred to as myalgic encephalomyelitis (ME) or ME/CFS, is a disabling illness estimated to affect between 0.76% (95% CI: 0.23-1.29)and 3.28% (95% CI: 2. 24-4.33) of the population, depending on whether the case definition is met through comprehensive clinical assessment or self-reporting of symptoms, respectively [1].The diagnosis of CFS rests upon the subjective report of six or more months of unexplained fatigue, including "physical" fatigue, which is exacerbated by exercise or other physical activity; as well as a comparable phenomenon of neurocognitive difficulties, which are exacerbated by concentration-demanding mental activities [2,3].The most commonly used diagnostic criteria were drafted by an international expert group convened by the Centers for Disease Control (USA) and require disabling fatigue, and four of eight additional symptoms (neurocognitive difficulties, sore throat, tender lymph nodes, muscle pain, joint pain, headaches, unrefreshing sleep and post-exertional malaise), as well as the careful exclusion of alternative medical or psychiatric conditions [4].Other diagnostic criteria (i.e., the 2003 Canadian Consensus Criteria [5], the 2011 International consensus criteria [6], and the Institute of Medicine criteria [7]) also place emphasis on additional symptom elements such as orthostatic intolerance.
In the absence of curative treatments, behavioural strategies including cognitive behavioural therapy (CBT) [8,9] and graded exercise therapy (GET) [10][11][12] have emerged as the only evidenced-based interventions for the management of CFS, albeit with some recent debate about efficacy [13][14][15].CBT and GET may be delivered alone or in combination, and often incorporate activity pacing [16].Activity pacing describes a group of behavioural strategies commonly used in the management of pain [17][18][19], and fatigue [20].While definitions of activity pacing in the literature are varied, a 2012 consensus document defines activity pacing as "an approach where patients are encouraged to be as active as possible within the limits imposed by the illness … pacing requires the individual to determine a level at which they can function but which does not lead to a marked increase in fatigue and other symptoms for up to five days.Minor and transient symptoms which do not make the person feel unwell may be ignored" [21].
The rationale for activity pacing is to address maladaptive behaviours, which include periods of over-activity triggering worsened symptoms as part of the recognised post-activity exacerbation phenomenon, followed by prolonged periods of rest to recover.This pattern is commonly referred to as a "boom-bust" activity cycle [22]; whereas "activity avoidance" refers to a dominant pattern of inactivity in anticipation of worsened fatigue symptoms [23].Activity pacing generally involves segmenting daily tasks into shorter episodes according to the patient's threshold for a prolonged post-activity exacerbation of symptoms, as well as dispersing larger activities over a week or more to accommodate rest.Activity pacing encompasses all daily physical activities (e.g., household chores) as well as cognitive activities (e.g., reading) as both have the potential to worsen symptoms in patients with CFS.Historically, activity pacing is related to the "energy envelope theory" which encourages patients to maintain activity levels within their perceived energy limits [20].Contemporary approaches to activity pacing encourage patients to first achieve a baseline level of activity where symptom fluctuations are reduced to manageable levels (activity pacing), then to gradually increase activities (both physical and cognitive) over time [24,25]-this approach we refer to here as "graded activity."The distinction between graded activity and GET is subtle, but there are important distinguishing features.Notably, graded activity encompasses both physical and cognitive activities, as well as activities that combine the two (e.g., increased frequency of social engagement outside of the home).In contrast, GET is confined to physical activities only and is typically undertaken in the form of planned and structured physical exercise (e.g., walking or cycling) [10], rather than encompassing the breadth of daily physical activities (e.g., housework) that are considered within activity pacing and graded activity [21].
Despite the wide application of activity pacing in the management of CFS, there is limited evidence for its effectiveness as a stand-alone intervention [12,26,27].The large sample (n ¼ 641) PACE trial [12] is one of very few randomised controlled trials (RCTs) that evaluated the effectiveness of activity pacing as a stand-alone intervention in patients with CFS.In this trial, participants in the "adaptive pacing" arm were advised "not to undertake activities that demanded more than 70% of their perceived energy envelope."The study showed that adaptive pacing was ineffective at reducing fatigue or improving physical function when compared to usual care [12].When included as part of multi-component CFS management programs however, activity pacing has been shown to be effective [28][29][30], but the magnitude of the findings vary.This discrepancy could be due to differences in the diagnostic criteria applied for inclusion, and how activity pacing was defined, explained to participants, and implemented.Specifically, whether participants were advised to remain well within their perceived limits (as in the PACE trial), to undertake activities to within 100% of the perceived energy envelope, and/ or if participants were advised to stabilise their activities, before gradually increasing them over time.That is, whether the advice was "cautious pacing," "more liberal pacing" or also included "graded activity." To date, activity pacing interventions have been inconsistently defined or applied (e.g., cautious pacing versus graded activity) in people with CFS and their effect on outcomes is yet to be synthesised.Therefore, the purpose of this systematic review was to determine the effectiveness of activity pacing interventions on fatigue, physical function, psychological distress, depression, and anxiety in people with CFS.

