Effects of citrulline on endurance performance in young healthy adults: a systematic review and meta-analysis

Figures & data

Table 1. ~TC~

Study (first author, year)	Source	Study design	Location	Sample size (Dropouts)	Age, years (mean ± SD)	Sex	Training level	Supplement dosage and form	Dosage regime (loading -days/bolus: timing before exercise)	Comparators	Type of endurance exercise: performance protocol	Outcome of interest	Number of participants included in analyses	Study sponsorship (conflict of interest)	Study findings
Bailey et al. 2015	Journal of Applied Physiology	Randomized, double-blind, placebo-controlled, crossover	Exeter, United Kingdom	10	19 ± 1	M	Recreationally active	6g L-Citrulline	Loading − 7 days: 90 minutes before exercise	6g L-Arginine, Placebo	Cycling: severe intensity on day 7	Time-to-exhaustion	NS	ND	↑ TTE
Bailey et al. 2016	Nitric Oxide	Randomized, double-blind, placebo-controlled, crossover	Exeter, United Kingdom	8	22 ± 2	M	Recreationally active	300ml WJ (approximately 3.4g of L-Citrulline)	Loading − 16 days: 75 minutes	Placebo	Cycling: Moderate and severe intensity on day 14	Time-to-exhaustion	NS	ND	↔ TTE
Cutrufello et al. 2014	Journal of Sports Sciences	Randomized, double-blind, placebo-controlled, crossover	Louisville, USA	22: 11 male/11 female	20.6 ± 1.2 (male); 21.0 ± 1.3 (female)	Mix	Recreationally trained	6g L-Citrulline	Bolus: 10 participants 60 minutes and 12 participants 120 minutes	710ml WJ (approximately 1g of L-Citrulline), Placebo	Running: graded treadmill test	Time-to-exhaustion	22	ND	↔ TTE
Gills et al. 2020	European Journal of Sport Science	Randomized, double-blind, placebo-controlled, crossover	NS	28 [4]	20.9 ± 2.8	M	Recreationally trained (endurance-based exercise twice a week for >1 year)	8g CM	Bolus: 60 minutes	Placebo	Cycling: at 90% of previously recorded VO2peak above 40 RPM pedalling cadence	Time-to-exhaustion	28	ND	↔ TTE
Hickner et al. 2006	Medicine & Science in Sports & Exercise	Randomized, double-blind, placebo-controlled, crossover	Greenville, USA,	17	NA (range 18–40 years old)	Mix	Physically active	3g or 9g L-Citrulline	Bolus: last meal before exercise	Placebo	Running: graded treadmill test	Time-to-exhaustion	17	ND	↓ TTE
Martínez- Sánchez et al. 2017	Food & Nutrition Research	Randomized, double-blind, placebo-controlled, crossover	Cartagena, Spain	21	35.3 ± 11.4	M	Amateur runners	WJ + L-Citrulline (approximately 6.91g L-Citrulline)	Bolus: 60 minutes	Placebo	Running: half-marathon time trial	Time to completion	NS	Ministerio de Economía y Competitividad; Fashion Group Association (ND)	↔ TTC
Shanely et al. 2016	Nutrients	Randomized controlled, crossover	North Carolina, USA	20	48.5 ± 2.3	M	Competitive cyclists	980mL WJ (approximately 1.47g L-Citrulline)	Loading − 14 days: ingested during time trial	No treatment plus 6% carbohydrate beverage during trial	Cycling: 75km time trial	Time to completion	NS	National Watermelon Promotion Board (One author is a science advisor to this Board)	↔ TTC
Stanelle et al. 2020	The Journal of Strength and Conditioning	Randomized, double-blind, placebo-controlled, crossover	Texas, USA	10	24 ± 3	M	Trained triathletes or cyclists	6g L-Citrulline	Loading − 7 days: 120 minutes	Placebo	Cycling: 40km time trial	Time to completion	9	The Sydney & J.L. Huffines Institute for Sports Medicine & Human Performance (ND)	↔ TTC
Suzuki et al. 2016	Journal of the International Society of Sports Nutrition	Double-blind, placebo-controlled, crossover	Osaka, Japan	22 [3]	29 ± 8.4	M	Trained	2.4g L-Citrulline	Loading − 8 days: 60 minutes	Placebo	Cycling: 4km time trial	Time to completion	22	KYOWA HAKKO BIO CO (Two authors are employees)	↓ TTC

SD standard deviation, NS not stated, USA United States of America, CM citrulline malate, M male, ND not declared.

