Changes in resistance training performance, rating of perceived exertion, and blood biomarkers after six weeks of supplementation with L-citrulline vs. L-citrulline DL-malate in resistance-trained men: a double-blind placebo-controlled trial

ABSTRACT

Purpose

This study aimed to investigate and compare the effects of chronic supplementation with L-Citrulline (LC) vs. L-Citrulline DL-malate (CM) on resistance training (RT) performance.

Methods

Thirty-three resistance-trained men were randomly assigned to ingest LC (8 g), CM (12 g), or Placebo (PL) daily, along with participation in a 6-week RT protocol. Muscular strength (1-repetition maximum [1RM] for hack squat [HS] and bench press [BP]), muscular endurance (repetitions to failure [RTF] for HS, leg extension [LE], BP, and incline press [IP]), rating of perceived exertion (RPE), and blood biomarkers (lactate, urea, and nitric oxide metabolites [NO_X]) were assessed before and after the intervention. This study was registered on irct.ir (IRCTID: IRCT20221128056642N1).

Results

Comparing mean ∆ scores revealed a significant difference between LC and PL (p < 0.001) and between CM and PL (p = 0.026) for total upper body (the sum of BP and IP) RTF, but only a trend for difference between LC and PL (p = 0.070) for total lower body (the sum of HS and LE) RTF. A significant time effect for NO_X was detected only for LC (p = 0.014) and CM (p = 0.003). In addition, a significant difference between CM and PL (p = 0.009) and a marginally significant difference between LC and PL (p = 0.057) was detected regarding post-exercise NO_X values at post-intervention. There were no other between-group differences for any outcome measure.

Conclusion

Chronic citrulline supplementation seems to enhance upper body muscular endurance and post-exercise NO_X response to RT, but there is no apparent difference between LC and CM in these aspects.

1. Introduction

L-Citrulline (LC) is a non-essential, non-proteinogenic amino acid that has gained popularity as a Nitric Oxide (NO) enhancer in sports nutrition [1]. The primary physiological effect of NO during exercise is vasodilation, leading to an increased supply of oxygen and energy substrates to muscle tissues [1,2]. Research shows that LC supplementation may also help reduce muscle soreness and fatigue, making it a promising option for resistance-trained athletes seeking to optimize their training performance [3]. Moreover, LC may positively affect muscle protein synthesis, further supporting hypertrophic adaptations in response to resistance training (RT) [4]. Therefore, supplementing with LC may improve performance, aid recovery, and promote long-term adaptations to RT.

Citrulline (CIT) supplements are available in different forms, including pure LC and L-Citrulline DL-malate (CM), which is an organic salt made up of LC and malic acid (or malate) in ratios ranging from 1:1 to 2:1. [1,5,6]. Most of the ergogenic properties of CIT supplements are attributed to LC’s role as an L-arginine precursor, which leads to increased L-arginine bioavailability and enhanced NO production. NO-mediated vasodilation during RT may delay fatigue and allow for higher training volumes by improving oxygen and nutrient delivery to the working muscles and facilitating the replenishment of ATP between RT sets [7]. In addition, LC may facilitate the clearance of ammonia and decrease lactate production through the urea cycle [3]. In the case of CM, malic acid, as a Krebs cycle intermediate, may also contribute to the potential ergogenic effects of LC by positively affecting ATP production rate [8]. Although recent studies, including those by Olmedo et al. [9], have started to explore this relationship, the proposed synergistic effect of malate on LC’s ergogenic benefits still requires further investigation to be fully confirmed [10].

Based on these proposed mechanisms, most studies in resistance-trained populations have investigated the acute effects of CIT supplementation on factors like muscular endurance, rating of perceived exertion (RPE), muscle soreness, vasodilation, and levels of related blood biomarkers [5,11–16]. While earlier studies reported promising results regarding RT performance improvement [14,16–19], most recent studies have failed to observe such effects [5,9,12,13,15,20–23]. The mixed findings may stem from variations in study design, dosage, and participant characteristics, which can influence the outcomes [4,10]. The inconsistency in findings related to CM supplementation can also be attributed to discrepancies in the LC to malate ratios used in different studies, as many commercial products do not meet the expected ratios [5]. Overall, it appears that acute CIT supplementation delays fatigue, increases repetitions to failure (RTF), and generally provides a small ergogenic effect [1,7].

Research is scarce regarding chronic supplementation with CIT [7,24]. In a study by Fick et al., 7-day supplementation with 8 g of CM in recreationally trained men failed to improve muscle contractile properties or fatigue rate of the quadriceps [22]. Gonzalez et al. also showed that 7 days of supplementation with watermelon juice (containing 2.2 g of LC) fails to improve isometric force production or bench press RTF [25]. In the only longer-term (8 weeks) study conducted to date, 8 weeks of supplementation with either LC (2 g/day, combined with 200 mg of glutathione) or CM (2 g/day) did not improve body composition or lean body mass [24].

The minimum effective dose of LC appears to be around 3 g/day [4], but higher doses up to 15 g are shown to be safe and well-tolerated [26]. Interestingly, [4,10] most prior studies have delivered an LC dose of 5.3 g or less [5,9,14–18]. Thus, we hypothesized that administering higher doses of LC or CM in conjunction with a long-term RT protocol would secure observing an ergogenic effect. In addition, most studies showing an ergogenic effect of CIT supplementation have used CM, suggesting that malate may enhance RT performance [4,10]. Excluding Hwang et al.‘s study [24] on underdosing and glutathione use, no long-term studies have compared the effects of isolated LC to CM. We hypothesized that both LC and CM would improve outcomes compared to placebo (PL), with CM potentially offering greater performance enhancement. Thus, this study is the first, to our knowledge, to investigate the effects of six weeks of supplementation with LC (8 g/day) or CM (12 g/day; a 2:1 ratio of LC to malate) on resistance-trained men’s muscular performance, RPE, and blood biomarkers.

