Folic Acid Does Not Improve Endothelial Function in Obese Children and Adolescents

A total of 56 children and adolescents with mild-to-moderate obesity (BMI Z score 1.7–3.0) were recruited consecutively from the outpatient clinics at Women's and Children's Hospital (Adelaide, Australia) between January 2004 and January 2006. Exclusion criteria were smoking, diabetes, hypertension, lipid-lowering treatment, vitamin B12 deficiency, and recent use of folate supplements, as well as syndromal obesity and/or endocrinological causes of obesity. These 56 subjects were part of our cross-sectional study (5). Sample size was calculated with an estimated increment of 3 ± 3.3% on FMD based on our previous studies (17,18), with 80% power at a 5% significance level. The Women's and Children's Hospital human research ethics committee approved the study. Written informed consent was obtained from the parents/guardians and the subject if he/she was aged >16 years.

A randomized, double-blind, placebo-controlled, parallel trial was performed comparing oral folic acid (5 mg daily) with placebo (both supplied by Sigma Pharmaceuticals, Melbourne, Australia). Two assessments were done before intervention during visits at −8 and 0 weeks to review endothelial function and weight changes over time without intervention, and two assessments were done after the intervention during visits at 8 and 16 weeks to measure endothelial function at peak red cell folate (RCF) levels (8-week visit) and 8 weeks after discontinuing folate (16-week visit), as we have shown that 8 weeks are required to show loss of beneficial effects of folate on FMD (17) (Fig. 1). Randomization was performed using a Fisher table in blocks of 10 (Pharmacy Department, Women's and Children's Hospital, Adelaide, Australia). Subjects were randomized at 0 weeks to receive either folic acid or placebo for 8 weeks. The folic acid dose was chosen taking previous studies into account (17,18). Subjects were instructed not to take any other vitamins during the study period, and their diet was not modified. Compliance was assessed by pill counting.

Four assessments were done at −8, 0, 8, and 16 weeks (Fig. 1). Baseline clinical data were collected at −8 weeks (Table 1). Obesity duration was determined by a subjective answer by the subject and/or parents. At each assessment, vascular function (FMD and GTN), hsCRP, serum folate, RCF, tHcy, lipids, height, weight, blood pressure, and waist and hip circumferences were measured. Height was measured with a wall-mounted stadiometer to the nearest 0.1 cm. Weight, measured with subjects wearing minimal clothing, was taken on an electronic digital scale to the nearest 0.1 kg. BMI (calculated as weight in kilograms divided by the square of height in meters), BMI Z score, weight Z score, and height Z score were calculated using EpiInfo database, version 3.2.2, and Centers for Disease Control 2000 standardized reference charts. Waist circumference was measured at the midpoint between the lower edge of the ribs in the midaxillary line and the top of the iliac crest. Blood pressure was taken with a cuff of the appropriate size on the left arm after a 10-min rest in supine position. All subjects were well at the time of assessment.

Endothelial and smooth muscle function (FMD and GTN) were assessed after overnight fasting, as previously reported (5,17–19). The diameter of the brachial artery (2–15 cm above the elbow) was measured in longitudinal sections from two-dimensional ultrasound images with a 10.0 MHz linear array transducer (Advanced Technology Laboratories, Bothel, WA) using an HDI 3000 ultrasound system (Advanced Technology Laboratories). Experienced pediatric sonographers performed all studies. A suitable site for imaging the vessel was selected, with reproducible ultrasonic markers such as venous valves or vessel bifurcations, to ensure that the measurement occurred at the same place for each scan. An electrocardiogram was recorded with the ultrasound images. Each study included four scans. The first was taken at rest. Reactive hyperemia was then induced by occluding arterial blood flow for 4 min using a sphygmomanometer inflated to 250 mmHg. Arterial flow velocity was measured by a pulsed Doppler signal at 60° to the vessel during the resting scan and for the first 15 s after deflation of the cuff. The second scan (FMD) was recorded 30–90 s after cuff deflation, with measurements between 45 and 75 s after deflation. The third (recontrol) scan was taken after 10–15 min was allowed for vessel recovery. The last scan was taken 4 min after sublingual glyceryl trinitrate spray (400 μg Nitrolingual Spray; G. Pohl-Boskamp, Hohenlockstedt, Germany) administration.

