The Presence of Hyperhomocysteinemia Does Not Aggravate the Cardiometabolic Risk Imposed by Hyperuricemia in Young Individuals: A Retrospective Analysis of a Cross-Sectional Study

Background: Little research has been conducted into the effects of the combined manifestation of hyperuricemia and hyperhomocysteinemia on cardiometabolic risk factors and markers in young subjects. Methods: 1298 males and 1402 females, 14-to-20-year-olds, were classified into four groups: 1/normouricemic/normohomocysteinemic, 2/normouricemic/hyperhormohomocysteinemic, 3/hyperuricemic/normohomocysteinemic, and 4/hyperuricemic/hyperhomocysteinemic. Anthropometric measures, blood pressure, plasma glucose, insulin, lipids, markers of renal function, C-reactive protein, asymmetric dimethylarginine, and blood counts were determined. Results: Hyperuricemic males (but not females) had higher odds for hyperhomocysteinemia than normouricemic ones (OR: 1.8; 95% CI: 1.4–2.3; p < 0.001). Homocysteine and uric acid levels correlated directly (males: r = 0.076, females: r = 0.120; p < 0.01, both). Two-factor analysis of variance did not reveal a significant impact of hyperhomocysteinemia on any of the investigated cardiometabolic variables in females; in males, hyperuricemia and hyperhomocysteinemia showed a synergic effect on asymmetric dimethylarginine levels. Among four groups, subjects concurrently manifesting hyperuricemia and hyperhomocysteinemia did not presented the highest continuous metabolic syndrome score—a proxy measure of cardiometabolic risk; neither the multivariate regression model indicated a concurrent significant effect of uric acid and homocysteine on continuous metabolic syndrome score in either sex. Conclusion: In young healthy subjects, hyperhomocysteinemia does not aggravate the negative health effects imposed by hyperuricemia.


Introduction
Uric acid (UA) is a bioactive end-product of the metabolism of purines (adenine and guanine). Extracellular UA acts as an antioxidant; intracellularly, UA exerts pro-oxidant effects [1]. Experimental studies document that UA affects vascular cell functions via promotion of degradation of the vasodilator nitric oxide and boosting of activity of the reninangiotensin axis [2]. Both UA and xanthine oxidoreductase (the enzyme that generates uric acid) may induce oxidative stress, promote inflammation, endothelial dysfunction, and atherosclerosis [2,3]. Hypertension, kidney, and cardiovascular diseases, fatty liver, dyslipidemia, obesity, insulin resistance, metabolic syndrome (MetSy), or diabetes are associated with elevated serum UA (SUA) levels already in adolescents [4][5][6].
Homocysteine (Hcy) is a sulfur-containing amino acid produced from the essential amino acid methionine as a byproduct of trans-methylation reactions. It derives from S-adenosyl-homocysteine hydrolase-catalyzed hydrolysis of S-adenosylhomocysteine to adenosine. Experimental studies show that at the molecular level, the toxicity of Hcy includes mechanisms involving the formation of reactive oxygen species, hypomethylation, induction of unfolded protein response, and protein homocysteinylation. At the cellular level, Hcy is a pro-inflammatory, pro-thrombotic, pro-atherogenic factor, vasodilation impairing agent, and an inducer of endoplasmatic reticulum stress [7]. Even in children and adolescents, increased levels of Hcy are associated with a range of disorders, such as renal and cardiovascular diseases, obesity, diabetes, premature atherosclerosis, or impaired bone health [8][9][10][11].
Uric acid and Hcy are metabolically interrelated: adenosine produced during metabolic transit of Hcy might eventually be metabolized into UA. Most daily UA and Hcy disposal occur via kidneys. Both UA and Hcy may affect the oxidative status, exert proinflammatory and proatherogenic effects, and may alter vascular cell function via interfering with NO metabolism. Levels of SUA and Hcy rise in identical pathologies, and both variables display a positive direct relationship even in healthy adults [12][13][14][15][16][17]. A synergistic association of hyperuricemia and hyperhomocysteinemia (hyperHcy) with chronic kidney disease has been documented in middle-aged and elderly patients [18]; while combined hyperuricemia and hyperHcy additively increased the risk of manifestation of subclinical atrial fibrillation in patients with cardiac implantable electronic devices [19]. However, it is not fully clarified whether hyperuricemia and hyperHcy represent markers, or rather act as etiological agents in cardiovascular and other mentioned pathologies [20,21]. The current guidelines of professional societies do not consider SUA or Hcy as cardiovascular disease risk stratifiers.
The recent cross-sectional analysis indicated that in adolescents, Hcy levels were positively correlated with those of SUA, and the odds to present elevated SUA levels increased across the Hcy terciles [22]. Yet, it remains unclear whether the combined manifestation of elevated SUA levels and hyperHcy exerts additive effects on cardiometabolic risk factors and markers in young healthy subjects. This question is of particular importance, as there is robust evidence linking cardiovascular risk factors in childhood and adolescence (even in juveniles with mildly elevated risk factor scores) with clinical atherosclerotic cardiovascular disease events in adulthood [23].
We hypothesized that concurrent manifestation of elevated SUA levels and hyperHcy exerts an additive worsening effect on cardiometabolic risk factors or markers compared with the presence of isolated hyperuricemia or hyperHcy. To this point, we retrospectively analyzed data obtained from healthy young individuals.