2. Methods

2.1. Study design

A double-blind, placebo-controlled design, with a 6-week RT and supplementation protocol, was employed and participants were randomly assigned to ingest LC, CM, or PL. Participants performed 4 weekly RT sessions, and RT performance, body composition, and blood biomarkers were assessed before and after the 6-week RT and supplementation protocol (Figure 1). The anthropometric data collection and body composition assessment procedures were carried out at the Exercise Metabolism and Performance Lab, located at the Faculty of Sport Sciences in Razi University (Kermanshah, Iran). All training sessions and other visits were conducted indoors at the Iran Fit Gym (Kermanshah, Iran).

Figure 1.

Timeline of testing and training protocol.

1RM 1 repetition maximum, 10RM 10 repetition maximum, RPE rating of perceived exertion, RT resistance training.

2.2. Participants

An a priori power analysis (G*Power v. 3.1.9.4) indicated that a total sample size of 30 participants would be required to detect significant differences. This calculation was based on an effect size of 0.20, an α of 0.05, and a power of 0.80 for a repeated-measures, within-between analysis of variance (ANOVA) design with 3 groups and a correlation of 0.8 among repeated measures values. To account for drop-outs, thirty-six healthy male volunteers [between the ages of 18–35, with a body mass index between 18.5–29.9 kg/m², and with more than 6 months of RT experience before enrollment (≥3×/week)] were recruited. Only those individuals were allowed to participate in the study who were considered low risk by the American College of Sports Medicine [27], were nonsmokers, had not used any ergogenic supplements for 8 weeks prior to enrollment, and had no orthopedic problems or metabolic disorders that could interfere with the outcome measures or full commitment to the study protocol. Participants were informed of the risks and purposes of the study and then signed a university-approved written consent. Additionally, all the study procedures conformed to the ethical considerations of the Declaration of Helsinki. The study was approved by the research ethics committee of Razi University (Protocol # IR.RAZI.REC.1401.057; date of approval: 26/10/2022) and was prospectively registered on irct.ir (IRCTID: IRCT20221128056642N1; date of registry: 29/11/2022). Thirty-three participants (LC: n = 11, 26.3 ± 4.6 y, 179.9 ± 7.3 cm, 84.7 ± 10.9 kg; CM: n = 12, 25.7 ± 3.7 y, 177.8 ± 3.9 cm, 83.6 ± 8.1 kg; PL: n = 10, 24.3 ± 4.1 y, 179.4 ± 5.9 cm, 80.5 ± 7.5 kg) completed the study and their data was used in the final analyses (Figure 2).

Figure 2.

Participant flow diagram.

2.3. Anthropometrics and body composition

The height of barefooted participants was measured using a wall-mounted stadiometer to the nearest 0.1 cm. Body mass and percent body fat (PBF) were assessed via a bioelectrical impedance body composition analyzer (Zeus 9.9 PLUS; Jawon Medical Co., Ltd., Kungsang Bukdo, South Korea). For body composition measurement during visits 1 and 5, participants reported to the laboratory at 8 A.M., after 8 h of sleep, 10 to 12 h of fasting, and at least 24 h of abstinence from caffeine, alcohol, or exercising.

2.4. Muscle performance assessments

The maximal strength of participants was assessed by performing 1 repetition maximum (1RM) tests on the plate-loaded hack squat (HS) and barbell bench press (BP) exercises at baseline (visit 2) and after the 6-week RT and supplementation protocol (visit 6). Participants were instructed to report to the laboratory at their prescheduled time (between 8 A.M. and 12 P.M.), in a fed state. The 1RM testing procedure was conducted according to the National Strength and Conditioning Association (NSCA) guidelines [28]. Each session started with a 10-min warm-up: 5 min of treadmill jogging followed by dynamic exercises for the main muscle groups, including one set of 12 repetitions for body-weight squats, walking lunges, hyperextensions, BP, and lateral raises, as well as any additional exercises of their choice to complete the 10-min timeframe. After performing three progressive warm-up sets, the 1RM was obtained within 3 to 5 main attempts, with a 3- to 4-min rest period after each attempt.

During visit 2, 10 repetition maximum (10RM) testing for leg extension (LE) and plate-loaded incline chest press (IP) was also performed using an identical pattern; however, intensities were relative to a 10RM instead of a 1RM. After the completion of each exercise’s 1RM/10RM testing procedure, there was a five-minute break. The maximum load lifted with proper technique was recorded for each exercise as the participant’s 1RM/10RM. The RM values obtained during this visit were used to determine the exercise intensity for visits 3 and 4, as well as the 6-week RT program. Specifications like grip width, stance width, individual technique, and machine settings were recorded for all four tested exercises and were replicated during experimental visits and training sessions.

2.5. Experimental visits

Both visits (i.e. visits 3 and 4) were performed between 8 A.M. and 1 P.M., at the same time of the day for a given participant (±1 hr). Participants reported to the laboratory 90 min after breakfast and rested seated for 15 min. A resting venous blood sample was obtained and then another 10 min of rest was allowed before starting the RT protocol. The protocol started with the same 10-min warm-up as performed during visit 2 and consisted of performing four exercises in the following order: HS (3 sets at 75% 1RM), LE (3 sets at 10RM), BP (3 sets at 75% 1RM), and IP (3 sets at 10RM). A second blood sample was taken within a minute of completing the last set of IP.

During visit 3, participants were required to continue each set to momentary muscular failure (MMF); A trained research assistant monitored and recorded each participant’s performance during each set. During visit 4, the same RT protocol was replicated, but the participants were required to stop each set after reaching the recorded number of repetitions during visit 3 and to avoid continuing the set to MMF even if possible. Participants rested for 2 min after each set and 3 min after each exercise. During both visits, RPE was measured upon the completion of each set using the Borg Category Ratio (CR-10) scale. Participants were asked to indicate their RPE by selecting a number from a scale where “0” represents no effort (i.e. rest) and “10” represents maximal effort (i.e. extremely heavy physical exercise) [29].