Images were recorded onto VHS videotape and analyzed subsequently by a blinded observer. For each scan, measurements were made incident with the electrocardiogram R-wave (i.e., at end diastole) over four cardiac cycles using ultrasonic calipers. Then, the measurements were averaged and expressed as percentages of the first control (resting) scan. There were four final measurements in total: resting, FMD, recontrol, and GTN. Reactive hyperemia was calculated as the flow in the first 15 s after cuff deflation divided by the flow during the resting scan. The coefficient of variation between 20 subjects is 3.9% for FMD and 4.0% for GTN (19).

Overnight fasting blood samples were collected. hsCRP was measured using a near-infrared particle immunoassay method using IMMAGE Immunochemistry Systems regent (Beckman Coulter, Fullerton, CA). Triglycerides, total cholesterol, and HDL were measured using enzyme-based assays on the Beckman Synchron CX5 analyzer (Beckman Coulter). LDL cholesterol was calculated using the Friedewald equation. A1C was measured using a latex immunoagglutination inhibition methodology (DCA 2000 A1C Reagent; Kit-Bayer, Toronto, ON, Canada).

Serum folate and RCF were measured using chemiluminiscent microparticle folate-binding protein assay (Architect System; Abbott Laboratories). tHcy was measured by a fluorescence polarization immunoassay using the commercial IMx homocysteine assay (Abbott Diagnostic, Oslo, Norway).

Subjects had an oral glucose tolerance test after an overnight fast using 1.75 g/kg of anhydrous dextrose to a maximum of 75 g in a volume of 300 ml. Glucose was measured by the hexokinase spectophotometry method (Synchron cx5ce; Beckman Coultur). Insulin was measured by a two-sandwich immunoassay using cheminluminescent technology (ADVIA Centaur Insulin Assay; Bayer Health Care, Tarrytown, NY.) Homeostasis model assessment of insulin resistance was calculated as follows: insulin (mU/l) × glucose (mmol/l)/22.5.

The data were analyzed using SPSS software, version 14.0. Statistical significance was inferred with a P value <0.05. Natural logarithmic transformation was applied to abnormally distributed variables (hsCRP and triglycerides). An independent sample t test was used to compare baseline characteristics between treatment groups and assessed the adequacy of randomization. The analysis was completed by intention to treat. Linear mixed models with simple contrasts were used to assess the effect of the intervention on FMD, GTN, vessel diameter, hsCRP, serum folate, RCF, tHcy, height, weight, weight Z score, BMI, BMI Z score, waist-to-hip ratio, and blood pressure. Simple contrasts were used to compare the expected mean on the variables above between-treatment groups at each visit (−8, 0, 8, and 16 weeks). Pearson's correlations were used to access the linear association of change over the intervention period (0–8 weeks) between FMD, GTN, folate, RCF, and tHcy.

A total of 194 subjects were eligible for the trial, of whom 138 did not participate. The main reasons for refusal were lack of interest (n = 88), living in the countryside (n = 18), time constraints (n = 16), needle phobia (n = 10), and difficulties taking tablets (n = 4). Of the 56 subjects screened, 2 had the first assessment and did not wish to continue with the trial, and 1 was excluded because of a vitamin B12 deficiency. A total of 53 subjects (26 male) aged 13.3 ± 2.2 years participated in the study. One subject in the folic acid group withdrew from the trial before the 8-week assessment. There were no significant differences in the baseline characteristics of the folic acid and placebo groups (Table 1). Baseline tHcy, folate, RCF, and vitamin B12 were within normal range. Compliance according to pill counting for placebo and folic acid tablet ingestion was 81%. No adverse side effects were reported.