Subjects
The cross-sectional "Respect for Health" study has been described previously [22]. Briefly, students of state secondary schools in the Bratislava Region participated voluntarily in the survey. Exclusion criteria were any acute or chronic illness, pregnancy, or lactation in females.
The study was approved by The Ethics Committee of the Bratislava Self-governing Region and conformed to the Helsinki Declaration. In minors, participation was subject to the written informed consent of the legal representative and the verbal assent of the child. Written informed consent was obtained from full-aged participants.

Measurements
Anthropometric measurements were performed following standard guidelines, as described previously [24]. Briefly, height was measured using a portable extendable stadiometer, waist circumference using a flexible tape, and body weight employing digital scales (Omron BF510, Kyoto, Japan). BMI and waist-to-height ratio (WHtR) were calculated.
Blood pressure (BP) was measured on a dominant arm in a person relaxed for at least 5 min in the seated position, using a digital monitor (Omron M-6 Comfort, Kyoto, Japan). The mean of the last two measurements was recorded.

Definition of Elevated Uric Acid Levels, Hyperuricemia, Hyperhomocysteinemia, Cardiometabolic Risk Factors, and Metabolic Syndrome
Employing reference ranges of the laboratory of the National Institute for Children's Diseases in Bratislava, we classified hyperuricemia as SUA concentration >340 µmol/L in females; in males >360 µmol/L if aged < 17 years, and >420 µmol/L in those aged ≥ 17 years; individuals aged 15 and 16 years who presented with Hcy > 10.0 µmol/L, adolescents aged 16 and 17 years who displayed Hcy > 11.3 µmol/L, and those aged ≥ 18 years with Hcy concentrations >15.0 µmol/L were classified as hyperhomocysteinemic.
In 14-to-17-year-olds, general overweight/obesity was classified according to the international age-and sex-specific cutoff points for BMI [29], in individuals aged ≥ 18 years as BMI ≥ 25 kg/m 2 . Central obesity was defined as WHtR ≥ 0.5 [30]. The increased cardiometabolic risk was classified according to guidelines for the classification of metabolic syndrome components: systolic BP ≥ 130 mmHg, DBP ≥ 85 mmHg, triacylglycerols ≥ 1.7 mmol/L, fasting glycemia ≥5.6 mmol/L, and HDL-C as <1.03 mmol/L in males and females aged < 16 years; and <1.29 mmol/L in females aged ≥ 16 years [31]. Subjects presenting at least three cardiometabolic risk factors, e.g., central obesity, elevated fasting glycemia, BP, TAG, or low HDL-C concentrations were considered as suffering from metabolic syndrome [31]. Moreover, fasting insulin ≥20 µIU/mL [32], CRP > 3 mg/L [33], atherogenic index ≥0.11 [26], and the presence of microalbuminuria/albuminuria (urinary albumin/creatinine ≥2.5 mg/mmol in males and ≥3.5 mg/mmol in females [24]) were considered as markers of increased cardiometabolic risk. The number of cardiometabolic risk factors (RF) was calculated as a sum of the presence of binary coded elevated BP, adiposity (presence of central obesity or general overweight/obesity), dyslipidemia (elevated TAG or AIP or low HDL-C), alteration of glucose metabolism (elevated fasting glycemia or insulinemia), and elevated CRP.