Protocol and registration
This systematic review was registered with the international prospective register of systematic reviews PROSPERO (CRD42016036087) and was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis PRISMA guidelines.Data and code supporting the analysis are available via the Open Science Framework (https://osf.io/bk7t2/).

Data sources and searches
To identify relevant published studies, Medline, PubMed, EBSCOhost, Web of Science, PEDro, EMBASE, PsycINFO and The Cochrane Library were searched from inception until 13th August 2022.The search strategy, which included terms relating to CFS and activity pacing interventions, were adapted for each database (Supplementary table 1).Additional articles were identified by searching reference lists of included articles.Eligibility criteria were established using the Participants, Intervention, Comparator, and Outcome (PICO) framework [31]

Study selection
Two reviewers independently screened studies according to title, abstract and full text.Discrepancies regarding inclusion of specific studies were resolved by a third reviewer.
A priori planned sub-analyses were performed to contrast interventions that encouraged the maintenance of activity levels (pacing) compared to those which encouraged a gradual increase in activity levels (graded activity).Interventions were classified as "pacing" when the goal of activity pacing intervention was to simply achieve and then maintain manageable and stable activity levels.Interventions were classified as "graded activity" when the goal of the activity pacing intervention was to gradually increase activity over time, typically after stable activity levels and symptoms had been achieved.Two reviewers (S.C. and M.J.) independently categorised the interventions as either "pacing" or "graded activity."These initial categorisations of "pacing" or "graded activity" were then independently re-appraised by two other reviewers (C.S. and B.B.).

Data extraction and quality assessment
Customised data extraction tables were used by two reviewers to independently extract data from each article.Data extracted included patient demographics (e.g., age, sex, duration and severity of illness), intervention details (e.g., modality, method of delivery, frequency and duration of treatment), relevant outcome measures (fatigue, physical function, psychological distress, depression and anxiety) and duration of follow-up.
The Physiotherapy Evidence Database (PEDro) Scale [32] was used to assess the methodological quality of the included studies.The PEDro scale assesses the quality of RCTs in terms of their internal validity and interpretability and consists of 11 criteria.Studies that score 6 or more on the PEDro scale are considered to be high quality [32].

