Eligibility criteria
A. Population of interest
HIV positive male or female adults, 18 years and older, at all stages of infection (CD4, viral loads), with or without comorbidities in any part of the world.
B. Intervention
Prescribed exercise performed at home or in the community of at least four-week duration, with minimal or no supervision. The program must include either an aerobic component or resistance component.
We will consider “prescribed exercise” to be a structured physical activity plan made available to participants with specified type, intensity, duration, and frequency of activity, accompanied by some form of monitoring (such as an exercise diary).
“Community” will refer to local public settings and outdoor spaces that are freely accessible, such as parks and public recreational facilities (see Table 1).
Table 1
CRITERIA AND DEFINITION OF HOME/COMMUNITY EXERCISE
Criterion
|
HOME/COMMUNITY
MINIMALLY SUPERVISED
|
CENTRE-BASED SUPERVISED
|
Supervision
|
- Mostly self-managed.
- Instruction offered during less than half of exercise time
|
- Mostly supervised.
- Instruction offered during half or more of exercise time
|
Location
|
Home environment or local neighbourhood
|
Within a specialised exercise facility
|
Equipment
|
Present or absent
|
Specialised equipment designed for exercise
|
Accessibility/cost
|
Accessible to the public at little or no cost
|
Not routinely open to the public
|
Examples
|
Within home, public outdoor spaces such as parks, community centres, public recreational facilities
|
hospital clinics, exercise laboratories, private practice, physiotherapy clinics
|
We will exclude studies of interventions in a detention or prison facility, as well as those where the setting is not clearly reported and cannot be sourced from the authors.
“Minimal supervision” will be regarded as professional input (in-person or virtual) in the form of instruction during exercise sessions for less than half the total duration of exercise time. Remote observation through physical activity monitors or exercise logbooks will not be considered supervision. Professional support restricted to safety or compliance checks, reminders, counselling, or motivation will also not be considered supervision.
Aerobic exercise will be defined as any physical activity performed with the aim of strengthening the heart and lungs, that results in increased breath and heart rates. This will include (but not limited to): jogging, running, walking, swimming, stair climbing, stepping, rowing, jump rope, dancing, and cycling.
Resistance exercise will be defined as any physical activity performed with the aim of strengthening muscle, utilising muscle contractions against resistance aided by weights (free weights, weight stations amongst others) or unaided (resistance provided by body’s own weight).
C. Comparators
We will compare home/community exercise outcomes to “before exercise” in single-group design or to control arms with no exposure/intervention, standard care, other types of exercise programs such as centre-based exercise (see Table 1), or other therapeutic modalities.
D. Outcomes
Primary
Our primary outcomes shall include the following measures (classification was adapted from previous work (27)):
i. Total body adiposity
-
total fat mass
-
percent body fat
-
body mass index ((BMI = body weight in kilograms/(height in metres)²)
-
fat mass index ((FMI = total fat mass in kilograms/(height in metres)²) (28)
-
body adiposity index (BAI = ((hip circumference in centimetres)/((height in metres)1.5)–18)) (29)
ii. Body adipose distribution
iii. Regional adiposity
-
skinfold thickness measurements
-
visceral adipose tissue (VAT) volumes/mass
-
regional fat percentages
-
subcutaneous fat volumes/mass.
In this study weight will not be considered a primary outcome measure of adiposity because of its well-recognized inability to account for body build and its susceptibility to changes in lean mass (27).
Secondary: In addition to the above primary outcomes, the following will be evaluated: changes in weight, metabolic parameters (fasting glucose, lipids), virologic outcomes (viral load, CD4+ T cells), physical activity levels (self-report or device measured), adherence to exercise program, muscle/lean mass, muscle strength, cardiorespiratory fitness, quality of life measures, psychological health measures, neurocognitive function, adverse events related to exercise.
E. Study Characteristics
Randomized control trials, non-randomized control trials, and uncontrolled single arm interventional studies published before 31st January 2021 will be included.
