Clinical Manifestations

In recent years, many studies have investigated the effects of non-pharmacological interventions (e.g. physical and cognitive training) on cognition in patients with mild cognitive impairment. The SYNERGIC trial showed that combining physical exercise and cognitive training leads to greater gains on the ADAS-Cog than an active control condition, and physical exercise alone. Yet not all participants showed the same cognitive gains. To date, little is known about the cognitive characteristics of responders vs non-responders to non-pharmacological interventions.

To compare baseline cognitive profiles of responders and non-responders to physical exercise training alone or in combination with cognitive training.

Of the 175 initially randomized individuals with MCI in the SYNERGIC trial, 143 completed the 6-month assessment. Regardless of randomization, participants were identified as responders if the change in ADAS-Cog-13 (T6-T0) was negative, indicating improved performances, and non-responders were determined if change was null or positive, indicating decreased performance. Composite z-scores were calculated from the baseline neuropsychological assessment: Global cognition (MoCA), Language (Boston Naming Test, Semantic Verbal fluency), Memory (MoCA delayed recall subscore), Processing Speed (TMT A, Stroop Naming, Stroop Reading), Working Memory (Digit span forward and backward), and Executive Functions (TMT B, Stroop Inhibition and Switching).

Responders (n = 91; 72yrs, 46% female, ADAS-Cog(T0): 15.81) and non-responders (n = 52; 74.2yrs, 54% female, ADAS-Cog(T0): 15.13) were comparable for sex, baseline ADAS-Cog; a trend was observed for older age in non-responders (p = .062). Overall, at baseline, responders had better global cognition (0.14±0.94 vs -0.23±1.07) and executive function performances than non-responders (0.12±0.77 vs-0.21±0.97), ps < .05, but were otherwise comparable. In the intervention arms, the proportion of responders differed significantly: Combined (75%), Physical alone (61%), Active control (44%), (p < .05). In the Combined group, responders had better global cognition (0.38±0.96 Vs -0.20±0.97) and working memory performances (0.25±0.96 Vs -0.28±0.47). In the Physical Alone group, responders and non-responders performed similarly. In the Active Control group, responders had better processing speed performances (0.22±0.31 vs -0.33±0.79) than non-responders.

Responders and non-responders appear to have different baseline cognitive profiles. These results suggest that cognitive and physical training interventions should be tailored and individualized according to baseline cognitive condition.