In a cohort of apparently healthy adults, serum GAA, and creatine concentrations had significant associations with several cardiometabolic risk factors. Higher GAA levels were associated with an unfavorable cardiometabolic risk profile (including higher fasting insulin, higher total homocysteine, and higher body fat percentage), while participants with elevated serum creatine levels (≥31.1 µmol/L) had greater odds of being overweight. This suggests a possible role of GAA–creatine axis in the pathogenesis of cardiovascular and metabolic disease.
Our findings were consistent with previous studies reporting a link between GAA and cardiometabolic disorders. Elevated serum GAA levels were associated with higher tHcy levels [11
] and insulin hypersecretion [12
], unfavorable risk factors for cardiometabolic diseases. We found that tHcy is a very strong predictor (β
= 0.30, p
< 0.001) and so it could be that tHcys is the root issue, either as an independent factor or through affecting arginine availability, since GAA is synthesized from arginine [1
]. Hypothetically, enhanced GAA synthesis might restrain the availability of arginine, an amino acid that has been shown to strongly affect the risk factors of cardiovascular diseases in humans [13
]. However, previous reports about the associations between serum creatine and cardiovascular and metabolic risk factors are not in line with our results. In several small-scale studies, higher serum creatine levels (as provoked by oral intake) were associated with a favorable risk-factor profile, including lower tHcy [14
], reduced total cholesterol [15
], or improved insulin sensitivity [16
]. In contrast, a number of studies found no significant relationships between creatine alone and different cardiometabolic markers (for review see [10
]), and a recent large study (n
= 622) suggested no association between serum creatine and tHcy levels in apparently healthy men and women [17
]. The conflict between our study and results from previous creatine studies may have been partially due to whether the analyses were based on serum levels or dietary intakes of GAA and creatine. The above studies typically linked creatine levels with cardiometabolic risk factors during an exercise program, an intervention known to reduce cardiometabolic risks per se [18
], while we evaluated a diverse population not currently involved in an exercise program or dietary intervention. Our findings are in accordance with previous trials reporting an association between higher BMI and higher serum creatinine, an end-product of creatine metabolism [19
], suggesting a link between the GAA–creatine axis and risk of being overweight. Hypothetically, GAA overload (as indicated by higher levels of serum GAA) may perturb creatine metabolism, thereby resulting in enhanced creatine synthesis that accounts for an equivalent proportion of tHcy production and cardiometabolic burden. We found that serum GAA is positively associated with insulin levels (β
= 0.53), while no such relationship was found between serum creatine and insulin levels. Having higher GAA concentrations appears to be accompanied by elevated levels of insulin circulating in the blood, suggesting a possible link between excess GAA and hyperinsulinemia, a well-known cardiometabolic risk factor. Our results corroborate a previous in vitro study showing that GAA stimulates insulin release more potently than creatine, by triggering insulin secretion via kinase-sensitive mechanisms in pancreatic islets cells [12
]. However, before putting forward serum GAA as a proxy for insulin release, additional studies are highly warranted to analyze GAA dynamics in populations with normal and impaired insulin secretion and action. In addition, creatine synthesis from GAA might affect cardiovascular risk through depletion of arginine, a precursor of creatine and a key source of nitric oxide (NO) [20
]. NO plays an important role in the protection against the onset and progression of cardiovascular disease [21
], and creatine overload may induce disturbances in NO bioavailability leading to a loss of the cardioprotective actions and in some populations may increase disease progression [22
]. Therefore, synchronized monitoring of GAA-creatine and arginine-NO axes in future studies should provide more mechanistic evidence by which creatine synthesis affects cardiovascular risk.
Baseline levels of GAA and creatine were in accordance with previous studies reporting reference values for these two compounds [23
], and gender-related differences for serum creatine found in the present study could be attributed to the effects of testosterone and estrogen on creatine synthesis and transport. The CK activity, which regulates the use and consumption of creatine, is found to be sex-specific [26
]. Gender differences in muscle mass and responses to downstream signaling pathways in the skeletal muscle might also affect creatine utilization [27
]. In particular, sex hormones appear to differently affect creatine transporter (SLC6A8) expression, with SLC6A8 inhibited by estrogens and stimulated by testosterone [28
]. This has been suggested to provoke a higher leakage of body creatine in the urine of healthy women [25
], with perhaps less creatine retained in the blood, as we found in the present study. In addition, we found that overweight adults had higher creatine concentrations (by 13.6% on average) compared to adults of a normal weight, and the odds ratio of being overweight was 3.26-fold higher in participants with a serum creatine ≥ 31.1 µmol/L. Previous studies confirm our findings, with a positive association having been reported between serum creatinine and BMI in healthy men and women [19
]. The origin of creatinine from creatine may explain this correlation and the higher concentrations in men than women [30
]. However, the strong positive correlation found between serum creatine levels (also GAA) and body fat percentage in our cohort remains puzzling. This perhaps means that circulating values of these compounds are related not only to body size but also to body composition and other (patho) physiological mechanisms that need to be revealed.
The major limitation of the present study is its cross-sectional design that prevented causal conclusions between the GAA-creatine axis and cardiometabolic risk factors. Second, we recruited a specific cohort of participants (predominantly young normal-weight adults, with no major biochemical disturbances) that limited the interpretation of our results to a healthy population while being unable to generalize the results to patients with cardiometabolic diseases. Third, we used only a finite compendium of biochemical variables to understand a more complete picture of GAA-creatine metabolism. Fourth, no link has been established between GAA and creatine levels and clinical indicators of cardiometabolic diseases, including markers of vascular pathology or metabolic overload. Finally, no skeletal muscle mass was profiled and accounted for in the regression analysis. Since skeletal muscle is a major site of insulin resistance [31
], with creatine utilization found to be different among type I and type II muscle fibers [32
], controlling for muscle mass in future studies could help to better address the link between biomarkers of GAA-creatine metabolism and cardiometabolic risk.