2.6. Resistance training protocol

All participants performed the same 4-day per week, split routine RT program for six weeks (Table 1). For HS and LE (day 1) and also BP and IP (day 2), participants were required to exercise with the same loads used during the experimental visits (i.e. 75% 1RM for HS and BP and 10RM for LE and IP). As these exercises were used to evaluate muscular endurance (RTF), participants were asked to continue all sets to MMF. Research assistants directly supervised training on days 1 and 2 in both week 1 and week 6 to record each participant’s RTF in these four exercises for future analyses. For all other exercises, participants were instructed to self-select a load that limited them to the prescribed rep ranges. They were asked to stop the first two sets two repetitions before failure and only push to MMF on the last set of each exercise. Participants were advised to rest for 2 min after each set and 3 min after each exercise. To promote compliance, participants were permitted to train at their preferred time of the day but were required to maintain this training schedule throughout the study. Participants were required to keep a daily training log and submit it at the end of the study to monitor adherence and compliance.

Table 1.

Six-week resistance training program.

Day 1: Monday		Day 2: Tuesday
Exercise	Sets/Reps (RM)	Exercise	Sets/Reps (RM)
Hack Squat	3×75% 1RM	Barbell Bench Press	3×75% 1RM
Leg Extension	3×10RM	Machine Incline Press	3×10RM
Standing Calf Raises	3×8–12	Machine Shoulder Press	3×8–12
Wide-Grip Lat Pulldown	3×8–12	DB Lateral Raises	3×8–12
Seated Row	3×8–12	EZ Bar Skull Crusher	3×8–12
EZ Bar Curl	3×8–12	Triceps Pushdown	3×8–12
DB Hammer Curl	3×8–12	Reverse Crunch	3×15–20
Hyperextension	3×15–20	Cable Crunch	3×15–20
Day 3: Thursday		Day 4: Friday
Leg Press	3×8–12	Barbell Incline Press	3×8–12
Lying Leg Curl	3×8–12	Machine Chest Fly	3×8–12
Romanian Deadlift	3×8–12	DB Shoulder Press	3×8–12
Seated Calf Raises	3×8–12	Machine Reverse Fly	3×8–12
Straight-Arm Pulldown	3×8–12	OH Triceps Extension	3×8–12
One-Arm DB Row	3×8–12	Cable Kickback	3×8–12
EZ Bar Preacher Curl	3×8–12	Crunch	3×15–20
Rope Cable Curl	3×8–12	Russian Twist	3×15–20

Reps repetitions, RM Repetition maximum, 1RM 1 repetition maximum, DB dumbbell, OH overhead.

2.7. Supplementation protocol

Pure L-Citrulline and L-Citrulline DL-malate 2:1 (Batch No.: YTLC2022011002 and YTCM22011001, respectively) were obtained from YT (XI’AN) Biochem Co., LTD. (Shaanxi, China), and were accompanied by certificates of analysis; The total citrulline malate content in the CM product was measured at 99.3% (comprising 70.0% L-citrulline and 29.3% DL-Malate); The LC product was found to contain 99.5% L-citrulline. Participants were matched according to body mass and randomized to one of three groups: 1) 8 g of L-citrulline + 15 g of Maltodextrin + 3 g of Citric Acid (LC), 2) 12 g of Citrulline malate + 15 grams of Maltodextrin (CM), or 3) 15 g of Maltodextrin + 3 g of Citric Acid (PL). Based on the certificates of analysis and prescribed doses, the actual dose of L-citrulline in the LC condition was 7.96 g (8 g x 99.5% L-citrulline), whereas the actual dose of L-citrulline in the CM condition was 8.4 g (12 g L-citrulline DL-malate x 70.0% L-citrulline). Adding sour cherry-flavored Maltodextrin (Arshia Supplement Co., Isfahan, Iran; Batch No: 1228169397) to all supplement powders and incorporating citric acid into LC and PL was done to ensure consistency in color, taste, and carbohydrate content. Each participant in a supplement group received a study kit that contained 42 single-dose supplement sachets, an opaque shaker cup, printed supplementation instructions, and logs to report supplementation compliance. Participants took their supplements 60 min before exercising on training days and after breakfast on non-training days.

2.8. Dietary intake

The participants’ diets were not standardized and they were asked to maintain their normal dietary intake for the duration of the study. Participants were provided with a 3-day dietary booklet to record their food intake for 3 days (2 weekdays and 1 weekend day) before and after the 6-week intervention. The 3-day diet records were evaluated using the Nutritionix online database (available at https://www.nutritionix.com) to determine the average daily energy intake and macronutrient consumption.

2.9. Blood sampling

To evaluate the plasma levels of lactate, urea, and NO metabolites (NO_X), 9-mL venous blood samples were obtained from the antecubital region of the arm 10 min before and immediately after exercise during both experimental visits. Samples were immediately centrifuged at 3000 rpm for 10 min, and then aliquots of plasma were frozen at −40${ }^{\circ }$ C until future analysis. The levels of lactate and urea in plasma samples were analyzed using commercially available enzymatic/colorimetric assays (lactate: Product Code: BXC0622; Biorex Fars, Iran: https://biorexfars.ir; and urea: Urease-GLDH; Delta Darman Part, Iran: https://delta-dp.ir). A commercially available colorimetric assay kit determined the NO_X levels in plasma samples (CAT No. ZB-NO-96A, ZellBio GmbH, Germany).