Information sources
We will search general databases (PubMed, PubMed Central, Embase and Scopus), databases specific to allied health (Cumulative Index to Nursing and Allied Health Literature (CINAHL), Physiotherapy Evidence Database (PEDro) and SPORTDiscus) and trial registries (Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews and Clinicaltrials.gov).
Databases will be searched from inception up to 31st January 2021 for peer-reviewed articles and peer-reviewed conference papers published or accepted for publication. Reference lists of the included studies and relevant systematic reviews will be manually back searched to identify any relevant articles. In addition, forward citation searching of included studies will be undertaken in Google Scholar and Scopus.
When full texts are unavailable or in the case of missing relevant data, the authors will be approached via email to obtain this information. No language restrictions shall be applied to our search. The services of a professional translator will be sought for articles that are not available in the English language.
Search strategy and study selection
The search strategy was designed by review authors in consultation with a professional librarian and pretested prior to the formal search. Key terms pertaining to “exercise”, “HIV”, “adiposity” and “interventional study” were included, see details in Table 2. Records retrieved will be exported to and managed in the EndNote referencing software. After elimination of duplicates, screening will be performed independently by two review authors. Studies with irrelevant titles will be excluded. Abstracts of the remaining studies will then be examined using pre-established inclusion checklist. Where additional information is required to determine eligibility, the reviewers will contact the authors concerned via email. Where there are several publications of the same study (such as a protocol paper cited in the principal article), the publications will be collated and considered together. When abstracts are deemed relevant or where eligibility is unclear, full texts will be obtained. Full texts will be assessed independently by two review authors. Any disagreements will be resolved through consensus or by the decision of a third independent reviewer.
Table 2
Field limits
|
#1
title and abstract
|
#4
title and abstract
|
#9
full text
|
#12
title and abstract
|
Keyword query
|
"human immunodeficiency" OR "human immune deficiency" OR HIV OR "acquired immune deficiency" OR "acquired immunodeficiency" OR "aquired immunodeficiency" OR "aquired immune deficiency" OR “HIV/AIDS”
|
exercis* OR sport* OR workout OR physiotherapy OR kinesiotherapy OR walking OR jogging OR running OR swimming OR bicycling OR cycling OR "weight-lifting" OR dancing OR “resistance bands” OR “home gym*”
|
"visceral fat" OR "visceral adipos*" OR "abdominal adipos*" OR "abdominal fat" OR "truncal adipos*" OR "truncal fat" OR “trunk fat” OR “trunkal fat” OR “trunk adipos*” OR "body fat" OR "limb fat" OR "leg adipos*" OR "waist circumference" OR "abdominal circumference" OR "sagittal abdominal diameter" OR "waist girth" OR waist-to-hip-ratio OR waist-hip-ratio OR waist-to-height-ratio OR waist-height-ratio OR waist-to-thigh-ratio OR waist-thigh-ratio OR skinfold OR body-mass-index OR BMI OR “regional fat” OR “regional adipos*” OR “fat mass index” OR (Quetelet* AND index)
|
Intervention* OR program* OR trial
|
#5
title and abstract
|
(physical OR aerobic OR resistance) AND (fitness OR therapy OR training OR rehabilitation)
|
#6
title and abstract
|
(weight OR endurance OR strength OR circuit) AND training
|
|
#2
|
#7
|
#10
|
#13
|
Index term query
|
"HIV"[Mesh] OR "HIV Long-Term Survivors"[Mesh] OR "HIV Infections"[Mesh]
|
"Exercise"[Mesh] OR "Exercise Therapy"[Mesh] OR "Sports"[Mesh] OR "Exercise Movement Techniques"[Mesh]
|
"Body Composition"[Mesh] OR "Body Weights and Measures"[Mesh] OR "Adipose Tissue, White"[Mesh]
|
"Clinical Trials as Topic"[Mesh] OR "Clinical Trial"[Publication Type]
|
Combined
|
#3 = #1 OR #2
|
#8 = #4 OR #5 OR #6 OR #7
|
#11 = #9 OR #10
|
#14 = #12 OR #13
|
Final string
|
#3 AND #8 AND #11 AND #14
|
Data extraction and management
A pre-piloted data extraction form developed in Microsoft Excel by the authors will be used to extract information (see form attached (Additional file 2)). The primary outcome characteristics will be extracted by two independent reviewers. Study characteristics and secondary outcome measures will be extracted by one reviewer and randomly cross-checked by a second. Disputes between the two extractors will be resolved by consensus or the decision of a third reviewer. We will attempt to obtain missing information from supplementary files if available (such as public data repositories) or contact the authors via email.
