Table 4 Prespecification of outcomes measured with BMDs in the included trials (n = 464 outcomes). For each included trial, we looked for the corresponding entry in a public clinical trial registry (e.g., ClinicalTrials.gov) by looking for the trial registration number reported in the articles on January 2020 (and June 2020 for articles added during peer-review)

Impossible because the trial is not registered*

–

Outcome definition is not clear enough for comparison

In the trial [22]:

Published outcome: “MVPA was calculated using 3 axes based on 60 s epochs. Freedson-VM cut-off points were used to distinguish between light, moderate and vigorous PA.”

Outcome registered: “Objective physical activity (accelerometer data)”

Similar outcome definitions reported in the 2 sources

In the trial [23]:

Published outcomes: “time with glucose concentration […] hyperglycemic (> 10.0 mmol/L and > 16.7 mmol/L)”

Outcomes registered: “Time spent above target glucose (10.0 mmol/l) (180 mg/dl) [Time Frame: 12-week intervention phase]” and

“The time with glucose levels in the significant hyperglycemia (glucose levels > 16.7 mmol/l) (300 mg/dl) [Time Frame: 12 week intervention phase]”

Outcome definition was modified in the published article

In the trial [24], time frames during which BMD outcome data analyzed were modified; published outcome: “M-value (24 h […], 08.00–12.00 h […], 12.00–24.00 h […], 00.00–06.00 h […])”

Outcomes registered: “M-value (24:00, 8:00–12:00, 12:00–24:00, 0:00–8:00)”

Outcome in the published article was not registered

Among outcomes registered for the trial [25], only “% time within target range (3.9–10 mmol/L)” and “Incidence rate of hypoglycemic events” were registered. Published article also reported:

“% time > 10 mmol/L”; “% time > 15 mmol/L”

“% time > 20 mmol/L” and “Mean (SD) glucose values”