MATERIALS AND METHODS

This study consists of two parts. For study A, carnosine was measured in two muscles of the arm to investigate in which muscle of the upper body carnosine is reliably measurable by ¹H-MRS. Study B was a nutritional intervention study focusing on the difference in loading between arm and leg muscles in nonathletes and different athlete populations.

Muscle Carnosine Quantification by ¹H-MRS

All the MRS measurements were performed on a 3-T whole body MRI scanner (Siemens Trio, Erlangen, Germany). Carnosine content was measured by ¹H-MRS, as described by Baguet et al. (2). The subjects were lying in supine position on their back. To measure the calf muscles, the lower leg was fixed in a spherical knee coil with the angle of the ankle at 20° plantar flexion. A shoulder coil was used to measure carnosine in the deltoid and triceps muscle (Fig. 1A). Single voxel point-resolved spectroscopy (PRESS) sequence with the following parameters was used: repetition time (TR) of 2.000 ms, echo time (TE) of 30 ms, number of excitations is 128, 1.024 data points, spectral bandwidth of 1.200 Hz, and a total acquisition time of 4.24 min. The average voxel size for the soleus, gastrocnemius, deltoid and triceps brachii muscle was 40 × 12 × 30 mm, 40 × 12 × 30 mm, 40 × 13 × 30 mm, and 30 × 9 × 27 mm, respectively. The line width of the water signal for the soleus, gastrocnemius, deltoid, and triceps brachii muscle was on average, respectively, 25.5, 26.8, 27.4, and 25.2 Hz following shimming procedures. The integral of the C2-H peak (at ∼8 ppm) was quantified relative to the water peak integral (×1000) and reported as arbitrary units (11) (Fig. 1B). The reason why we could not calculate into millimolar concentrations is because the shoulder coil was a receive-only coil, whereas the knee coil was a transmit/receive coil. With a receive-only coil it is not possible to use the phantom method to calculate the millimolar concentrations, as we usually do. In order to allow comparison between the leg muscles and the deltoid, we expressed the concentration relative to the water peak (arbitrary units) for all muscles. We found significant correlation coefficients for the water signal pre- and postsupplementation in the three muscle groups: 0.684 (P < 0.001) for soleus, 0.632 (P < 0.001) for gastrocnemius, and 0.661 (P < 0.001) for deltoid. The variation coefficients were 4.23% for soleus, 4.59% for gastrocnemius and 7.1% for deltoid. There was no significant difference in water signals between pre- and postsupplementation for the different groups in the different muscles. These results showed that muscle water content was not influenced during supplementation which made it possible to use these arbitrary units.

Download figureDownload PowerPoint

RESULTS

Reliability of arm muscle carnosine quantification (study A).

The CV of the carnosine concentration was calculated for each muscle. Deltoid muscle had a smaller variation than the triceps brachii (6.6% vs. 9.2%: Table 2). Intraclass correlation coefficients (ICC) were significant for both deltoid (ICC: 0.903, P < 0.01) (Fig. 1C) and triceps muscle (ICC: 0.656, P = 0.01), respectively. Over a time period of 4 wk, we found a CV of 13.3% in the deltoid (Table 3) and a correlation coefficient of 0.734 (P = 0.03).

Table 2. The coefficient of variation (CV) of the carnosine content in deltoid and triceps muscle measured twice on 1 day (n = 11)

Muscle	Measurement 1	Measurement 2	Mean	SD	CV (%)
Deltoid
1	0.146	0.153	0.149	0.005	3.3
2	0.211	0.197	0.204	0.010	4.8
3	0.131	0.099	0.115	0.022	19.4
4	0.164	0.166	0.165	0.002	1.1
5	0.123	0.137	0.130	0.010	8.0
6	0.178	0.187	0.183	0.006	3.5
7	0.172	0.164	0.168	0.006	3.3
8	0.106	0.123	0.115	0.012	10.5
9	0.127	0.136	0.132	0.006	4.5
10	0.116	0.122	0.119	0.005	4.0
11	0.094	0.108	0.101	0.010	10.2
Mean	0.142	0.145			6.6
SD	0.035	0.031
Triceps
1	0.204	0.167	0.185	0.027	14.3
2	0.201	0.201	0.201	0.000	0.1
3	0.155	0.126	0.141	0.021	14.9
4	0.116	0.110	0.113	0.004	3.9
5	0.169	0.170	0.170	0.001	0.6
6	0.132	0.167	0.150	0.025	16.7
7	0.196	0.176	0.186	0.015	7.9
8	0.174	0.182	0.178	0.006	3.3
9	0.149	0.104	0.126	0.031	24.9
10	0.137	0.136	0.136	0.001	0.4
11	0.137	0.167	0.152	0.021	14.0
Mean	0.161	0.155			9.2
SD	0.030	0.031

