The increment is too close to the noise level, so to reach any conclusion it's probably better to wait until there is better study with a clear result. IANAMD.
> ADAS-cog results run on a scale from 0 to 70, with higher scores suggesting greater impairment. The usual-care group had an average score of 23.8, compared with 25.2 for the exercise group (adjusted estimate -1.4, 95% confidence interval [CI] -2.6 to -0.2).
They have to repeat the study with more patients to be able to reduce the error estimation in the average. (Or run study for a longer time so the increment or decrement effect accumulates and it's easier to measure.) But if the first study had no interesting result it may be more difficult to get the foundlings for another similar study.
> ADAS-cog results run on a scale from 0 to 70, with higher scores suggesting greater impairment. The usual-care group had an average score of 23.8, compared with 25.2 for the exercise group (adjusted estimate -1.4, 95% confidence interval [CI] -2.6 to -0.2).
They have to repeat the study with more patients to be able to reduce the error estimation in the average. (Or run study for a longer time so the increment or decrement effect accumulates and it's easier to measure.) But if the first study had no interesting result it may be more difficult to get the foundlings for another similar study.