Data synthesis and analysis
Data for the primary outcome of fatigue and secondary outcomes of physical function, psychological distress, depression, and anxiety were extracted into the customised extraction tables.All studies that assessed physical function used the physical function subscale of the Medical Outcomes Survey Short Form-36 questionnaire.For each available relevant outcome, the sample size, mean and standard deviation for each group at baseline, postintervention and follow-up (where applicable) were logged.Data reported as standard error or 95% confidence interval (CI) were converted to standard deviation using standardised equations [33].If a study did not provide outcome data, the authors were contacted to request the data.If no response was received, data was extracted from the published figures of the study using WebPlotDigitiser where applicable [34].This was only done for one study [9].
Meta-analysis was performed in R using the metafor package [35,36].When necessary, data were direction adjusted so that an improvement in fatigue, psychological distress, depression or anxiety was signified by a negative value (i.e., less fatigue) and an improvement in physical function was indicated by a positive value.Meta-analyses for the effect of activity pacing on fatigue, physical function, psychological distress, depression, and anxiety were calculated between the intervention and control groups for post-intervention and, where possible, follow-up.For these analyses, data from studies that included two similar intervention groups (e.g., activity pacing using web diaries and accelerometers compared to activity pacing using paper diaries and step counters) or two control groups (e.g., a no treatment/usual care control group as well as a control group matched for therapistattention) were combined to create a single group which was then compared to the intervention group [33].For studies that included two different intervention groups (e.g., pacing and graded activity), these were kept as independent groups and were compared to a split control group [33].Additional metaanalyses were calculated for the effect of activity pacing compared to control groups matched for therapist attention and for the a priori planned comparison of pacing versus graded activity.For all meta-analyses, heterogeneity was quantified using Cochran's Q, s 2 , the I 2 statistic and 95% prediction intervals.The level of significance was set at p < 0.05.Data for fatigue, psychological distress, depression and anxiety are reported as standardised mean difference (Hedges' g) and 95% CIs and are interpreted as negligible (<0.2), small (0.2-0.49), moderate (0.5-0.79) or large (>0.8) [37].Data for physical function are reported as mean difference and 95% CI.
The studies varied in terms of their duration, frequency, mode of delivery and the type of health professional who delivered the intervention.The interventions were delivered over a period of three to 18 months and the length of follow-up varied between 0 and 15 months.All but one study [12] implemented activity pacing within a CBT approach.Studies described as "self-management packages" [11,47] were regarded as part of CBT since the interventions included key elements of CBT (e.g., sleep/wake cycle management, coping skills, managing stress, cognitions).Only one study included a stand-alone activity pacing intervention as well as a CBT intervention that included activity pacing instruction [12].Five studies [9, 38,39,41,48,49] specifically adapted the treatment to the participant's baseline level of physical activity (i.e., depending on whether participants were classified as "active" or "passive").This distinction was made on the basis of data from both actimetry [9,39,41] and self-report [38,41,48,49].Generally, active participants were those identified as engaging in periods of over-activity resulting in an exacerbation of symptoms and consequently requiring periods of rest and inactivity to recover.These participants were advised to divide their activities more evenly before gradually building them back up.Passive participants were those identified as being activity avoidant and were encouraged to start immediately gradually building up their level of activity.
Regarding the mode of treatment delivery, three studies [38,44,47] used guided self-instruction where no face-to-face contact with a therapist was provided, and instead interaction between the participant and therapist was made through fortnightly emails [38,44] or not at all [47].In two studies [40,46], participants received two face-to-face visits for up to an hour and then engaged in a home-based self-management program.One of these two studies also provided two telephone counselling sessions during treatment [46].In the other eight studies, participants received CBT face-to-face, but the duration of the intervention and number of face-to-face consults varied considerably (e.g., 10 sessions over 5 months [39], 16 sessions over 8 months [9], and 40 to 60 sessions over 18 months [41]).
Regarding other treatments, three studies required participants in the intervention group to cease all other fatigue-related treatments during the intervention, whereas participants in the control groups were free to have other treatments during the study period [9,39,42].Four studies utilised a control group matched for therapist-attention [9, 40,42,43].In two of these studies, the attention-matched control group was included in addition to a usual  Classification of activity pacing: Studies classified as "Pacing"-the goal of activity pacing was to achieve manageable and stable activity levels (for both physical and cognitive activities).Studies classified as "Graded activity"-where the goal of activity pacing intervention was to gradually increase physical and/or cognitive activity over time typically after first achieving manageable and stable activity levels.
care control group [9,40] whereas in the two other studies, the attention-matched control was the only control group [42,43].
Participants allocated to the attention-matched groups received an equivalent amount of therapist contact and support throughout the study period but were not counselled using CBT principles or advised on activity pacing.

Methodological quality assessment
The PEDro assessment revealed that the included studies were of high quality (mean score 6.2 (0.7) out of 10; range 5-7; Supplementary table 2).The most common limitation of the included studies was related to blinding (items 5-7) with subjects, therapists, and assessors never blinded to the treatment group.

Effect of activity pacing on psychological distress
Compared to no treatment/usual care, there was a small effect of activity pacing on reducing psychological distress post-intervention (-0.43 (-0.60 to À 0.26), I 2 ¼ 0%, Figure 4).Subgroup analysis showed that both graded activity (-0.39 (-0.65 to À 0.13), I 2 ¼ 29.2%) and pacing (-0.47 (-0.76 to À 0.19), I 2 ¼ 0%)) had small effects on reducing psychological distress.There was insufficient data to perform a meta-analysis comparing the effect of activity pacing to no treatment/usual care on psychological distress at follow-up.One study found no significant difference betweengroups at follow-up [42].There was also insufficient data to perform meta-analyses for the effect of activity pacing compared to no treatment/usual care post-intervention or at follow-up.
One study found no significant difference between-groups for either time point [42].