Table 3. Coefficient of variation of the carnosine content in deltoid muscle at week 0 and 4 (n = 6)

Deltoid	0 wk	4 wk	Mean	SD	CV (%)
1	0.069	0.102	0.086	0.023	27.29
2	0.149	0.133	0.141	0.011	8.02
3	0.131	0.129	0.130	0.001	1.09
4	0.089	0.111	0.100	0.016	15.56
5	0.133	0.168	0.151	0.025	16.44
6	0.089	0.076	0.083	0.009	11.14
Mean	0.110	0.120			13.3
SD	0.032	0.031

CV, coefficient of variation.

BA-induced carnosine loading in arm vs. leg muscles (study B).

For the nonathletes, there was a significant increase in carnosine concentration in both arm (deltoid) and leg muscles (soleus + gastrocnemius) after supplementation, but the absolute and relative increase in carnosine in the arm vs. leg muscles showed no significant difference (P = 0.286 and P = 0.230, respectively) (Fig. 2). The relative increase was 47.44 ± 32.29% in arm muscle and 33.01 ± 23.71% in the leg muscles.

Download figureDownload PowerPoint

There were no differences between the athletic groups concerning height and weight (Table 1). Yet, swimmers were significantly younger than the cyclists (P = 0.019). There were no differences in baseline deltoid carnosine concentration between the groups. Baseline carnosine concentration in the leg muscles (soleus + gastrocnemius) was significantly lower in the cyclists than in nonathletes (P = 0.028) and kayakers (P = 0.047).

In gastrocnemius muscle, the absolute increase in carnosine (expressed as arbitrary units) was significantly higher in the cyclists (0.082 ± 0.035) and the swimmers (0.076 ± 0.015), compared with the nonathletes (0.043 ± 0.026) and the kayakers (0.035 ± 0.014) (Fig. 3 and 4). The swimmers (0.107 ± 0.029) had a higher absolute increase in the deltoid muscle compared with cyclists (0.065 ± 0.040) and the nonathletes (0.065 ± 0.035) (Table 4). There were no differences between the groups in the soleus. For mean muscle carnosine in the legs (soleus and gastrocnemius), the cyclists (0.084 ± 0.027) and swimmers (0.082 ± 0.017) had a higher absolute increase than the nonathletes (0.051 ± 0.031). We found no significant correlations between training volume or years of training and increases in carnosine concentration with BA supplementation within or across athlete groups.

Download figureDownload PowerPoint

Download figureDownload PowerPoint

Table 4. Muscle carnosine concentration pre- and postsupplementation and absolute increase in muscle carnosine concentrations after beta-alanine supplementation in the nonathletes, cyclists, kayakers, and swimmers

Muscle	Pre	Post	Δ (postpre)
M. Soleus
Nonathletes	0.149 (±0.029)	0.209 (±0.055)	0.060 (±0.041)
Cyclists	0.098 (±0.032)	0.183 (±0.035)	0.085 (±0.023)
Kayakers	0.155 (±0.066)	0.216 (±0.041)	0.061 (±0.027)
Swimmers	0.140 (±0.033)	0.227 (±0.032)	0.087 (±0.023)
M. Gastrocnemius
Nonathletes	0.171 (±0.050)	0.214 (±0.052)	0.043 (±0.026)
Cyclists	0.121 (±0.043)*†	0.202 (±0.039)	0.082 (±0.035)*†
Kayakers	0.178 (±0.038)	0.214 (±0.030)	0.037 (±0.013)
Swimmers	0.157 (±0.042)	0.233 (±0.040)	0.076 (±0.015)*†
M. Deltoideus
Nonathletes	0.149 (±0.032)	0.214 (±0.036)	0.065 (±0.035)
Cyclists	0.142 (±0.036)	0.207 (±0.027)	0.065 (±0.040)
Kayakers	0.135 (±0.033)	0.223 (±0.028)	0.089 (±0.026)
Swimmers	0.131 (±0.039)	0.238 (±0.045)	0.107 (±0.029)*$

Data are means ± SD.

