Note that this is calorie counting (i.e. just tracking calories consumed), not calorie restriction (trying to consume fewer calories). The study population was mostly normal weight (average BMI of 23.1).
You can read the full study here: https://sci-hub.se/uptodate/S2212267221007346.pdf
edit: Also: an interesting fact. 1 member of the (100-person) intervention group dropped out of the study because the app "made them feel bad" (p. 3) and was thus excluded from the results. One wonders if that would happen in a larger sample size--if 1% of people have to drop out because calorie counting affects their mood so badly, that'd be an important result.