Effects of Afternoon Napping, Caffeine and Recovery on Evening Athletic Performance (NAP-CAF-REC)

This is a mechanistic, randomised, double-blind, placebo-controlled crossover trial designed to examine the isolated and combined effects of afternoon napping, caffeine ingestion, and a personalised active recovery protocol on evening cognitive and physical performance in male and female athletes stratified by chronotype.

STUDY DESIGN:

The study employs a Latin square counterbalanced crossover design with five experimental conditions separated by washout periods of at least 72 hours to eliminate carryover effects of caffeine and sleep manipulation:

1. PLA: Placebo capsule at 18:00, no nap opportunity
2. NAP: 90-min nap opportunity (13:00-14:30) + placebo capsule at 18:00
3. CAF: Caffeine ingestion (5 mg/kg anhydrous caffeine) at 18:00, no nap opportunity
4. NAP+CAF: 90-min nap + caffeine (5 mg/kg) at 18:00
5. NAP+CAF+REC: 90-min nap + caffeine (5 mg/kg) at 18:00 + 15-min personalised active recovery protocol (dynamic stretching + carbohydrate-protein snack) from 18:45-19:00

Condition allocation was determined using a Latin square design separately for each combination of sex and chronotype (morning-type males, evening-type males, morning-type females, evening-type females). Condition allocation was concealed in sequentially numbered, opaque, sealed envelopes prepared by an independent researcher not involved in data collection. Both participants and the experimenters responsible for testing were blinded to condition assignment. Capsules were prepared by an independent pharmacist and were identical in appearance, mass, colour, and odour. The researcher administering the capsules and the one preparing the EEG and biomarker materials were different from the testing team to ensure blinding.

PARTICIPANTS:

Sixty elite university athletes (30 males, 30 females; age 20.6 ± 1.5 years) were recruited from the Higher Institute of Sport and Physical Education of Sfax, Tunisia, and from regional sport clubs. Participants were classified as definite morning-type (MEQ score 59-86) or definite evening-type (MEQ score 16-41) using the Morningness-Eveningness Questionnaire, with exclusion of intermediate types (score 42-58) to maximise contrast between chronotype groups. Female participants were tested during the early follicular phase of the menstrual cycle (days 2-5) to control for hormonal variability.

EXPERIMENTAL PROCEDURES:

Participants arrived at the laboratory at 19:00 on the evening before each trial for standardisation. They consumed a standardised dinner (35 kcal/kg, 55% carbohydrates, 15% protein, 30% fat) at 20:00, followed by two hours of low-demand leisure activities. Lights were turned off at 22:15 after a 10-min adaptation period in a darkened, sound-attenuated room, and wake-up time was fixed at 06:30. Sleep duration was confirmed by wrist actigraphy and sleep diaries; only nights with total sleep time between 7.5 and 8.5 hours were accepted. A standardised breakfast (20 kcal/kg, 65% carbohydrates, 15% protein, 20% fat) was provided at 07:00.

Throughout the morning, participants remained in the laboratory and engaged exclusively in sedentary activities (reading, quiet games, internet browsing) to avoid unplanned physical exertion. At 12:00, they consumed a standardised isocaloric lunch (35 kcal/kg, 55% carbohydrates, 15% protein, 30% fat).

In nap conditions (NAP, NAP+CAF, NAP+CAF+REC), participants entered the nap room at 12:50 and were fitted with a portable EEG headband (Dreem 3, Dreem, Paris, France) for objective sleep recording. Earplugs and eye masks were provided, and the nap opportunity commenced at 13:00 for a duration of 90 minutes (until 14:30). The room was maintained quiet (ambient noise < 30 dB), dimly lit (< 10 lux), and at a constant temperature (22 ± 1°C). Awakening was by a silent vibrating alarm placed under the pillow to avoid startling. For the no-nap conditions (PLA and CAF), participants remained in a quiet, dimly lit room and engaged in passive activities (reading, quiet games) during the same 13:00-14:30 time window without lying down.

After the nap or rest period, participants remained in the laboratory and continued with passive, seated activities until 18:00. At 18:00, participants ingested a single opaque capsule containing either anhydrous caffeine powder (5 mg/kg of body mass) or a visually identical placebo (microcrystalline cellulose-starch mixture). All capsules were swallowed with 200 mL of still water under direct supervision to ensure compliance.

In the NAP+CAF+REC condition only, participants underwent a 15-min personalised active recovery protocol from 18:45 to 19:00, consisting of five dynamic stretching exercises targeting the lower limb musculature (leg swings, walking lunges, high knees, butt kicks), each performed for 10 repetitions per leg, followed by a carbohydrate-protein snack (30 g total, 2:1 carbohydrate-to-protein ratio: 20 g maltodextrin + 10 g whey isolate mixed with 200 mL water, 120 kcal).

At 19:00, all participants completed a standardised 10-min warm-up consisting of light jogging (5 min at a self-selected pace), five submaximal vertical jumps, and dynamic stretching. This was followed by a 5-min passive rest period, then the testing battery performed in fixed order (19:15-20:00).

STATISTICAL ANALYSIS:

Linear mixed-effects models will be fitted with fixed effects for sex, chronotype, condition, timepoint (where applicable), and their interactions, with random intercepts for subject. Degrees of freedom and p-values will be estimated using the Kenward-Roger approximation. Model fit will be assessed using conditional R². Post-hoc pairwise comparisons will use Bonferroni correction. Significance level: p ≤ 0.01 (adjusted for multiple comparisons across five conditions). Effect sizes: partial eta-squared (ηp²) for ANOVA effects and Cohen's d for pairwise comparisons. Sample size calculation (G*Power 3.1.9.7): based on repeated agility performance (ηp² = 0.14, effect size f = 0.40, α = 0.01, power = 0.95, two groups sex, two groups chronotype, five repeated measures, correlation between repeated measures = 0.5, nonsphericity correction ε = 1), minimum n = 56. With 10% dropout allowance, n = 60 will be recruited.