^* P < 0.05 vs. control group.

^† P < 0.05 vs. kayakers.

^$ P < 0.05 vs. cyclists.

In the kayakers, their deltoid muscle carnosine (0.089 ± 0.026) had a higher absolute increase than their leg muscles (soleus + gastrocnemius) (0.049 ± 0.019, P = 0.001). An opposite pattern was observed with the cyclists where the leg muscles (0.084 ± 0.027) showed a tendency to a higher absolute increase than the deltoid muscle (0.065 ± 0.040, P = 0.078).

Taking all the subjects together (n = 35), there was a negative correlation between baseline carnosine content and the absolute increase in muscle carnosine in the leg muscles (soleus + gastrocnemius) (P = 0.020; r = −0.404) and deltoid muscle (P = 0.005; r = −0.476) following BA supplementation. The lower the baseline muscle carnosine content was, the higher the increase in muscle carnosine, following supplementation. For each group individually, there was only a significant negative correlation between baseline carnosine content and the absolute increase in the deltoid muscle in the cyclists group (P = 0.010; r = −0.765).

DISCUSSION

The first goal of our study was to determine in which muscle of the upper body carnosine can be reliably measured via ¹H-MRS. For both the deltoid and triceps brachii muscles, the C2-peaks (at ∼8 ppm) in the proton MRS spectra were sufficiently large to allow quantification. We found relatively low CVs (6.6% for deltoid and 9.2% for triceps brachii) in the test-retest condition within the same day, indicating that the carnosine concentration in the arm can be reliably measured. When adding the biological variability (3–4 wk apart), the CV in the deltoid increased from 6.6% to 13.3%, which is in agreement with Baguet et al. (4), who found a variation of 9.8% in the soleus muscle and 14.2% in the gastrocnemius muscle, when measured several weeks apart, compared with 4.3% and 7.6%, respectively, when measured twice on the same day (19). This first part of the study gave us the opportunity to compare supplementation-induced changes in muscle carnosine concentrations within one subject in different parts of their body in the second part of the study. Due to the better reproducibility of the deltoid, we decided to exclude triceps brachii in study B.

A subsequent goal was to compare BA-induced muscle carnosine loading in arm vs. leg muscles in a nonactive control group. To date, the increase of carnosine concentration after BA supplementation has not yet been investigated in the upper body musculature, despite several reports of ergogenic effects in exercise types that entirely (27) or predominantly (24) depend on this part of the body. Our results indicate that BA supplementation is equally effective in raising carnosine concentrations in upper vs. lower body muscles in nonathletes. On the basis of the above results, we started measuring athletes from different sports disciplines to compare carnosine loading in specifically trained and untrained muscles. We selected three well-trained athletic populations: swimmers, kayakers, and cyclists. Kayakers almost exclusively train their upper body, cyclists the lower body, whereas swimmers train the entire body. It was hypothesized, on the basis of the positive effects of exercise training on muscle creatine loading (12, 21), that trained muscles would have a higher absolute increase in carnosine concentration compared with untrained muscles. In line with this hypothesis, the results of the current observational study clearly indicate that there is an effect of training on the carnosine loading, as we observed a nearly doubling of the increase in muscle carnosine content after supplementation in specifically trained muscles compared with the untrained muscles (77.95% vs. 42.88%, respectively). In soleus, we did not find a greater accumulation in cyclists compared with nonathletes and kayakers. Soleus is a tonic muscle that is recruited very frequently also in daily activities like walking and standing in nonathletes. Therefore, the difference in activity degree between the groups is smaller for soleus than for gastrocnemius. In the gastrocenemius, we found a higher absolute increase in carnosine concentrations in the swimmers and cyclists compared with the nonathletes. Vice versa, the deltoid of the swimmers had a higher absolute increase in carnosine concentration compared with the nonathletes. This was further supported when comparing loading after supplementation in different muscles within each athlete group: cyclists were able to accumulate more carnosine in their leg muscles (soleus + gastrocnemius) than arm (deltoid) muscles, whereas a reverse pattern was observed in kayakers. Even though we did not conduct a training intervention (only nutritional intervention) in the current study, these results confirm that trained muscles are more efficiently loaded with carnosine than untrained muscles (Fig. 4).