There was no significant difference in the change in FMD from −8 to 16 weeks between the folic acid and placebo groups (P = 0.7). There was no significant difference in FMD between treatment groups at −8 weeks (P = 0.8), 0 weeks (P = 0.3), 8 weeks (P = 0.6), and 16 weeks (P = 0.4) (Table 2). Smooth muscle function (GTN), vessel diameter, hsCRP, lipids, BMI Z score, and weight Z score did not significantly change with the intervention (Table 2).

Folic acid supplementation increased serum folate by 18.4 nmol/l (95% CI 13.8–23.0; P < 0.001) and RCF by 240.1 nmol/l (201–364; P < 0.001) during the intervention (0–8 weeks). This increment represents an increase of 100% for serum folate and 67% for RCF. Folic acid decreased tHcy by 0.95 μmol/l (−1.45 to −0.45; P = 0.008), which represents a 17% reduction in tHcy. Changes in serum folate and RCF related to changes in tHcy levels (r = −0.40, P = 0.004 and r = −0.36, P = 0.016, respectively). Changes in serum folate related to changes in RCF (r = 0.39, P = 0.01).

A subgroup analysis of the obese children with three or more metabolic syndrome criteria (triglycerides ≥1.24mmol/l, HDL cholesterol ≤1.03 mmol/l, waist circumference higher than or equal to the 90th percentile according to age and sex, fasting glucose level ≥6.11 mmol/l, and blood pressure higher than or equal to the 90th percentile according to age and sex) (20) was performed. Of 53 children, there were 12 with metabolic syndrome (folic acid group, n = 7; placebo group, n = 5), and there was no significant change in FMD (P = 0.46) or GTN (P = 0.22) with the intervention in these subgroups.

BMI and weight Z scores improved to a small degree but significantly during the trial (−8 to 16 weeks) in both groups (from 2.30 to 2.27, P = 0.03 and from 2.51 to 2.48, P = 0.04, respectively), and the change did not significantly differ between groups (P = 0.2 and P = 0.3, respectively). Waist-to-hip ratio improved from −8 to 16 weeks in both groups from 0.915 to 0.905 (P = 0.002). Total cholesterol, HDL cholesterol, LDL cholesterol, triglycerides, and hsCRP did not change during the trial in either group.

We have shown, for the first time, that short-term folic acid supplementation does not improve endothelial function in children and adolescents with mild-to-moderate obesity. This was despite a sustained increase in serum folate and RCF, above the normal range, and a significant decrease in tHcy.

We had hypothesized that obese children would respond to folate, as they have higher tHcy levels and lower folate status than children with type 1 diabetes, in whom folate normalizes endothelial function. Possible explanations for the lack of effect of folic acid on endothelial function in obese children compared with children with type 1 diabetes are different baseline characteristics and different responses to folic acid. We have shown that obese children have higher levels of inflammatory markers, higher LDL-to-HDL ratios, higher triglycerides, higher tHcy levels, and lower folate status compared with children with type 1 diabetes (5). Therefore, despite their metabolic derangement, children with type 1 diabetes have a more favorable cardiovascular risk factor profile and less inflammation than obese children of the same age (5). We and others have shown that inflammation is more pronounced in obese children than in children with type 1 diabetes with a similar degree of endothelial dysfunction and increased intima-media thickness (5,21). Inflammation and lipid abnormalities may be more important components of the underlying mechanisms for endothelial dysfunction in obesity than in type 1 diabetes (4,21) and be less amenable to folate supplementation.