Study characteristics extracted will include study date, study aim/objectives, setting of the study, description of participants, study design including recruitment procedure. Details of the home/community exercise intervention and the outcomes of interest including adherence will equally be extracted. In addition, the details of the descriptions of comparators for all studies will be extracted (Additional file 2). These general characteristics will be summarized and presented in a table in the results.
Quality assessment
Two reviewers will independently assess the quality of included studies. Comprehensive reporting and risk of bias in individual studies will be assessed using the Cochrane Effective Practice and Organisation of Care (EPOC) Risk of bias tool for controlled and uncontrolled before-after studies (30). The EPOC Risk of Bias tool for controlled studies assesses the quality of both randomized and non-randomized trials. The scale consists of nine items ranging from participant allocation to outcome reporting. Each criterion is rated as “unclear risk of bias” (plausible bias unlikely to seriously alter the results), “low risk of bias” (plausible bias that raise some doubt about the results), or “high risk “(plausible bias that seriously weakens confidence in the results) (30). The EPOC tool for uncontrolled before-after studies is a seven-item tool with criteria that ranges from intervention independence to selective outcome reporting and has similar grading to the EPOC tool for controlled trials (30). Assessments will be performed for each of the primary outcomes evaluated in the study separately. We will perform an overall assessment of risk per outcome and per study and summarize in a risk of bias table.
Data analysis, assessment of heterogeneity and publication bias
Exercise volume will be translated to metabolic equivalent (MET) hours per week based on standard definitions in the compendium of physical activities (31), to ease comparison across studies. In studies with multiple intervention groups, we shall report outcomes for all groups relative to the outcome in the minimally supervised home exercise group. Where a study reports outcome at multiple time points, we will consider the timepoints that immediately precedes and the timepoint that immediately follows the intervention in our analysis.
Continuous outcome measures will be reported as means or standardized mean differences where there are different scales for the same outcome. Where possible, missing outcomes will be computed from other reported statistics such as percentage body fat from skinfold thickness using validated formulae (32).
Where there are at least two studies with common or similar outcome measures (reported or computed) we will undertake a quantitative synthesis. We anticipate some heterogeneity in participant characteristics, study quality, sample sizes, and type and volume of exercise, hence we will assume and conduct a random effects meta-analysis. The Cochran’s Q test will be used to assess for heterogeneity and the I² statistic to quantify it. I² of 0–30% will be considered minimal, 30–55% moderate, 55–75% substantial heterogeneity and 75–100% considerable heterogeneity (33, 34). If we find that have considerable heterogeneity across studies, we will perform a narrative synthesis.
In the event where authors report medians and interquartile ranges we will use this to compute corresponding means and standard deviations as described in previous literature (35) and perform subgroup analysis to evaluate the impact this has on effect estimates. We will perform subgroup analysis (or a narrative synthesis if insufficient studies) to compare the effect size for primary outcomes across studies with “minimal supervision” versus “no supervision”, “technology-assisted delivery” versus “no technology” as well as by type and intensity of exercise.
Sensitivity analysis will be carried out to restrict analysis to published studies, studies with low risk of bias, studies without co-interventions and studies that performed comparisons with “no exercise”; to assess the impact of these on the effect sizes. Publication bias will be evaluated using funnel plots and tested using the Egger’s regression test.
Meta-analysis will be performed with the Comprehensive Meta-Analysis Software (CMA) software. Our findings will be summarized by considering the five Grading of Recommendations Assessment, Development and Evaluation (GRADE) criteria (study limitations, inconsistency of effect, imprecision, indirectness of evidence, and publication bias) as outlined previously (36) to determine the strength and quality of evidence for each outcome.