Effect of activity pacing on anxiety
Compared to no treatment/usual care, activity pacing did not reduce anxiety immediately post-intervention (-0.21 (-0.43 to 0.01), I 2 ¼ 0%, Figure 6) but it did at follow-up (-0.20 (-0.36 to À 0.04), I 2 ¼ 0%, Supplementary figure 7).Subgroup analysis showed that reductions in anxiety were significant for pacing (-0.26 (-0.50 to À 0.02), I 2 ¼ 0%) but not for graded activity (0.00 (-0.52 to 0.52), though this latter effect was based on a single study [43].There was insufficient data to perform meta-analyses comparing the effect of activity pacing to attention-matched controls on anxiety post-intervention or at follow up, but the two studies found no significant differences between-groups [40,43].

Discussion
This systematic review and meta-analysis showed that activity pacing interventions, often embedded as part of CBT, are effective in reducing fatigue, psychological distress, and depression, and improving physical function in people with CFS.Moreover, activity pacing interventions are more effective when the advice encourages a gradual increase in activity over time ("graded activity") as opposed to simply maintaining manageable activity levels where symptom fluctuations are stabilised ("pacing").The results of our review are similar to those observed in a recent systematic review and meta-analysis (n ¼ 6 studies) in people with CFS and other chronic conditions associated with fatigue (e.g., post-cancer fatigue, fibromyalgia and osteoarthritis), where pacing had a comparable significant effect on improving fatigue (SMD 0.49, 95% CI 0.08 to 0.90) [50].Interestingly, Abonie et al. found a larger effect when activity pacing was combined with GET or CBT (SMD 0.68, 95% CI 0.28 to 1.08) compared to activity pacing alone (SMD 0.27, 95% CI À 0.12 to 0.67) [50], which is similar to our subgroup analysis where graded activity had a larger  effect on improving fatigue compared to pacing alone.The effect of graded activity on improving fatigue observed in our review was also comparable to effect sizes reported in previous metaanalysis of people with CFS using CBT (SMD 0.39, 95% CI À 0.60 to À 0.19) [8] and GET (SMD 0.66, 95% CI 0.31 to 1.01) [10].Thus, the results of our review are broadly consistent with the effect sizes observed for other behavioural interventions in people with CFS.
The magnitude of the intervention effects varied depending on when outcomes were assessed (i.e., post-treatment versus follow-up) and whether participants in the comparator group(s) received an equivalent amount of therapist attention.That is, activity pacing generally had larger effects immediately post-treatment compared to follow-up, whereas the effects were smaller when participants in the activity pacing group were compared to attention-matched controls who received an equivalent amount of therapist attention.This implies that the benefits of activity pacing diminish over time and are partly due to being in a supportive environment.Nonetheless, the effects of activity pacing on reducing fatigue compared to attention-matched controls were still significant post-intervention, demonstrating a benefit of activity pacing interventions on fatigue additional to mere patient support.
Although the PEDro ratings indicated good overall study quality, the findings must still be interpreted cautiously due to the moderate-high level of heterogeneity between studies.Contributors to heterogeneity included variation in the training of therapists who delivered the intervention (only five of the 14 studies specified that training was given), duration of interventions (three to 18 months), the age of patients (adolescents to middle-aged adults), the duration and severity of illness (although most studies included a similar cut-off for severity of illness), diagnostic criteria for inclusion, and the instruments used to assess fatigue.For instance, the largest study included in this review, the PACE trial, used the Oxford criteria [51] which is more inclusive and may give higher prevalence estimates for CFS [52].These factors may have contributed to the varied effect sizes observed between studies.For example, the study showing the greatest improvements in fatigue [41] recruited adolescents (mean age 11.9 years) and the intervention was delivered intensively (40-60 sessions) over the course of 18 months; whereas the majority of the other studies' interventions were 3-6 months and included Figure 5. Forest plot (Hedges' g (95% confidence intervals)) of findings from studies examining the effect of activity pacing on depression in people with chronic fatigue syndrome.Separate meta-analyses were calculated for the effect of pacing and graded activity on depression, as well as for the overall effect of activity pacing on depression.Figure 6.Forest plot (Hedges' g (95% confidence intervals)) of findings from studies examining the effect of activity pacing on anxiety in people with chronic fatigue syndrome.Separate meta-analyses were calculated for the effect of pacing and graded activity on anxiety, as well as for the overall effect of activity pacing on anxiety.
20-50 year old adults.Another study showing large improvements in fatigue was unique in that the intervention was delivered in group sessions and was compared to a waitlist control [45].Several studies also allowed patients to have other fatigue-related treatments during the study period.While this introduces some heterogeneity and may reduce the reproducibility of results if the co-interventions were not well-reported, it is also reflective of real-world clinical practice where patients with CFS may utilise multiple treatments concurrently-some reported to their clinicians, some not.