Our findings seem to be in contrast with Kendrick et al. (17) who found no difference in carnosine loading between a trained and untrained leg in a single-leg training study. However, it is worth noting that training volume (10 × 10 maximal knee extensions, 3–4 times per week) was limited in that study. The athletes in our study trained at least 8 h/wk in their specific sport. If we estimate the amount of contractions, the current study had ∼130 times the number of contraction cycles compared with Kendrick's study. Kendrick et al. (16) probably did not undertake enough contractions to really assess whether training had an effect on muscle carnosine synthesis. We suspect that the apparent contradiction between both studies is attributable to differences in training modalities, which should be subject to further investigation. The observed differences in carnosine loading in trained vs. untrained muscles could either be attributed to the acute effects of muscle contractile activity, i.e., the effect of the actual exercise during the supplementation period, or to the beneficial structural and metabolic properties of muscle induced by prior training. The current study design is not able to distinguish between training status vs. exercise effects. However, an advantage of the current study design is that—similar to the one-leg training studies—the comparison occurs within one subject, ensuring identical nutritional and environmental factors, and identical circulating hormone and BA concentrations. Yet in contrast to the one-leg training studies, effects of larger training load and volumes can be investigated in the currently adopted design.

Although this is still the first study to document a training effect on muscle carnosine loading, it is tempting to speculate on its underlying physiological mechanisms. A possible explanation for the acute effect of exercise training is the increased blood flow in contracting muscles, which results in better BA delivery to the muscle cells. Another possibility is that there is a contraction-induced stimulation of transporters such as TauT and PAT1, which are expressed in human skeletal muscle (9), to take up BA in the myocyte. This would be in accordance with other transsarcolemmal metabolite transporters, such as the glucose and fatty acid transporter (GLUT4 and FAT/CD36), which are recruited by a contraction stimulus (16, 18). On the other hand, if it is the result of the training status, rather than the contractile activity, it could be related to the endurance training-induced increases in the capillary density (20), which ensures also better BA delivery in the trained muscles. Consequently, this same training-induced alteration can also increase the expression of transporters involved in BA uptake and enzymes involved in carnosine synthesis (9). However, all the proposed explanations are speculative at present.

From a practical point of view, this scientific finding can be translated into attractive implications to athletes. It seems wise to supplement BA during a period of substantial training volume, rather than in a rest or recovery period, in order to optimize supplementation effectiveness throughout a training season. In addition, the physiological carnosine loading mechanism is probably more effective in trained vs. untrained individuals. This is opposite to some other ergogenic supplements, where biological activity and effectiveness is often less pronounced in a trained population.

In summary, we managed to measure carnosine reliably and noninvasively in a single muscle of the upper body (deltoid) by ¹H-MRS. We also showed that muscle carnosine loading via BA supplementation is of the same magnitude in arm and leg muscles in a nonactive control group. Finally, we demonstrated that carnosine loading is more pronounced in the trained vs. untrained muscles of athletes. These findings suggest that training is a possible determinant of carnosine loading, but it remains to be determined whether these effects are due to the acute exercise effects and/or to chronic adaptations of training.

GRANTS

This study was financially supported by grants from the Research Foundation - Flanders (FWO G.0243.11 and G.0352.13N).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: T.B. and W.D. conception and design of research; T.B. and J.S. performed experiments; T.B., J.S., and E.A. analyzed data; T.B., J.S., and W.D. interpreted results of experiments; T.B. prepared figures; T.B. and W.D. drafted manuscript; T.B., W.C., A.B., and S.S. edited and revised manuscript; T.B. and W.D. approved final version of manuscript.