2.10. Statistical analyses

The normality of distribution for each group at each time point was assessed using the Shapiro–Wilk test and evaluation of skewness and kurtosis. The primary analysis performed was a 3 (groups: LC vs. CM vs. PL) × 2 (time: pre- vs. post-training) repeated-measures analysis of variance (RMANOVA) to determine differences between groups over time for each of the outcome measures. Post-hoc tests were performed as needed using pairwise comparisons with the least significant difference (LSD) method. A one-way ANOVA was used as a secondary analysis to assess absolute mean change (∆) scores (post mean – pre mean) between groups for body mass, PBF, diet, muscular performance, RPE, and blood biomarkers. In cases where the assumption of homogeneity of variances was met, the results of the standard one-way ANOVA were reported; wherever this assumption was violated, the Welch ANOVA, which is robust to violations of homogeneity, was reported. For outcome measures where the assumption of homogeneity of variances was violated and Welch’s ANOVA indicated significant differences between group means, post-hoc comparisons were conducted using the Game–Howell test. Effect sizes were reported using partial eta squared (η² x_p), which represents the proportion of variance explained by a given factor while accounting for error variance. Effect size interpretation followed Cohen’s (1988) guidelines, where η² x_p values of 0.01, 0.06, and 0.14 correspond to small, medium, and large effects, respectively [30]. Data are presented as means ± SD. Statistical significance was set at an alpha level of p < 0.05. All statistical analyses were performed using SPSS version 23 (SPSS, Inc., Chicago, IL).

3. Results

3.1. Body composition and diet

There was no significant time (p = 0.955, η² x_p = 0.000) or group (p = 0.391, η² x_p = 0.061) effect for body mass, but a significant group × time interaction (p = 0.044, η² x_p = 0.188) was detected. As shown in Table 2, further analysis revealed a statistically significant body mass loss for PL (∆: −1.3 ± 2.6; 95% Confidence Interval (CI) [−2.65, −0.01]; p = 0.048), with a slight, non-significant increase in body mass for LC (∆: 0.6 ± 1.0; CI [−0.67, 1.85]; p = 0.345) and CM (∆: 0.8 ± 2.2; CI [−0.4, 2.0]; p = 0.185). However, there was no significant difference between groups regarding mean ∆ in body mass (Welch’s ANOVA: p = 0.111, η² x_p = 0.188). For BPF, there were no significant main effects or group × time interaction (time: p = 0.348, η² x_p = 0.029; group: p = 0.773, η² x_p = 0.017; interaction: p = 0.894, η² x_p = 0.007), and neither was a significant difference in mean ∆ scores (Welch’s ANOVA: p = 0.901, η² x_p = 0.007). Regarding dietary composition, no significant main effects or group × time interaction existed for total daily calories (time: p = 0.598, η² x_p = 0.013; group: p = 0.500, η² x_p = 0.061; interaction: p = 0.256, η² x_p = 0.116), carbohydrates (time: p = 0.698, η² x_p = 0.007; group: p = 0.761, η² x_p = 0.024; interaction: p = 0.418, η² x_p = 0.076), fat (time: p = 0.489, η² x_p = 0.022; group: p = 0.196, η² x_p = 0.138; interaction: p = 0.855, η² x_p = 0.014), or protein (time: p = 0.847, η² x_p = 0.002; group: p = 0.337, η² x_p = 0.094; interaction: p = 0.188, η² x_p = 0.141) intake, neither was a significant difference in mean ∆ scores (calories: p = 0.256, η² x_p = 0.116; carbohydrates: p = 0.418, η² x_p = 0.076; fat: p = 0.855, η² x_p = 0.014; protein : p = 0.188, η² x_p = 0.141) (Table 2).

Table 2.

Mean absolute change (∆) from baseline to 6 weeks for body composition and dietary analysis.

Variable	Group	Baseline	Post-intervention	∆
Body mass (kg)	LC	84.7 ± 10.9	85.3 ± 11.5	0.6 ± 1.0
	CM	83.6 ± 8.1	84.4 ± 8.6	0.8 ± 2.2
	PL	80.5 ± 7.5	79.2 ± 5.8*	−1.3 ± 2.6
Body fat (%)	LC	22.3 ± 4.7	21.7 ± 5.3	−0.6 ± 1.5
	CM	20.6 ± 4.6	20.4 ± 6.2	−0.1 ± 2.9
	PL	21.4 ± 4.9	21.1 ± 4.2	−0.4 ± 1.6
Total calories (kcal)	LC	2566.2 ± 629.2	2541.2 ± 528.8	−25.1 ± 140.0
	CM	2576.3 ± 590.7	2624.1 ± 631.5	−47.8 ± 156.0
	PL	2347.9 ± 356.6	2279.1 ± 300.9	−68.8 ± 130.2
Carbohydrate (g/d)	LC	328.1 ± 107.5	321.5 ± 72.2	−6.6 ± 49.5
	CM	297.0 ± 111.2	313.1 ± 122.0	16.1 ± 26.0
	PL	292.2 ± 54.7	291.3 ± 41.6	−1.0 ± 30.9
Fat (g/d)	LC	84.7 ± 13.8	81.3 ± 18.3	−3.5 ± 11.1
	CM	90.2 ± 19.2	90.0 ± 17.4	−0.2 ± 15.2
	PL	74.7 ± 23.3	73.2 ± 19.2	−1.4 ± 8.2
Protein (g/d)	LC	116.3 ± 37.0	129.1 ± 48.9	12.9 ± 23.1
	CM	143.7 ± 37.1	139.6 ± 28.4	−4.1 ± 17.1
	PL	119.7 ± 45.2	113.5 ± 23.6	−6.2 ± 26.0

Mean ± SD.

LC l-citrulline, CM citrulline malate, PL placebo, Kcal kilocalorie, g/d gram per day.

*Significantly different from the baseline of the same group.

3.2. Muscular strength

A main effect for time was observed for HS and BP (p < 0.001, η² x_p = 0.910 and 0.559, respectively), with all groups showing a significant increase in HS and BP 1RM compared to baseline, regardless of treatment. However, there was no significant main effect for group (HS: p = 0.603, η² x_p = 0.033; BP: p = 0.096, η² x_p = 0.145), and also no significant group × time interaction (HS: p = 0.763, η² x_p = 0.018; BP: p = 0.517, η² x_p = 0.043). Comparing mean ∆ scores for both HS and BP showed a similar pattern, with no significant difference between groups for 1RM (HS: p = 0.763, η² x_p = 0.018; BP: p = 0.517, η² x_p = 0.043) (Figure 3).