Different approaches to activity pacing
Only four of the included studies [12,46,48,49] published an intervention protocol [53][54][55], but most studies provided a detailed description of the intervention either within the article or through reference to a preceding study [56][57][58].Some studies even specified the content of each of the individual treatment sessions [40,42,43].Activity pacing advice in the 14 included studies was rarely, if ever, labelled "activity pacing," but on careful examination of the methods and after independent review, it was agreed that they met the definition for activity pacing.In particular, attention was paid to any advice on changing activity levels or activity patterns as well as any education provided on the relationship between activity and CFS symptoms.
Interventions were then classified as "pacing" when the goal was to achieve and maintain stable activity levels, or "graded activity" when the goal was to gradually increase activity over time, typically after stable symptoms and activity levels had been achieved.Cases where studies specified intervention exclusions were also considered in making the distinction between pacing and graded activity.For example, Wiborg et al. [45] specified that "a formal exercise programme was not part of the intervention," which helped to distinguish it as "graded activity" rather than a GET intervention.

Distinguishing activity pacing from other CFS therapies
Despite activity pacing interventions being well accepted in clinical practice [59] and guidelines [60] for the management of CFS, there is a gap in the literature that assesses its effectiveness as a stand-alone intervention.In a clinical setting activity pacing interventions are typically delivered as part of a multi-component program.This is reflected in the included studies where the activity pacing interventions were most commonly embedded in CBT.Indeed, there was only a single study which investigated pacing as a stand-alone intervention [12].It is therefore important to distinguish activity pacing from other therapies used in the management of CFS.Cognitive behavioural therapy aims to address inappropriate/unhelpful cognitions that drive behaviour by asking patients to undertake "behavioural experiments" that obtain corrective information and test the validity of their dysfunctional beliefs [61].In CFS, the initial aim of CBT is to address common comorbidities that can worsen symptoms or impair function (e.g., disrupted sleep, mood disturbances and anxiety) before addressing behaviours that perpetuate fatigue.Through behavioural experimentation, CFS patients are encouraged to "address beliefs such as fear regarding the relative benefits of activity, high selfexpectations, and all-or-nothing thinking" [62].In many cases, this can involve activity pacing (e.g., stopping a task before it is completed to challenge "all-or-nothing" thinking) or GET to overcome activity avoidance.For the purpose of this review, CBT interventions were only included if they specified the activity pacing advice provided to patients.
GET for CFS refers to the establishment of a dose or "threshold" of exercise or physical activity that can be completed without causing a prolonged worsening of symptoms, after which incremental increases in the duration and/or intensity of the activity are made [12].Whereas GET relates specifically to physical activities, activity pacing applies to both physical activities and concentration-demanding tasks (cognitive activities).Importantly, there is a risk that commencing GET without prior activity pacing may be more likely to trigger symptom exacerbations.It is clear that excessive energy expenditure by any means contributes to the problem of post-activity exacerbation, either short-lived or prolonged.Unless daily physical and cognitive activities are managed and stabilised first, then the addition of GET to a weekly routine could lead to increased risk of symptom exacerbation because of the combined effect of the GET with other incidental physical activities and cognitive activities.