Statistical Analyses
Data not fitting the normal distribution (Shapiro-Wilk test) were log-transformed before statistical analyses. Males and females were compared using the two-sided independent samples Student's t-test. According to the reference ranges of SUA and Hcy, subjects were classified into four groups: 1/concentrations of SUA and Hcy below the upper reference range (e.g., normal, n), 2/SUA levels below and Hcy above the upper reference range, 3/SUA above and Hcy concentration below, and 4/both SUA and Hcy concentrations above the reference ranges. Four groups were compared using the twofactor analysis of variance (ANOVA) with the presence/absence of hyperuricemia and presence/absence of hyperHcy as fixed factors. Normally distributed data are given as mean ± standard deviation (SD), those failing assumptions of normality are described with a back-transformed geometric mean (interval −1 SD, +1 SD). Categorical data were compared using the Fisher's exact test or Chi-square test with Yates correction (Y) if appropriate, and are given as counts and frequencies. Pearson correlation coefficients and odds ratios (OR) were calculated. p value < 0.05 was considered significant. Statistical software SPSS version 16 (SPSS, Chicago, IL, USA) was used.
Multivariate regression of independent factors on continuous metabolic syndrome score was performed using the orthogonal projection to latent structures model (OPLS, Simca v.16 software, Sartorius Stedim Data Analytics AB, Umea, Sweden). In Model 1, age, insulin, CRP, eGFR, ACR, leukocyte (WBC) and erythrocyte (RBC) counts, and ADMA, were entered as independent variables; in Model 2, SUA was forced into the model; in Model 3 SUA was replaced by Hcy; in Model 4 both SUA and Hcy were entered. Before fitting the OPLS models, all variables with high skewness and a low min/max ratio were log-transformed and all data were mean-centered. Variables with a variable of importance for the projection (VIP) values ≥ 1.00 were considered significant.

Results
Cohort characteristics are given in Table 1. Males differed from females in all variables except for age, QUICKI, the prevalence of elevated fasting insulinemia, elevated TAG levels, or microalbuminuria/albuminuria.

Correlations between Cardiometabolic Risk Factors and Markers with Uricemia or Homocysteinemia
Correlations between age or glycemia and SUA were insignificant; significant inverse correlations were revealed between SUA and HDL-C, QUICKI, ln urinary albumin/creatinine, and eGFR; all other variables showed a direct significant relationship with SUA (Table 2). Age, DBP, non-HDL-C, AIP, and lnTAG showed a positive significant relationship with lnHcy; while eGFR, ln urinary albumin/creatinine, and lnADMA correlated inversely. However, all significant associations were weak (Pearson r: SUA: −0.062-to−0.316, lnHcy: −0.056-to−0.094; Table 2).