Figure 3.

1RM strength change from baseline to post-intervention for hack squat (a) and bench press (b).

1RM 1 repetition maximum, LC l-citrulline, CM citrulline malate, PL placebo.

3.3. Muscular endurance

Pre- to post changes in RTF are depicted in Table 3. Regarding lower body performance, a significant main effect for time (p < 0.001, η² x_p = 0.776) was observed for HS RTF, as all groups demonstrated a significant improvement in the total number of repetitions achieved over 3 sets of HS. However, no significant main effect was observed for group (p = 0.718, η² x_p = 0.022) or group × time interaction (p = 0.176, η² x_p = 0.109). Similarly, comparing mean ∆ scores showed no difference between groups for HS (Welch’s ANOVA: p = 0.061, η² x_p = 0.109). For LE, a significant main effect for time was observed (p < 0.001, η² x_p = 0.655), indicating an overall improvement in RTF across all groups. However, no significant main effect for group (p = 0.096, η² x_p = 0.145) or group × time interaction (p = 0.468, η² x_p = 0.049) was found. Comparing mean ∆ scores showed no difference between groups for LE (p = 0.468, η² x_p = 0.049). For total lower body (TLB; the sum of HS and LE values) RTF, there was a significant main effect of time (p < 0.001, η² x_p = 0.817), with no significant group effect (p = 0.670, η² x_p = 0.026) or group × time interaction (p = 0.127, η² x_p = 0.129). However, comparing mean ∆ scores showed a significant difference between groups for TLB (Welch’s ANOVA: p = 0.035, η² x_p = 0.129); Further analysis showed only a trend for difference between LC and PL (∆ difference: 9.427; CI [−0.73, 19.58]; Games–Howell Test: p = 0.070).

Table 3.

Mean absolute change (∆) from baseline to 6 weeks for muscular endurance (RTF).

Variable	Group	Baseline	Post-intervention	∆
HS RTF	LC	36.0 ± 6.4	58.1 ± 13.5	22.1 ±11.1
	CM	39.5 ± 6.6	60.3 ± 15.8	20.8 ± 13.2
	PL	42.3 ± 4.8	56.1 ± 7.9	13.8 ± 5.05
LE RTF	LC	30.6 ± 2.5	34.8 ± 3.2	4.1 ± 3.5
	CM	30.2 ± 3.5	34.8 ± 3.7	4.5 ± 2.9
	PL	28.8 ± 2.6	31.8 ± 2.5	3.0 ± 2.2
TLB RTF	LC	66.6 ± 6.8	92.9 ± 13.8	26.2 ± 11.7
	CM	69.8 ± 7.7	95.1 ± 15.0	25.3 ± 14.2
	PL	71.1 ± 6.2	87.9 ± 8.4	16.8 ± 5.0
BP RTF	LC	29.6 ± 2.7	34.5 ± 3.1	4.9 ± 2.2
	CM	31.0 ± 3.6	36.4 ± 4.6	5.4 ± 2.2
	PL	29.2 ± 4.1	32.3 ± 3.7	3.1 ± 1.5
IP RTF	LC	22.9 ± 2.9	28.1 ± 3.7	5.2 ± 3.2
	CM	23.3 ± 4.5	26.9 ± 4.5	3.6 ± 3.4
	PL	24.7 ± 4.0	26.2 ± 3.7	1.5 ± 1.7
TUB RTF	LC	52.5 ± 4.9	62.6 ± 4.6	10.1 ± 2.6
	CM	54.3 ± 6.5	63.3 ± 8.0	9.0 ± 4.7
	PL	54.0 ± 7.0	58.5 ± 5.8	4.5 ± 2.2

Mean ± SD.

RTF repetitions to failure, HS hack squat, LE leg extension, TLB total lower body, BP bench press, IP incline press, TUB total upper body, LC l-citrulline, CM citrulline malate, PL placebo.

For upper body performance, there was a significant main effect for time (p < 0.001, η² x_p = 0.840) and also for Group × time interaction (p = 0.031, η² x_p = 0.206) for BP RTF, while no significant group effect (p = 0.174, η² x_p = 0.110) was observed. Further post hoc analyses revealed no significant differences between groups at baseline (LC vs. CM: p = 0.357, CI [−4.34, 1.61]; LC vs. PL: p = 0.802, CI [−2.73, 3.50]; CM vs. PL: p = 0.251, CI [−1.30, 4.80]), but a significant difference was observed at post-intervention. The CM group achieved significantly more repetitions compared to the PL (mean difference = 4.075, p = 0.020, CI [0.68, 7.47]), with no significant difference between the LC and PL (mean difference = 2.25, p = 0.196, CI [−1.22, 5.71]) or between the LC and CM (mean difference = −1.83, p = 0.268, CI [−5.14, 1.48]). The result of 1-way ANOVA comparing mean ∆ in BP RTF from baseline to post-intervention was also significant (p = 0.031, η² x_p = 0.206); Follow-up comparisons revealed a significant difference between LC and PL (∆ difference: 1.859; CI [0.51, 3.67]; p = 0.044) and between CM and PL (∆ difference: 2.325; CI [0.55, 4.10]; p = 0.012). Regarding IP, there was a significant main effect for time (p < 0.001, η² x_p = 0.598), and also a significant group × time interaction (p = 0.023, η² x_p = 0.222), but no significant group effect (p = 0.959, η² x_p = 0.003) was found. Follow-up comparisons showed that the significant time effect was driven by the LC (∆: 5.2 ± 3.2; CI [3.38, 6.98]; p < 0.001) and CM (∆: 3.6 ± 3.4; CI [1.86, 5.31]; p < 0.001) groups (p < 0.001), with no significant change over time for the PL group (∆: 1.5 ± 1.7; CI [−0.44, 3.34]; p = 0.127). Additionally, the results of 1-way ANOVA for mean ∆ in IP RTF from baseline to post-intervention were significant (Welch’s ANOVA: p = 0.008, η² x_p = 0.222); Further analysis detected a significant difference between LC and PL (∆ difference: 3.731; CI [0.88, 6.58]; Games–Howell Test: p = 0.01), with no other significant difference between groups. For total upper body (TUB; the sum of BP and IP) RTF, a significant effect for time (p < 0.001, η² x_p = 0.850) and group × time interaction (p = 0.002, η² x_p = 0.338) was detected, but no significant group effect (p = 0.614, η² x_p = 0.032) was observed. Further analysis revealed no significant difference between groups. However, a significant difference was detected by comparing mean ∆ scores (Welch’s ANOVA: p < 0.001, η² x_p = 0.222), with follow-up comparisons indicating a significant difference between LC and PL (∆ difference: 5.590; CI [2.94, 8.24]; Games–Howell Test: p < 0.001) and between CM and PL (∆ difference: 4.458; CI [0.52, 8.40]; Games–Howell Test: p = 0.026).

