Children exhibit a developmental advantage in the offline processing of a learned motor sequence

Abstract

Changes in specific behaviors across the lifespan are frequently reported as an inverted-U trajectory. That is, young adults exhibit optimal performance, children are conceptualized as developing systems progressing towards this ideal state, and older adulthood is characterized by performance decrements. However, not all behaviors follow this trajectory, as there are instances in which children outperform young adults. Here, we acquired data from 7-35 and >55 year-old participants and assessed potential developmental advantages in motor sequence learning and memory consolidation. Results revealed no credible evidence for differences in initial learning dynamics among age groups, but 7- to 12-year-old children exhibited smaller sequence-specific learning relative to adolescents, young adults and older adults. Interestingly, children demonstrated the greatest performance gains across the 5 h and 24 h offline periods, reflecting enhanced motor memory consolidation. These results suggest that children exhibit an advantage in the offline processing of recently learned motor sequences.

Fig. 1. Experimental procedures.

a Serial reaction time task (SRTT). A stimulus appeared in 1 of 8 spatial locations and the participant responded with the corresponding key/finger as fast as possible. Stimuli appeared in either a pseudorandom or sequential (i.e., 4-7-3-8-6-2-5-1) manner. b To increase motivation and attention, the task was built around a story in which participants were asked to catch butterflies. c Design of Experiments 1 and 2. Experiment 1 consisted of the first session only (i.e., Day 1 in blue) and Experiment 2 of all three sessions displayed. Session 1 included four blocks (bl) of pseudorandom SRTT (Ran), followed by 20 blocks of sequential SRTT, which were divided into 16 blocks of training and 4 blocks of test after a 1-min rest, and four additional blocks of pseudorandom SRTT. In Experiment 2, session 2 and 3 were completed approximately 5 and 24 h, respectively, after session 1. Both sessions 2 and 3 consisted of 4 blocks of sequential SRTT and 4 blocks of pseudorandom SRTT. Note that the first session of Experiment 2 was constrained between 9 am and 2 pm to give ample time for the 5 h delayed retest. The participants in Experiment 1 were instructed to complete the session anytime between 9 am and 7 pm. Butterfly images in a and b were taken from https://openclipart.org and fall under the Creative Commons Zero 1.0 Public Domain License.

Fig. 2. Initial motor sequence learning.

a Average normalized response time (RT) for both task variants. b Average normalized accuracy for both task variants. For a and b, shaded regions represent standard errors of the mean. c Average learning magnitude for performance speed per age group. Shaded regions represent the kernel density estimate of the data, colored circles depict individual data, open circles represent group medians, and the horizontal lines depict group means. CH children, AD adolescents, YA young adults, OA older adults. *p < 0.05 for pairwise group comparisons. d Learning magnitude as a function of age. Quadratic fit from childhood into young adulthood: Learning magnitude = (-0.0008572*age²) + (0.040²1*age) + (−0.2226). Quadratic fit for older adults: Learning magnitude = (0.00157*age²) + (−0.1996*age) + 6.4675. n = 33, 33, 32 and 32 for groups of children, adolescents, young adults, and older adults, respectively.

Fig. 3. Micro-online (left) and -offline (right) gains.

a, b Depict micro-online and offline gains, respectively, displayed as a function of practice blocks (N children = 31, N adolescents = 33, N young adults = 32, N older adults = 32). c, d Contain violin plots of micro-online and -offline, respectively, gains averaged across blocks. Note that pairwise comparisons in these panels included blocks or rest periods in the statistical models. Shaded regions represent the kernel density estimate of the data, colored circles depict individual data, open circles represent group medians, and the horizontal lines depict group means. CH children, AD adolescents, YA young adults, OA older adults. e, f show averaged micro-online and -offline gains, respectively, as a function of age from childhood into young adulthood and within older adulthood. Lines of fit childhood into young adulthood: micro-online gains = (−0.318*age)^−0.459; micro-offline gains = 0.298*exp(−0.377*age). Lines of fit within older adulthood: micro-online gains = (0.003*age) + (−0.238); micro-offline gains = (227.361*age) + (−1.897). n = 33, 33, 32 and 32 for groups of children, adolescents, young adults, and older adults, respectively.

Fig. 4. Normalized motor performance in Experiment 2.

Average normalized response time (RT; a) and accuracy (b) for the three experimental sessions and two task variants of Experiment 2. Shaded regions represent standard errors of the mean. Corresponding statistical analyses are in Supplementary Table 5. n = 27 in each of the 4 age groups, except n = 26 for young adult data in Session 3 and adolescent data in post-test random of Session 3.

Fig. 5. Macro-offline performance gains.

Sequential offline gains across the 5 h (a) and 24 h (c) offline periods for the four age groups. Shaded regions represent the kernel density estimate of the data, colored circles depict individual data, open circles represent group medians, and the horizontal lines depict group means. CH children, AD adolescents, YA young adults, OA older adults. *p < 0.05 and ~p < 0.10 for pairwise group comparisons. n = 27 in each of the 4 age groups. Five-hour (b) and 24h (d) sequential offline gains are plotted as a function of age. Lines of fit from childhood into young adulthood: 5h gains = (−0.003 * age) + 0.075; 24 h gains = (0.872*age)^−0.682. Lines of fit within older adulthood: 5h gains = (1.814e⁻⁶  * exp(0.256 * age)) + (−1.858e⁻⁶  * exp(0.256 * age)); 24h gains = (−0.006 * age)   0.381. For better visualization of the age-related changes, the scale of the y-axes was set to range from −0.4 to 0.4 and thus the young adult with the outlier 5h sequential gain (visible in a) is not depicted in (b). This individual, however, was included in the fitting procedure. Note that if this individual was excluded from statistical analyses, the pattern of results observed in (a) remains largely similar, with the exception that the pairwise difference between young and older adults becomes statistically significant (i.e., older adults exhibit worse offline consolidation over a 5h wake interval as compared to young adults).