Multivariate Regression Models
The OPLS multivariate regression model was employed to elucidate whether and how SUA and/or Hcy affect the continuous metabolic syndrome score-a proxy measure of cardiometabolic risk. If neither SUA nor Hcy was considered, the model indicated that insulinemia and inflammatory markers independently affect the continuous metabolic syndrome score (Table 4, Model 1). After the inclusion of SUA, it became an additional significant predictor of cardiometabolic risk, regardless of the absence (Model 2) or presence (Model 4) of Hcy in the model. However, the variance of the continuous metabolic syndrome score explained by models 2 and 4 increased only slightly (by 3%) after the inclusion of SUA. After forcing Hcy into the model, erythrocyte counts but not Hcy became a significant predictor of continuous metabolic syndrome score (Models 3 and 4). The variability of the continuous metabolic syndrome score explained by the models was not affected.

Multivariate Regression Models
The multiple regression model not adjusted for SUA and Hcy selected insulinemia, inflammatory markers, and ADMA as significant independent predictors of continuous metabolic syndrome score (Model 1; Table 4). Forcing SUA (Model 2), Hcy (Model 3), or their combination (Model 4) into the model neither affected the selection of independent variables modulating the continuous metabolic syndrome score, nor its variability explained by the OPLS model, which was generally low.

Discussion
Retrospectively, we tested the hypothesis that in healthy young individuals, the concurrent presence of hyperuricemia and hyperHcy is associated with less favorable cardiometabolic status in comparison with the manifestation of only one of them. We did not confirm our hypothesis. In males, two-factor ANOVA indicated that out of nineteen investigated risk factors and markers, hyperuricemia significantly affected fifteen; hyperHcy only five. Paradoxically, hyperHcy was associated with lower measures of obesity, lower SBP, number of manifested risk factors, and CPR concentration regardless of the absence or presence of hyperuricemia. The synergic effect of hyperuricemia and hyperHcy was observed only for lnADMA. In females, hyperuricemia was associated with worsening of eleven cardiometabolic risk factors and markers; while none of the endpoints was affected significantly by hyperHcy. No significant interaction between hyperuricemia and hyperHcy was observed in either sex.
Numerous studies in the general population of adults indicate that SUA and Hcy levels show a linear positive correlation [12][13][14][15]17]. In line with the recent study on teenagers [22], we show that this direct relationship is manifested already in young healthy subjects-a population not affected by age-associated comorbidities. Similar to other studies [14,19,34], this correlation was tighter in females compared with males, despite that females generally present with lower levels of SUA and Hcy, and a lower prevalence of cardiometabolic risk factors and markers compared with males. As in the aforementioned studies, correlations between SUA and Hcy were weak. This evokes a question of whether such weak statistical correlations might be of clinical impact.
Associations between uricemia or homocysteinemia and variables characterizing cardiometabolic risk are widely studied in different populations. In the general population of adolescents, rising SUA concentrations go hand in hand with worsening of the components of MetSy, other cardiovascular disease indicators, such as inflammatory markers or glomerular filtration rate, as well as increased cardiometabolic risk [35][36][37][38]. With some minor sex differences, several cardiometabolic risk factors and markers worsened, and the number of cardiometabolic risk factors and cardiometabolic risk (evaluated as continuous metabolic syndrome score) increased significantly with increasing SUA levels also in our study.
The reports on whether variables characterizing cardiometabolic status worsen with increasing Hcy levels in adults are inconsistent [12][13][14][15][16][17]. Large studies in adolescents show that neither MetSy nor the rising number of its components is associated with increased Hcy levels [39,40]. In our study, all significant simple correlations indicated a less favorable cardiometabolic status with increasing lnHcy in both sexes, albeit significant correlations were less frequent compared with those observed for SUA. However, two-factor ANOVA indicated a concurrent significant impact of hyperuricemia and hyperHcy only in males; and in five out of six cases-BMI, WHtR, SBP, risk factors number, and lnCRP-the presence of hyperHcy paradoxically associated with a partial amelioration of the negative effect of hyperuricemia. The cross-sectional nature of our study does not allow for commenting on potential mechanisms. We are not aware of similar reports in the literature; thus, our findings open the field for further research.
Although the combination of hyperuricemia and hyperHcy was associated with the highest ADMA levels among the four groups, the effect was significant only in males; and was synergic, not additive. Reports on the association of SUA and ADMA in the general population are scarce. The Polish study reported elevated ADMA levels in adolescents with hyperuricemia [41]. The hyperhomocysteinemia-associated rise in ADMA might reflect the interconnection of their metabolism. S-adenosylmethionine methyltransferases and proteinarginine methyl transferases participate simultaneously in Hcy and ADMA synthesis and hyperHcy may increase ADMA levels by reducing the activity of dimethylarginine dimethylaminohydrolase-an enzyme that metabolizes ADMA [42,43]. Moreover, ADMA is eliminated also by renal excretion. Thus, some of the deleterious effects of hyperHcy may involve ADMA-related cardiovascular effects. ADMA acts as a competitive inhibitor of NO synthase and may cause a further decrease in the bioavailability of NO by increasing the production of reactive oxygen species [44]. Elevated serum ADMA is associated with MetSy, endothelial dysfunction, and cardiovascular diseases such as hypertension and atherosclerosis [45].
In our adolescents, the prevalence of combined hyperuricemia and hyperHcy was low, particularly in females. Thus, the study could be underpowered concerning the ability of two-factor ANOVA to detect an additive effect of both factors. To this point, we used multivariate regression to test whether and how uricemia and/or homocysteinemia affect the continuous metabolic syndrome score. However, only SUA appeared as a significant independent predictor, and only in males; and the addition of both biomarkers into baseline prediction models did not improve their prediction abilities for the continuous metabolic syndrome score in either sex.
There are several limitations associated with this study. This is a retrospective analysis of a cross-sectional study; thus, a causal relationship cannot be inferred. The measurements were taken at a one-time point. Our results cannot be generalized to different populations. Information on circulating levels of vitamins essential for Hcy metabolism, genetics, lifestyle factors, or dietary habits that could potentially affect SUA or Hcy levels were not available. However, the study in Slovak adolescents indicated that 677 C → T mutation of the methylenetetrahydrofolate reductase gene, one of the most frequent genetic causes of moderate hyperHcy, was not associated with increased Hcy levels [46]. Regarding diets, daily intake of vitamin B12 and folate in these adolescents exceeded the recommended daily allowance, while that of vitamin B6 reached only 60-66%, thus could contribute to hyperHcy. On the other hand, to our knowledge, this is the largest study exploring the combined effects of hyperuricemia and hyperHcy on several cardiometabolic risk factors and markers in apparently healthy young individuals. The number of participants allowed for a separate evaluation of both sexes thus pointing out physiological sex disparities which may remain undetected if the sexes are not analyzed apart.

