Homocysteine, folate, hs-C-reactive protein, tumor necrosis factor alpha and inflammatory proteins: are these biomarkers related to nutritional status and cardiovascular risk in childhood-onset systemic lupus erythematosus?

Cardiovascular disease (CVD) is a leading cause of morbidity and mortality in adult SLE and has been associated with chronic inflammation [1]. Although there has been a significant improvement in all-cause mortality rates in adult SLE patients over time, the mortality secondary to atherosclerosis and cardiovascular disease has not significantly changed [2]. One study reporting the long-term burden of CVD in c-SLE, demonstrated that patients with onset of SLE during childhood had a similar incidence of myocardial infarction (MI) compared to subjects with adult-onset disease (7.9% versus 4.5%, respectively; p = 0.23), and also had the first MI earlier in life (mean of 32 years old) [3].

The incidence of acute myocardial infarction and risk of premature atherosclerosis is approximately 50 times greater in women with SLE compared to healthy controls. This association is the result of the complex interaction between classic SLE risk factors, chronic inflammation, elevated concentrations of homocysteine, the production of autoantibodies, and SLE treatment [4].

Elevated homocysteine (Hcy) levels are associated with increased risk for CVD and can indicate micronutrient insufficiencies especially cobalamin and/or folate which are required cofactors for Hcy metabolism through the remethylation pathway [5]. Inflammatory biomarkers such as high-sensitivity C-reactive protein (hs-CRP) [6] and tumor necrosis factor – α (TNF – α) [7] have been consistently associated with the presence of CVD in multiple studies in different populations including individuals with SLE.

The status of the inflammatory system has been analyzed in multiomic studies which included associations of gene variants (genomics), mRNA (transcriptomics), protein (proteomics), and metabolites (metabolomics) [8]. Until now, only one study analyzed selected plasma proteins in c-SLE [9]. The present study describes and compares anthropometric measurements, food intake, levels of folate, plasma proteins, and the cardiovascular risk factors homocysteine, TNF-α, hs-C reactive protein in c-SLE and healthy controls. These cardiovascular risk factors were correlated with nutritional status, lipid profile, and duration of corticosteroid therapy in c-SLE. We tested the hypothesis that nutritional status, lipid and proteomic profile, homocysteine, folate, hs-CRP and TNF-α levels would be consistent with CRF in c-SLE compared to the levels in healthy controls.

Twenty c-SLE subjects who met the inclusion and exclusion criteria were recruited and nineteen agreed to participate completing all follow-up procedures in the study. Forty-five healthy controls who met inclusion and exclusion criteria were recruited and thirty-eight agreed to participate in the study.

c-SLE and control groups were matched for age (c-SLE 15.7 ± 1.95 years old; Control 14.9 ± 2.5 years old, p = 0.117) and pubertal status (breast p = 0.158 and pubic hair p = 0.578). The age at diagnosis in lupus group was 10.8 ± 2.7 years old and mean disease duration until data collection day was 6.3 ± 2.6 years old. The median SLEDAI 2 k in c-SLE participants was 4 (range 0–29); 16/19 (84.2%) had active disease, with 4 having important activity (> 12), 3 with a moderate activity (5–12), 9 with mild activity (2–4), and 3 were classified as not having an active disease (equal to zero) [21]. The median SLICC/ACR-DI score was 0 (range 0–9) and six patients had a score of one or higher. The mean duration of steroid use was 5 years. Fifteen out of nineteen subjects used prednisone in the last 6 months prior data collection (0.36 ± 0.32 mg/kg). Systolic and diastolic blood pressure levels were not different between lupus and controls (systolic: 105.51 ± 2.26 versus 102.80 ± 1.61, p = 0.354; diastolic: 68.5 ± 2.19 versus 69.00 ± 1.56, p = 0.860).

