Circadian Phenotype Composition is a Major Predictor of Diurnal Physical Performance in Teams

Abstract

Team performance is a complex phenomenon involving numerous influencing factors including physiology, psychology, and management. Biological rhythms and the impact of circadian phenotype have not been studied for their contribution to this array of factors so far despite our knowledge of the circadian regulation of key physiological processes involved in physical and mental performance. This study involved 216 individuals from 12 different teams who were categorized into circadian phenotypes using the novel RBUB chronometric test. The composition of circadian phenotypes within each team was used to model predicted daily team performance profiles based on physical performance tests. Our results show that the composition of circadian phenotypes within teams is variable and unpredictable. Predicted physical peak performance ranged from 1:52 to 8:59 p.m. with performance levels fluctuating by up to 14.88% over the course of the day. The major predictor for peak performance time in the course of a day in a team is the occurrence of late circadian phenotypes. We conclude that circadian phenotype is a performance indicator in teams that allows new insight and a better understanding of team performance variation in the course of a day as often observed in different groupings of individuals.

Keywords: circadian; mental; performance; physical; sleep.

Figure 1

Relevant sleep/wake parameters validate circadian phenotyping. (A) Average wake-up time on weekdays. (B) Average wake-up time at weekends. (C) Average sleep onset on weekdays. (D) Average sleep onset at weekends. (E) Average sleep duration during weekdays. (F) Average sleep duration weekends. White boxes represent early circadian phenotypes (ECT), light gray boxes represent intermediate circadian phenotypes (ICT), late circadian phenotypes are shown in dark gray (LCT). Boxplots show 25th–75th percentile. Whiskers and outliers are plotted by the Tukey method and the mean is shown within the box as a +. Statistical analysis was carried out using Kruskal–Wallis non-parametric tests combined with Dunn’s multiple comparison test. ns, not significant, **p < 0.01, ***p < 0.001.

Figure 2

Variability of within-team circadian phenotype composition determines predicted diurnal performance curves and peak performance times. Each team (T1–T12) is represented by a pie chart showing the composition of circadian phenotypes and a graph showing predicted performance over the course of a day calculated from performance tests conducted at six different times of day (25). Peak performance times are indicated by the dotted vertical lines and shown in the top left hand corner of each graph. Early circadian phenotypes (ECT) are shown in white, intermediate circadian phenotypes (ICT) in gray and late circadian phenotypes (LCT) in black. (A) T1. (B) T2. (C) T3. (D) T4. (E) T5. (F) T6. (G) T7. (H) T8. (I) T9. (J) T10. (K) T11. (L) T12.

Figure 3

Team performance undergoes significant changes in the course of a day within teams. Boxplots represent predicted team performance levels of T1 (A) to T12 (L). (M) = morning, i.e., 07.00–10.00 h, light gray bars; (A) = afternoon, i.e., 13.00–16.00 h, dark gray bars; (E) = evening, i.e., 19.00–22.00 h, black bars. Tukey boxplots show 25th–75th percentile; mean values are shown within the box as a +. Statistical analysis was carried out using Kruskal–Wallis non-parametric tests combined with Dunn’s multiple comparison test. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 4

Between-team performance differences are most pronounced in the morning and afternoon. (A) Average predicted team performance in the morning ([M] = 07.00 and 10.00 h performance tests), (B) average predicted team performance in the afternoon ([A] = 13.00 and 16.00 h performance tests), and (C) average predicted team performance in the evening ([E] = 19.00 and 22.00 h performance tests). Bars represent mean values + SE of predicted team performance levels based on the composition of circadian phenotypes within each team. Statistical analysis was carried out using Kruskal–Wallis non-parametric tests combined with Dunn’s multiple comparison test. ***p < 0.001. Predicted performance represents percentage of maximum performance attained. Dunn’s multiple comparison test results are shown in Table S1 in Supplementary Material.

Figure 5

Age and gender are negligible predictors of circadian phenotype composition and peak performance times. (A) Male teams average age vs. percentage of ECTs. (B) Male teams average age vs. percentage of ICTs. (C) Male teams average age vs. percentage of LCTs. (D) Female teams average age vs. percentage of ECTs. (E) Female teams average age vs. percentage of ICTs. (F) Female teams average age vs. percentage of LCTs. (G) Age vs. predicted peak performance for all teams. (H) Age vs. predicted peak performance in male teams. (I) Age vs. predicted peak performance in female teams. Statistical analysis was carried out using linear regression analysis; ns, not significant, *p < 0.05. Early circadian phenotypes are labeled as ECT, intermediate circadian phenotypes as ICT and late circadian phenotypes as LCT; male teams, ♂; female teams, ♀.

Figure 6

Subjective mental and physical performance are strongly linked. Frequency plots show both self-reported high mental and physical performance depending on time of day. Subjective physical performance is shown by the solid black line and subjective mental performance by the stippled line. X-axes = time of day in hours. Y-axes = percentage of total sample (%). (A) ECTs. (B) ICTs. (C) LCTs.