Conclusions
In our young healthy subjects, presence of combined hyperuricemia and hyperHcy was not associated with worse cardiometabolic status compared with that imposed by isolated hyperuricemia. While speculative, there might be several potential explanations. First, the relationship of SUA and Hcy with cardiometabolic risk variables rather reflects their metabolic interconnection than the pathophysiological link. It is also possible that in young healthy individuals SUA or Hcy may not be as specific as the conventional cardiometabolic risk factors concerning cardiovascular risk prediction. Our data do not rule out that there are (even non-clinical) populations where the combined hyperuricemia and hyperHcy associate with an increased cardiometabolic risk-thus, it requires validation in diverse cohorts. If combined risk is confirmed in certain populations, it remains to be elucidated whether successful intervention restoring SUA and Hcy to normal levels concomitantly reduces cardiometabolic risk. Moreover, potential sex-specific disease risk in later life imposed by combined hyperuricemia and hyperHcy should be evaluated.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki, and approved by the Ethics Committee of the Bratislava Self-governing Region, Bratislava, Slovakia, on 19 April 2011. Informed Consent Statement: Written informed consent was obtained from all full-aged participants, in all minors from their parents/guardians after the study purpose and procedures had been explained.

Data Availability Statement:
The data that support the findings of this study are available from the corresponding author upon reasonable request.