3.4. Perceived exertion

TLB (the sum of HS and LE values) showed a significant main effect for time (p < 0.001, η² x_p = 0.664), as the reported RPEs significantly reduced at post-intervention compared to baseline, regardless of treatment. Baseline values were: LC: 37.8 ± 9.4; CM: 41.3 ± 7.5; PL: 40.1 ± 8.3 and post-intervention values were: LC: 28.2 ± 7.6; CM: 27.2 ± 10.5; PL: 25.1 ± 10.2. There were no significant main effects for group (p = 0.873, η² x_p = 0.009) or group × time interaction (p = 0.392, η² x_p = 0.061). Comparing mean ∆ scores also showed no difference between groups for TLB RPE (∆: LC: −9.6 ± 9.4, CI [−15.98, −3.30]; CM: −14.1 ± 10.2, CI [−20.57, −7.60]; PL: −15 ± 9.0, CI [−21.43, −8.58]; p = 0.392, η² x_p = 0.061). Similarly, TUB (the sum of BP and IP values) showed a significant main effect for time (p < 0.001, η² x_p = 0.513). Baseline values were: LC: 39.3 ± 8.3; CM: 43.3 ± 7.3; PL: 40.1 ± 8.1 and post-intervention values were: LC: 30.8 ± 9.9; CM: 30.9 ± 11.7; PL: 29.8 ± 9.8. There were no significant main effects for group (p = 0.748, η² x_p = 0.019) or group × time interaction (p = 0.672, η² x_p = 0.026). Comparing mean ∆ scores also showed no difference between groups (∆: LC: −8.5 ± 8.1, CI [−13.90, −3.01]; CM: −12.4 ± 12.1, CI [−20.12, −4.72]; PL: −10.3 ± 11.02, CI [−18.18, −2.42]; p = 0.672, η² x_p = 0.026) regarding TUB RTF.

3.5. Training and supplementation compliance and blinding efficacy

The participants’ overall training compliance was 96.5 ± 3.9%, with no significant difference between groups (LC: 96.6 ± 3.6%; CM: 95.8 ± 4.7%; PL: 97.1 ± 3.4%; p = 0.762, η² x_p = 0.018). Overall supplementation compliance was 96.9 ± 3.0%, also with no significant difference between groups (LC: 95.5 ± 3.6%; CM: 97.8 ± 2.4%; PL: 97.4 ± 2.6%; p = 0.141, η² x_p = 0.123). To assess blinding efficacy, after the study, participants were asked which group they thought they had been assigned to. The results indicated an overall blinding efficacy of 66.7%, where 30% (3/10), 36.4% (4/11), and 33.3% (4/12) of the participants correctly guessed the PL, LC, and CM conditions, respectively. Even though this information was not officially recorded or statistically analyzed, it was noted that the majority of participants who guessed correctly did so based on their subjective feelings and energy levels, not on the taste or appearance of the supplement they consumed.

3.6. Blood analytes

The data for blood analytes is shown in Figure 4. For lactate and urea, comparing baseline mean ∆ (visit 3: post-exercise values – pre-exercise values) with post-intervention mean ∆ (visit 4: post-exercise values – pre-exercise values) showed no significant main effects for time (lactate: p = 0.271, η² x_p = 0.040; urea: p = 0.251, η² x_p = 0.044), group (lactate: p = 0.726, η² x_p = 0.021; urea: p = 0.478, η² x_p = 0.048), or group × time interaction (lactate: p = 0.685, η² x_p = 0.025; urea: p = 0.971, η² x_p = 0.002). The results of 1-way ANOVA comparing mean ∆ (i.e. post-intervention mean ∆ - baseline mean ∆) also showed no significant difference between groups (lactate ∆: LC: −1.30 ± 3.72, CI [−3.80, 1.20]; CM: −0.09 ± 2.99, CI [−1.98, 1.81]; PL: −0.57 ± 3.24, CI [−2.89, 1.75]; p = 0.685, η² x_p = 0.025; and urea ∆: LC: −0.27 ± 0.99, CI [−0.94, 0.39]; CM: −0.18 ± 1.06, CI [−0.85, 0.50]; PL: −0.18 ± 1.05, CI [−0.93, 0.57]; p = 0.971, η² x_p = 0.002).

Figure 4.

Comparison of baseline mean change (∆) (visit 3: post-exercise values – pre-exercise values) with post-intervention mean ∆ (visit 4: post-exercise values – pre-exercise values) in the plasma levels of lactate (a), urea (b), and nitrate/nitrite (NO_x) (c).

LC l-citrulline, CM citrulline malate, PL placebo.

