Accuracy of Fitbit devices: a systematic review and narrative syntheses of quantitative data (Preprint)
Bibliographic record
Abstract
UNSTRUCTURED OBJECTIVE: To systematically evaluate and report measurement accuracy for FitbitT activity trackers in controlled and free-living settings. DATA SOURCES: Electronic searches using PubMed, Embase, CINAHL and SportsDiscus databases with a supplementary Google Scholar search. ELIGIBILITY: Original research published in English comparing Fitbit to a gold- or research-standard criterion in healthy adults and those living with any health condition or disability. APPRAISAL: Risk of bias was assessed using a modification of the COnsensus-based Standards for the selection of health status Measurement INstruments (COSMIN). SYNTHESES: We explored measurement accuracy for steps, energy expenditure, sleep, time in activity and distance using group percent differences as the common rubric for error comparisons. We conducted descriptive analyses for frequency of accuracy comparisons within a +/-3% error in controlled and +/-10% error in free-living settings and assessed for potential bias of over- or under-estimation. We secondarily explored how variations in body placement, ambulation speed or type of activity influenced accuracy. RESULTS: Sixty-seven studies were included. Consistent evidence indicated that Fitbit devices were likely to meet acceptable accuracy for step count approximately half the time, with a tendency to underestimate steps in controlled-testing and overestimate steps in free-living settings. Findings also suggest a greater tendency to provide accurate measures for steps during normal/self-paced walking with torso placement, during jogging with wrist placement, and during slow/very slow walking with ankle placement in adults with no mobility limitations. Whereas, consistent evidence indicated that Fitbit devices were unlikely to provide accurate measures for EE in any testing condition. Evidence from a limited number of studies also suggest that compared to research-grade accelerometers Fitbit devices may provide similar measures to for time in bed or time sleeping, while likely markedly overestimating time spent in higher intensity activities. LIMITATIONS: Our point estimations for potential bias (mean or median percent error) gives equal weighting to all accuracy comparisons, possibly mis-representing the true point-estimate for measurement bias for some of the testing conditions we examined. CONCLUSION:Fitbit devices are most likely to provide accurate measures of steps in adults with no mobility limitations when the device is worn on the torso while walking at normal or self-paced walking speeds. Whereas, Fitbit devices are unlikely to provide accurate measures of EE. Limited evidence suggests that Fitbit activity trackers may not provide accurate measures for sleep, distance or time spent in activity, however, further accuracy studies are warranted. IMPLICATIONS: Other than for measures of steps in adults with no limitations in mobility, discretion should be used when considering the use of Fitbit devices as an outcome measurement tool in research or to inform health care decisions as there are seemingly a limited number of situations where the device is likely to provide accurate measurement. REGISTRATION: n/a
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.127 | 0.466 |
| Meta-epidemiology (narrow) | 0.002 | 0.002 |
| Meta-epidemiology (broad) | 0.009 | 0.009 |
| Bibliometrics | 0.030 | 0.027 |
| Science and technology studies | 0.001 | 0.003 |
| Scholarly communication | 0.007 | 0.009 |
| Open science | 0.003 | 0.005 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.006 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".