Acute Moderate Exercise Improves Working Memory Efficiency In Humans

Abstract

BACKGROUND: Despite evidence indicating that a single bout of exercise can reduce state anxiety and improve cognitive function, (for review Chang et al. 2015, Ensari et al. 2015) few studies have examined the impact of acute exercise on the relationship between anxiety and cognition. The current study explores the modulatory effects of exercise on anxiety and working memory (WM) performance. Moderate exercise, compared to light exercise, was expected to improve WM performance, while decreasing self-reported anxiety and physiological measures (anxiety-potentiated startle) of anxiety. METHODS: Healthy adults (N = 21) of varied physical fitness levels underwent three exercise sessions on a cycle ergometer: one maximal exercise test to assess maximal heart rate (HR) and two subsequent randomized 30-minute exercise sessions at moderate (60-70% HR reserve (HRR)) and low (10-20% HRR) intensities. After exercise, subjects performed a WM task (n-back) under threat of unpredictable electric shock and safe conditions. Anxiety was probed using the acoustic startle reflex and retrospective subjective measures. WM performance (accuracy and reaction time), subjective anxiety, and startle variables were analyzed using 3-way ANOVAs, with Condition (threat, safety), Load (0-back, 1-back, 3-back), and Exercise (light, moderate) as within-subject factors. RESULTS: Participants reported higher subjective feelings of anxiety and had a greater startle response in the threat compared to the safe condition. Moderate intensity exercise resulted in a faster reaction time at a trend level time (F(1,19) = 3.34, p = 0.08) without compromising accuracy. This was most notable during the 3-back WM level (F(2,18) = 3.69, p = 0.05) during both safe and threat contexts. There was no effect of exercise level on subjective anxiety or anxiety-potentiated startle. CONCLUSION: Acute moderate exercise improves cognitive efficiency (shorter reaction times) without compromising accuracy but does not reduce elevated state anxiety. This might be due to an increase in arousal that affects reaction time but not the distribution of working memory resources between cognitive function and anxiety. If this pattern strengthens with the full sample (N=36), we will (1) move to a patient population, and (2) apply this protocol to a neuroimaging study.