For NO_X, a significant main effect for time was detected (p = 0.001, η² x_p = 0.316). Follow-up comparisons revealed that this effect was driven by the LC (∆: 6.9 ± 6.0; CI [1.48, 12.34]; p = 0.014) and CM (∆: 8.1 ± 10.5; CI [2.92, 13.31]; p = 0.003) groups, with no significant change over time for the PL group (∆: 2.1 ± 9.2; CI [−3.56, 7.82]; p = 0.450). No significant main effects for group (p = 0.601, η² x_p = 0.033) or group × time interaction (p = 0.270, η² x_p = 0.084) were observed for NO_X. Similarly, the 1-way ANOVA result comparing mean ∆ did not show a significant difference between groups (p = 0.270, η² x_p = 0.084). However, comparing post-exercise values at baseline (visit 3) versus post-intervention (visit 4) revealed a marginally significant main effect for group (p = 0.51, η² x_p = 0.180). Follow-up comparisons showed no difference between groups in post-exercise NO_X values at baseline; However, there was a significant difference between CM and PL (mean difference: 11.924; CI [3.27, 20.58]; p = 0.009) and also a marginally significant difference between LC and PL (mean difference: 8.559; CI [−0.27, 17.39]; p = 0.057) at post-intervention, with no significant difference between LC and CM (mean difference: −3.365; CI [−11.80, 5.07]; p = 0.422) (Figure 5).

Figure 5.

Comparison of post-exercise NO_x values at baseline versus post-intervention.

LC l-citrulline, CM citrulline malate, PL placebo.

*Significant difference compared to PL at the same time point.

†Marginally significant (p = 0.057) difference compared to PL at the same time point.

4. Discussion

This was the first study to explore the effects of a 6-week RT program in conjunction with daily LC, CM, or PL on resistance-trained men’s muscular performance, RPE, and blood biomarkers. It was hypothesized that higher doses of LC (8 g) or CM (12 g) used in the current study would elicit an ergogenic effect. It was also hypothesized that although both LC and CM would positively affect the outcome measures when combined with the six-week RT protocol, CM may produce more substantial effects due to the potential synergistic effects of malate. We found no significant difference between LC, CM, and PL in muscular strength, RPE, lactate, and urea. However, we observed a significant difference between groups regarding upper body endurance, which indicated a notable enhancement of RTF for LC and CM, but not the PL group. We also noticed a trend indicating a difference between LC and PL concerning TLB RTF. In addition, CIT supplementation increased post-exercise NO_X levels compared to PL. Thus, the key finding of our study is that chronic supplementation with LC or CM does not positively affect muscular strength, RPE, lactate, or urea. However, it significantly improves upper body muscular endurance and NO response to exercise, and also shows potential for enhancing lower body endurance.

The present study showed significant HS and BP 1RM improvements across all groups. This indicates that the RT protocol effectively enhanced upper and lower body strength in resistance-trained men. However, the lack of significant differences between groups implies that LC or CM may not offer additional advantages for maximal strength development in a 6-week training and supplementation period. Although there is a distinct lack of studies investigating the longer-term and even acute effects of CIT on muscular strength, our results are consistent with the limited available literature. Hwang et al. [24] conducted the first long-term (8-week) study and reported no increase in upper or lower body maximal strength in response to 2 g of LC (combined with glutathione) or CM supplementation in resistance-trained men. In a study by Chappell et al. [5], an acute 8 g dose of CM failed to improve peak force during isometric, concentric, or eccentric muscle actions in an isokinetic leg extension protocol. In a meta-analysis including only four eligible studies, Aguiar, and Casonatto [2] concluded that CM does not enhance upper or lower body strength in healthy resistance-trained individuals. This may be because short-term strength gains are mainly associated with neural adaptations, while CIT mechanisms of action are more of a metabolic nature and may better serve muscular endurance [2].

As the first-ever study to explore the impacts of long-term CIT supplementation on muscular endurance, the analysis of TUB RTF mean ∆ scores (achieved over a total of 6 sets of BP and IP) and also BP RTF mean ∆ scores revealed that both the LC and CM groups significantly outperformed the PL; however, there was no significant difference between LC and CM. For IP, only the LC group showed a significant difference in RTF mean ∆ compared to the PL. However, in both cases, mean ∆ differences always favored the supplementation groups compared to PL, regardless of statistical significance. Overall, it appears that both LC and CM may have a similar positive impact on upper body muscular endurance. Regarding lower body endurance, the reported mean ∆ scores from baseline to post-intervention always favored supplementation groups over PL, but the differences were not statistically significant in most cases (Table 3). This suggests that while LC and CM may have conferred a marginal benefit to lower-body endurance, these effects were not robust enough to reach statistical significance.

There are several potential reasons for the inconsistent results in the analysis of upper body exercises in isolation or for the difference in endurance improvement between upper and lower body. It is possible that these non-significant differences could become significant with more experienced participants, longer-duration supplementation, and/or a larger sample size to detect even smaller effect sizes. Overall, the observed significant improvement in TUB endurance aligns with previous acute studies that reported similar ergogenic effects in different populations. [14,16,18,19,31]. In contrast, more recent acute studies have failed to report similar ergogenic effects [9,13,20,22,23,32]. Obviously, more chronic studies are required to better understand the chronic effects of CIT supplementation on muscular endurance.

One of the main proposed mechanisms for the potential ergogenic properties of CIT supplementation is increased NO production, as increases in blood flow mediated by NO may enhance muscular performance, hypertrophy, and strength [1,2]. Although some previous acute studies have failed to report an improvement in various indicators of blood flow and vasodilation [11,13,15,23], it was hypothesized that the long-term nature of this study might reveal significant differences that chronic CIT supplementation could potentially produce in blood biomarkers. Our findings suggest that LC and CM supplementation positively influenced NOx levels after exercise, while the PL group did not show significant change. The marginally significant difference between LC and PL and the significant difference between CM and PL in post-exercise NOx values at post-intervention suggest a potential benefit of CIT in enhancing NO production, albeit without a significant difference between LC and CM.

