REF-1756 · Klier, Masur & Düking, 2026 · Frontiers in Sports and Active Living, · Paper · PubMed

Sources on consumer sleep trackers in active people most often named missing validation and weak sleep-stage evidence as weaknesses and misleading training or recovery advice as the main threat; the authors advise reading multi-week trends beside how the athlete feels: a systematic SWOT synthesis of 21 studies, 11 of them narrative reviews, with no accuracy numbers of its own

Main caveat: A qualitative synthesis: themes are counted by how often studies mention them, which reflects what authors write about, not how large any effect or error is.
limited, mixed evidenceMeasured on recreational to highly trained athletesRead in full

Klier, Masur and Düking searched three databases for studies of consumer sleep technology (wearables, smartphones and apps) in healthy active adults and athletes, and sorted what the 21 included studies said into strengths, weaknesses, opportunities and threats. They extracted 9 strengths, 12 weaknesses, 8 opportunities and 8 threats (37 aspects, 4 counted twice).

The most often named strength was the growing availability of devices and sensors (10 studies); the most often named weakness was that most devices lack validation studies (10 studies); motion-based devices were described as over-reading sleep and under-reading wake (8 studies), and sleep stages as the least supported output (7 studies).

The leading opportunity was regular, longitudinal monitoring in real settings (13 studies); the leading threat was misleading training or recovery advice from poor sleep data (7 studies), followed by orthosomnia, an unhealthy pursuit of perfect tracker numbers (5 studies).

From these, the authors write twelve best-practice recommendations: favour sleep duration and timing over proprietary scores, read rolling averages over 14, 21 or 28 nights, and interpret device data together with subjective sleep and training context.

Participants

21 studies of healthy, physically active people or athletes of any sport and sex, aged 18-60; clinical and ill populations excluded.
Sample sizes, sexes and competition levels per study are in a supplementary table that was not available.

Protocol

Search: PubMed, Scopus and EBSCO, 14 July 2025, no filters, terms built with a librarian; 2,470 records, 96 full texts assessed, 21 included.
Titles and abstracts sifted by one reviewer; full texts by two, with a third resolving disagreements.
Risk of bias: adapted Cochrane ROB-ME for the systematic review, RoB 2 for the RCT and the study within an RCT, ROBINS-I for the rest; not applied to the narrative reviews and the survey.
Synthesis: three-stage thematic coding into SWOT categories, ordered by how many studies named each theme.

Key Results

SWOT categoryNumber extractedMost often named (studies naming it)
Strengths9More devices and sensors available (10); sleep-wake detection better when several signals are combined, most often heart rate, HRV, breathing and motion (9); practical and cheap (9)
Weaknesses12Most devices not validated (10); unknown, changing manufacturer algorithms (8); sleep over-read and wake under-read by motion-based devices (8); weakest evidence for sleep stages (7)
Opportunities8Regular longitudinal monitoring in different places (13); a fuller picture of health and performance (12); early detection of health problems or overtraining (8); combining objective with subjective data (8)
Threats8Misleading recovery and training advice from poor data (7); orthosomnia (5); false ideas of healthy sleep (4); over-reading single data points (4)
Risk of bias: the meta-analysis, the RCT and the laboratory validation studies were judged low risk; all other studies had some or serious concerns.
One study named reduced accuracy of optical (PPG) sensors on darker skin and tattoos.
Recommendations (the authors' practice): pick devices and parameters with evidence of validity, such as sleep duration and timing; treat proprietary scores cautiously; do not over-read single nights; use rolling averages of 14, 21 or 28 nights; combine with subjective sleep, recovery, fatigue and mood; add context (training load, nutrition, travel, stress, illness); keep firmware updated, accepting that updates may change the data; let wearable data support, not replace, body awareness.

Practical Application

Use a sleep tracker for the multi-week trend in duration and timing, read with how the athlete feels, and do not let one night or a stage number change training. This is the authors' recommendation from a thematic synthesis of 21 sources, 11 of them narrative reviews; it carries no new accuracy data and was not tested.
This agrees with [SYNTH-91] and [SYNTH-07]: sleep and its stages are estimated by an algorithm from movement and heart rate, not measured, so they are context. The sizes of the error are in the agreement studies, not here (REF-1252, REF-364, REF-1698).
Name orthosomnia when an athlete worries over tracker numbers: chasing perfect sleep scores was one of the threats the sources named most often; the authors' answer is to read data as trends and let it support, not replace, body awareness.
Keep firmware updated, knowing an update can change the numbers (the authors' recommendation). Inference, not tested here: a step change in sleep metrics right after an update is more likely the algorithm than the athlete.

Limitations

A qualitative synthesis: themes are counted by how often studies mention them, which reflects what authors write about, not how large any effect or error is; no device accuracy numbers.
11 of 21 sources are narrative reviews; apart from the meta-analysis, the RCT and the laboratory validation studies (low risk), the rated studies had some or serious risk-of-bias concerns, and the narrative reviews and the survey were not rated.
Title and abstract selection, data extraction and risk-of-bias rating were each done by one author.
No distinction between devices, technologies or wear sites; the authors flag that findings from elite samples may not transfer to recreational users.
The recommendations (rolling averages of 14-28 nights, at least 72 h of monitoring drawn from actigraphy guidance) are the authors' practice, not tested outcomes.
Funding: open-access fees from the University of the Bundeswehr Munich; no commercial conflicts declared.
How this entry was checked and changed
Why this level

a systematic search with risk of bias read, but the synthesis counts how often themes were mentioned rather than measuring anything, more than half the sources are narrative reviews, most studies had some or serious concerns of bias, extraction and bias rating were done by one author, and the best-practice recommendations are the authors' own practice