Folic acid did not improve endothelial function in obese children compared with children with type 1 diabetes despite increasing serum folate and RCF at a higher level (100 vs. 50% and 67 vs. 56%, respectively) and decreasing tHcy levels (12% vs. no change). Of interest, this increment in serum folate and RCF in our obese children was greater than that in children with type 1 diabetes after supplementation with the same folic acid dose and with similar compliance (17,18). This is probably explained by the lower basal folate levels in obesity, which has been described before and has been related to diet (5,22). Folic acid reduced tHcy in our obese children, who have higher tHcy levels, but not in children with type 1 diabetes (17,18) This reduction in tHcy was comparable with that obtained in children with chronic renal failure with high tHcy levels, in whom endothelial function did not improve (8). This might demonstrate differences in the folic acid mechanism of action in obesity or chronic renal failure in comparison with type 1 diabetes; i.e., folic acid reduced tHcy levels but did not cause direct endothelial effects.

Obese children are not the only population studied in whom folate supplementation has not improved FMD; folic acid does not improve FMD in adults with chronic renal failure and hyperhomocysteinemia (23) despite increments in folate status and a decrease in tHcy levels, which were comparable with our obese children. Folic acid improves FMD in children with chronic renal failure and raised tHcy levels, but this improvement was not significant in comparison with placebo (8).

The majority of studies of the effect of B vitamins on endothelial dysfunction have investigated their effect through the lowering of homocysteine levels (6,7). In our study, endothelial function in obese children did not improve despite a significant decrease in tHcy. However, direct benefits of folate within 2–4 h on endothelial function, independent of homocysteine lowering, are reported in vitro (13–16) and in type 1 and type 2 diabetes in our studies and others’ (10,17,18,24,25). Consistent with these findings is the evidence that low RCF independent of homocysteine in adults is associated with increased intima-media thickness (26). Potential mechanisms of folate's direct action on the endothelium are on increments in nitric oxide levels (13), direct interactions with endothelial nitric oxide synthase enzyme (14), effects on cofactor availability (15), and antioxidant actions (16). These effects of folic acid (13–16) were not achieved in an obese population with the folic acid dose used in the trial. A total of 5 mg folic acid might be sufficient to increase antioxidant activity and therefore improve endothelial function when baseline antioxidant status is reduced, as in diabetes, but it has less effect in obesity. Although we did not measure antioxidant activity, Bennett-Richards et al. (8) have shown that 5 mg folic acid does not improve endothelial function or antioxidant activity in children with chronic renal failure and hyperhomocysteinemia.

Folic acid (2.5 mg/day) with a hypocaloric diet improves inflammatory markers such as C-reactive protein, interleukin-8, and macrophage chemo attractant protein-1, as well as insulin sensitivity in overweight (BMI <95 percentile or BMI 25–29 kg/m²) but not obese (BMI >95 percentile or BMI ≥30 kg/m²) adults with normal glucose tolerance (11). Folic acid did not improve hsCRP in our study's obese subjects despite a higher folic acid dose (5 mg/day), probably due to their higher BMI (all with BMI >95 percentile for age and sex) and the lack of dietary intervention in addition to the folate supplementation. To our knowledge, this is the first study evaluating the effects of folate supplementation on endothelial function and/or inflammatory markers in obese children.

There was a statistically significant but clinically small improvement in BMI Z score, weight Z score, and waist-to-hip ratio over the study period in all children, irrespective of the intervention. These changes were not significant in comparison with improvements with diet and exercise interventions (27,28).

One limitation to our study was the participation rate, which is, however, comparable with previous trials in obese children (27,28) and did allow good retention of children during the 24-week study period.

In conclusion, short-term treatment with 5 mg folic acid daily did not improve endothelial function in children and adolescents with mild-to-moderate obesity despite improved folate status and reduced tHcy levels. This is in contrast to the response to folate in children with type 1 diabetes, highlighting the complexity of vascular dysfunction in obesity and emphasizing the difficulties in finding other strategies besides diet and exercise in the treatment of endothelial dysfunction in obesity.

This work was supported by Channel Seven Research Foundation and Women's and Children's Research Foundation grants. A.P. was supported by a University of Adelaide scholarship and an international postgraduate research scholarship.

Preliminary results of this manuscript were presented in abstract form at the 10th International Congress on Obesity (2006).