Study limitations
There was methodological and clinical heterogeneity among the included studies.Careful consideration was given by the authorship team to define the interventions as activity pacing and its classification as either pacing or graded activity.Additionally, a major limitation of the studies was the lack of data on how well patients adhered to activity pacing (or the intervention more broadly) and whether activities met 100% of the perceived energy envelope or if a more cautious approach was adopted.This reflects the lack of standardised measures available to capture the construct of activity pacing.Finally, we did not formally appraise risk of bias for the included studies using the Cochrane Risk of Bias tool (version 2), [63], although all studies would be deemed to have a high risk of bias because patients (the outcome assessors) were not blinded to the intervention they received and this would likely influence their self-report of fatigue and other secondary outcomes.However, we did appraise the quality of the included studies using the PEDro scale which is commonly used in systematic reviews of non-pharmacological (e.g., physical therapy, behavioural and lifestyle) interventions [64].

Conclusion
Activity pacing interventions are effective in reducing fatigue, psychological distress and depression, and improving physical function in CFS, especially when patients are advised to gradually increase their activities.Future high-quality RCTs directly comparing the two activity pacing approaches ("pacing" and "graded activity") for people with CFS are warranted to help verify the indirect evidence provided in this meta-analysis.Additionally, it is recommended that future studies measure activity pacing via questionnaires and incorporate capturing of intervention adherence.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Figure 1 .
Figure 1.PRISMA flowchart outlining study screening and selection process.

Figure 2 .
Figure2.Forest plot (Hedges' g (95% confidence intervals)) of findings from studies examining the effect of activity pacing on fatigue in people with chronic fatigue syndrome.Separate meta-analyses were calculated for the effect of pacing and graded activity on fatigue, as well as for the overall effect of activity pacing on fatigue.

Figure 3 .
Figure3.Forest plot (mean difference (95% confidence intervals)) of findings from studies examining the effect of activity pacing on physical function in people with chronic fatigue syndrome.Separate meta-analyses were calculated for the effect of pacing and graded activity on physical function, as well as for the overall effect of activity pacing on physical function.

Figure 4 .
Figure 4. Forest plot (Hedges' g (95% confidence intervals)) of findings from studies examining the effect of activity pacing on psychological distress in people with chronic fatigue syndrome.Separate meta-analyses were calculated for the effect of pacing and graded activity on psychological distress, as well as for the overall effect of activity pacing on psychological distress.
Council [No.1043067].CXS is supported by a Cancer Institute New South Wales Early Career Fellowship [2021/ECF1310].
as follows: The control comparators could include treatment as usual/standard care, evidence-based interventions that did not include specific instructions about activity pacing, and/or pharmacological interventions.� Language: Only articles in English were included.� Outcome measures: Studies were included if they used a validated scale to measure fatigue (primary outcome) and/or physical function, psychological distress, depression, or anxiety (secondary outcomes).
activities over time (referred to here as graded activity) (see study selection below for further detail).Interventions were included if they involved activity pacing (as per our above definition) either alone or in combination with other evidence-based interventions (i.e., CBT and/or GET).Studies were excluded if the intervention was not described in sufficient detail to determine whether it met our definition of activity pacing.�Comparator:

Table 1 .
Descriptive characteristics of studies.
Graded activity: "A schedule of planned, consistent and graded activity and rest was agreed on.The initial targets were modest and small enough to be sustained despite fluctuations in symptoms.Rather than being symptom dependent, activity and rest were divided into small, manageable portions spread across the day (for example, three 5-minute walks daily rather than a 45min walk once a week).Patients were encouraged to persevere with their targets and not to reduce them on a bad day or exceed them on a good day.Once a structured schedule was established, graded activity was gradually increased and rest was reduced, step by step, as tolerance developed.Therapist and patient agreed on specific daily targets covering a range of activities (such as walking, reading, visiting friends, or gardening)."Pacing: "Optimal self-management was intended to achieve a healthy balance between mental and physical exertion and period of rest.With information gathered from the Week 1 Web diary, the scheduling of home-based activities, rest/sleep assignments, and cognitive coping skills was individualised for each patient.Walking, if included, was intended as a voluntary leisure activity rather than a fitness regimen.For instance, a relatively low-functioning individual might be assigned a regular sleep/wake schedule and gradual low-effort walking to increase tolerance of physical activity.A higherfunctioning patient might respond more favourably to pacing of activity and low-effort pleasant activities."