BMI, waist circumference, Hcy, TNF- α, hs-CRP, and triglycerides levels were higher in participants with c-SLE, while serum folate was significantly lower (Table 1). No significant differences in energy (lupus 1566 ± 76 Kcal/day vs controls 1594 ± 52 Kcal/day, p = 0.766) or folate intake (lupus 83.91 ± 9.64 μg/day vs controls 77.30 ± 6.59 μg/day, p = 0.588) were found. Vitamin B12 intake was significantly higher in c-SLE participants compared to controls (lupus 3.14 ± 0.17 μg/day vs controls 2.65 ± 1.8 μg/day, p = 0.029). Anthropometric and biochemical parameters were also tested between control group, LCBMP, and LCWMP (Table 2) and as expected, all parameters related to CVD were statistically higher in lupus clusters compared to controls, while folate level was lower.

iTRAQ results identified 86 proteins in our total sample (Control, LCBMP, LCWMP) but after correcting for lysozyme and internal controls using Scaffold software, the abundance of only 10 proteins were statistically different between the three groups. These 10 proteins were involved in inflammatory processes and were more abundant in the c-SLE group (LCBMP) with the worse metabolic parameters (Table 3).

Spearman test showed no association between CVD risk factors and nutritional status parameters (data not shown). TNF-α was positively correlated with triglycerides in lupus group (r = 0.519, p = 0.023). The levels of cardiovascular parameters total cholesterol (r = 0.575, p = 0.025) and triglycerides (r = 0.707, p = 0.003) were correlated with dosage of corticosteroid even adjusted by Bonferroni, hs-CRP was positively correlated with duration of corticosteroid therapy (r = 0.626, p = 0.009).

We confirmed the hypothesis that childhood-SLE have worse nutritional status based on lipid and proteomic profiles, homocysteine, folate, hs-CRP, and TNF-α levels when compared to healthy controls. In the present study, SLE patients showed higher BMI and WC compared to controls and with increased levels of cardiovascular risk factors (homocysteine, TNF- α, hs-CRP, and folate levels). High BMI values have been associated with increased CVD risk factors including hs-CRP in lupus, which contributes to an increase in morbidity and mortality in these patients [22, 23]. Cross sectional studies in children suggest that increased BMI is associated carotid intima and media thickness, and left ventricular mass in early or mid-adulthood [24]. Truncal obesity is a risk factor for coronary heart disease, which is frequently seen in SLE and is associated with the development of atheroma and metabolic syndrome [25]. Waist circumference (WC) is the best overall predictor of abdominal visceral obesity [24]. High waist circumference values were independent predictors of increased carotid intima media thickness and metabolic syndrome in adults with lupus (Rizk 2012).

The childhood SLE patients in this population had higher homocysteine and lower folate levels when compared to controls, which corroborated results of other studies [26]. Hyperhomocysteinemia has been identified as a risk factor for atherothrombotic events and has been related to impaired 5,10-mehtylenetetrahydrofolate reductase activity and deficiency of folate in patients with SLE, including children and adolescents [21, 26‐29].

Dyslipidemia may be present in over half of the adult patients with lupus and CVD [23] and caused by multiple factors. Cytokine, autoantibodies, medications, dietary intake, renal disease, physical activity, and genetic factors are all likely important contributors [2]. Individuals at risk of CVD characteristically have higher levels of triglycerides and very-low-density lipoprotein cholesterol but lower levels of HDL [2, 22]. Triglycerides levels were higher in these childhood SLE patients compared to controls and were positively correlated with TNF-α, a well-known risk factor for CVD. In addition, total cholesterol and triglycerides were correlated with dosage of corticosteroid.

The autoimmunity and the inflammatory process of SLE are directly related to changes in lipid profile and metabolism of lipoproteins. Regardless of disease activity, SLE by itself and corticosteroid use promote pro-atherogenic lipoprotein profiles [22]. Some authors found a positive association between corticosteroid and serum triglycerides in adult-onset SLE [30], same results were found in our study.