↔ no significant difference (p>0.05) compared to placebo, ↑ significantly greater (p<0.05) than placebo, ↓ significantly less (p<0.05) than placebo.

Table 2. Results of risk of bias assessment using the ROB 2 tool.

Bias domain	Criteria within domain	Bailey et al. 2015	Bailey et al. 2016	Cutrufello et al. 2014	Gills et al. 2020	Hickner et al. 2006	Martínez- Sánchez et al. 2017	Shanely et al. 2016	Stanelle et al. 2020	Suzuki et al. 2017
Bias arising from the randomization process	Allocation sequence randomised?	PY	PY	PY	PY	PY	PY	PN	Y	Y
	Allocation sequence concealed?	PY	PY	PY	Y	PY	PY	NI	Y	Y
	Baseline imbalances suggesting problematic randomization?	NI	NI	NI	NI	N	NI	NI	N	NI
	Roughly equal proportion of participants allocated to each group?	NI	NI	NI	NI	Y	NI	NI	Y	NI
	Period effects included in analysis?	NI	NI	NI	NI	NA	NI	NI	NA	NI
	Overall risk of bias	Some concerns	Someconcerns	Some concerns	Someconcerns	Low risk	Some concerns	High risk	Low risk	Someconcerns
Bias due to deviations from intended interventions	Participants aware of assigned intervention during each trial period?	N	N	N	N	N	N	PY	N	N
	Trial personnel aware of assigned intervention during each trial period?	N	N	N	N	N	N	PY	N	N
	Deviations from intended interventions beyond usual practice?	NA	NA	NA	NA	NA	NA	PN	NA	NA
	Were deviations unbalanced between the two interventions?	NA	NA	NA	NA	NA	NA	PN	NA	NA
	Sufficient time for carry-over effects to disappear?	Y	Y	Y	Y	Y	Y	Y	Y	Y
	Overall risk of bias	Low risk	Low risk	Low risk	Low risk	Low risk	Low risk	Low risk	Low risk	Low risk
Bias due to missing outcome data	Outcome data available for all participants randomised?	PY	PY	Y	N	Y	PY	Y	Y	Y
	Are missing outcome data similar across interventions?	NA	NA	NA	Y	NA	NA	NA	NA	NA
	Results robust to presence of missing outcome data?	NA	NA	NA	Y	NA	NA	NA	NA	NA
	Overall risk of bias	Low risk	Low risk	Low risk	Low risk	Low risk	Low risk	Low risk	Low risk	Low risk
Bias in measurement of the outcome	Were outcome assessors aware of intervention allocation?	PN	PN	PN	PN	PN	PN	NI	N	PN
	Was outcome assessment likely to be influenced by knowledge of intervention received?	NA	NA	NA	NA	NA	NA	N	NA	NA
	Overall risk of bias	Low risk	Low risk	Low risk	Low risk	Low risk	Low risk	Low risk	Low risk	Low risk
Bias in selection of the reported result	Reported outcome data likely selected from multiple outcome measurements in the outcome domain?	N	N	N	N	N	N	N	N	N
	Reported outcome data likely selected from multiple analyses of the data?	N	N	N	N	N	N	N	N	N
	Reported outcome data likely selected from the outcome of a statistical test for carry-over?	N	N	N	N	N	N	N	N	N
	Overall risk of bias	Low risk	Low risk	Low risk	Low risk	Low risk	Low risk	Low risk	Low risk	Low risk
Overall risk of bias judgment		Some concerns	Some concerns	Some concerns	Some concerns	Low risk	Some concerns	High risk	Low risk	Some concerns

Y yes, PY probably yes, N no, PN probably no, NI no information, NA not applicable.

Figure 1. PRISMA flow diagram.

Figure 2. Forest plot for TTE data.

Figure 3. Forest plot for TTC data.

Figure 4. Post hoc sensitivity analysis of TTC outcome data.

Dhanakoti SN, Brosnan JT, HerzbergGR, et al. Renal arginine synthesis: studies in vitro and in vivo. Am J Physiol Endocrinol Metab. 1990;259(3):437.

 Google Scholar

Wu G, Morris SM. Arginine metabolism: nitric oxide and beyond. Biochem J. 1998;336(1):1–17.

 PubMed Web of Science ®Google Scholar