These findings align with existing research suggesting that CIT can enhance NO production, potentially contributing to improved blood flow and muscular endurance [33–36]. McKinley-Barnard et al. [34] reported that 7 days of supplementation with combined LC (2 g) and Glutathione (200 mg) resulted in significant increases in plasma NOx at 30 min post-exercise compared to the PL. In a more recent study by Valaei et al [37], acute ingestion of 12 g of LC prior to high-intensity interval training led to a significant increase in immediate post-exercise NOx levels compared to the PL. We cannot dismiss the possibility that the lack of strong interaction effects for NO_X in the current study may be related to methodological considerations. For example, using a larger sample size or evaluating NO_X levels at different time points post-exercise might have revealed a significant difference between groups. Given that previous studies examining the long-term ergogenic benefits of CIT have not measured NOx [24,38], it is important to conduct further long-term research to fully understand the specific mechanisms of these supplements in promoting NOx-related benefits.

Another potential mechanism for the performance-enhancing effects of CIT is reduced lactate levels via ammonia clearance through the urea cycle [2]. Ammonia accumulation in cells promotes glycolysis and inhibits aerobic pyruvate utilization, leading to lactate formation and fatigue. L-citrulline supplementation can reduce lactate production by buffering ammonia and enhancing aerobic pyruvate use [3]. Most acute studies have failed to show a positive impact on lactate concentration from CIT supplementation [5,16,19,20,39]. Similarly, Trexler et al. [15] reported no significant impact on lactate and urea in response to an acute 8 g dose of CM compared to PL. A recent meta-analysis [3] concluded that supplementing with CIT does not significantly reduce post-exercise blood lactate levels. The current study showed no significant difference in the plasma levels of lactate and urea between the groups either. However, these findings do not align with the few previous longer-term (2–4 week) studies that have reported a reduced lactate response following high-intensity exercise in athletes [40,41]. In a recent study, Viribay et al. [42] examined the effects of co-supplementation of Beetroot (500 mg NO₃⁻) and LC (6 g) in elite rowers and reported that co-supplementation resulted in improved lactate clearance after Wingate test compared to PL and Beetroot alone. Although speculative, possible explanations for the discrepancy in results could be the use of different athletic populations (e.g. rowers, pentathlon athletes, handball players), different exercise protocols, and different supplement dosing/combinations in previous long-term studies compared to the present study [40–42]. Overall, these results cast doubt on the significance of reduced lactate accumulation as a mechanism for CIT’s ergogenic properties and emphasize the necessity for more long-term studies to establish a consensus.

The current study was not without its limitations. First, although participants were recruited based on an a priori analysis to detect a small effect size of 0.20, recruiting an even larger sample size could have resulted in more robust positive results, as was mentioned previously. Second, using only recreationally resistance-trained men, and not including women, limits our ability to generalize to other athletic populations such as advanced lifters and women. Third, we did not conduct an independent chemical analysis of the supplements due to the lack of access to a third-party laboratory, as well as the absence of the necessary analytical equipment within our facilities. We relied on the manufacturer’s certificate of analysis to determine the composition of the supplements. Based on this method of dosing, the CM group may have received approximately 0.4 g more LC compared to the LC group. Additionally, while efforts were made to ensure similar taste and texture, there was variation in the total weight of the supplement doses across experimental groups; the PL condition was lower in total weight than the LC and CM. This may have compromised blinding efficacy, influencing treatment perceptions. Future studies should include independent verification of supplement content to ensure accurate nutrient dosing, and all supplement doses must be weight-matched to enhance blinding integrity.

Finally, subjects’ diets were not standardized and only 3-day diet records at baseline and post-intervention were used to assess dietary intake. Although participants claimed full compliance, this may not accurately reflect their nutritional intake throughout the study due to honesty and accuracy concerns. Interestingly, statistically significant body mass loss for PL and a slight (non-significant) increase in body mass for LC and CM (from baseline to post-intervention) were observed. In addition, there was no significant main effect or significant difference in mean ∆ scores for PBF. If the participants had been given personalized and supervised diet plans, we could hypothesize that the purported anabolic properties of LC may have helped to maintain or possibly even slightly increase lean body mass in the LC and CM groups compared to the PL group.

5. Conclusions

This is the first study to simultaneously examine and compare the ergogenic effects of LC (8 g) and CM (12 g) during chronic RT. We found that six-week supplementation with LC or CM in conjunction with an RT protocol improved upper body muscular endurance and NO response to exercise. There were no significant between-group differences for other outcome measures. Future research is warranted to further investigate the chronic effects of CIT supplementation on RT performance in different populations, preferably with larger sample sizes, along with the assessment of muscular hypertrophy and muscle protein synthesis. Overall, as CM did not provide additional benefits for muscular endurance beyond LC, different athletic populations may use LC and CM interchangeably to improve upper body muscular endurance during RT protocols.

Funding Statement

The author(s) reported there is no funding associated with the work featured in this article.

Author contributions

DB was the main investigator in charge of conceptualization, data collection, and subject monitoring. MA, NB, and GMT contributed to the study design and methodology. MA submitted the protocol for review to the ethics committee of Razi University. MA and NB supervised the process of data collection. MA, NB, and GMT contributed to data analysis and interpretation. DB wrote the original draft and MA, NB, and GMT reviewed the manuscript. All authors read and approved the final version of the manuscript.

Disclosure statement

DB is the founder and owner of Iran Fit Gym where training sessions and some experimental procedures were conducted, but he has no conflicts of interest to disclose. GMT provides consulting services to dietary supplement manufacturers through Tinsley Consulting LLC but has no affiliation with the supplement manufacturers whose products were used in the present study. The remaining authors have no competing financial interests to disclose.

Consent to participate

Informed consent was obtained from all participants included in the study.

Ethical approval

This study was approved by the research ethics committee of Razi University (Protocol # IR.RAZI.REC.1401.057; date of approval: 26/10/2022).