The present study found a significant positive correlation between TNF-α and triglycerides. High cytokine levels such as TNFα (a non-specific pro-inflammatory factor may play a role in the dysregulation of lipid levels in pediatric SLE [7, 31]. TNF-α acts as mediator to inhibit lipoprotein lipase resulting in elevated levels of triglyceride, including those individuals with SLE [7]. Another inflammatory marker correlated with CRF is hs-CRP [32]. Some studies have found higher levels of hs-CRP in pediatric SLE patients compared with healthy controls [33]. High hs-CRP levels were correlated with functional and morphological alterations in the cardiovascular system [33, 34]. To our knowledge no study has found correlations of the use and duration of corticosteroid in childhood-onset systemic lupus erythematosus with hc-CRP. The positive correlation between hs-CRP level with duration of corticosteroid therapy found in our study is controversial [35].

Although folate consumption was not different between groups and not statistically correlated with homocysteine, folic acid serum levels were significantly lower in c-SLE subjects, which has not been previously reported in the literature. Lower folate serum levels were found in adult SLE patients and can be associated with hyperhomocysteinemia [26].

Individuals in the LCWMP group had lower levels of anti-inflammatory proteins, apolipoprotein A-I, apolipoprotein E, alpha-2-macroglobulin, alpha-1-antitrypsin, ceruloplasmin, complement C3, hemopexin and serum transferrin compared with healthy control group and with LCBMP group. Apolipoprotein A-I (apo A-I) acts in cholesterol homeostasis and inhibits pro-inflammatory cytokines decreasing risk of cardiovascular disease [9]. Similar results were found by others who described lower plasma abundance of apo A-I in pediatric patients with SLE when compared to healthy controls [9]. An pro-atherogenic profile characterized by decreased plasma levels of apo A-I was previously observed in adolescents with SLE [36]. The decreased abundance of apolipoprotein A I, apolipoprotein E, α-2-macroglobulin, α-1-antitripsin, ceruloplasmin, complement C3, serotransferrin, haptoglobin, hemopexin and serum transferrin was considered a CRF [9, 37‐43]. Our study is the first one to describe the decreased abundance of these 10 proteins as potential early biomarkers of CRF in childhood-onset systemic lupus erythematosus.

The present study has some limitations. The cross-sectional design makes it difficult to establish a cause-and-effect relationship between CRF and nutritional status, lipid profiles, and inflammatory proteins. In addition, the sample size was small. The convenience sampling method often cannot be extrapolated to give population results and may be prone to volunteer bias. Hence, the conclusions stated in the study were limited to this specific group and cannot be extrapolated even though all participants shared a similar built environment. The proteomic method used here could not evaluate each sample individually due to cost considerations. However, the technique of grouping samples as done in the present study has been used in several published reports [44, 45]. Given these limitations, the present study is the first to identify and characterize distinct metabolic groups of cSLE patients, one of which had a proteomic profile consistent with increased risk of CVD when compared to healthy controls.

The present study was the first one to confirm the hypothesis that children and adolescents with childhood-onset systemic lupus erythematosus have a less healthy nutritional status, lipid and proteomic profile, homocysteine, folate, hs-CRP and TNF-α levels when compared to healthy controls.

Children with chronic rheumatologic diseases are exposed to a vast array of pro-atherogenic insults, but the prevalence and natural history of accelerated atherosclerosis remains poorly defined. Cardiovascular disease is increasingly found in adult SLE patients which leads to significant increases in morbidity and mortality. Identifying key risk factors, developing disease-specific stratification algorithms, and implementing medical and nutritional interventions may help develop strategies to prevent or at least delay atherosclerosis in SLE patients. Multidisciplinary collaboration with pediatric rheumatologists, nutritionists, and preventive cardiologists may help optimize care in this vulnerable pediatric population. These unique results deserve further investigations to better elucidate the whole of these biomarkers in the context of nutrition and systems biology in SLE pediatric patients.