N05037: The effects of copper depletion and repletion on plasma homocysteine concentration, folate and choline status

Study Duration: March 2002 to April 2003

Contractor: NICHE, University of Ulster

Coronary Heart Disease is the main cause of death in Northern Ireland. It has recently been discovered that the risk of developing heart disease is increased by having raised levels of a substance called homocysteine in the blood. Homocysteine levels may, in turn, be influenced by the levels of copper in the body. The relationship between copper status and homocysteine metabolism is currently unclear, but there is evidence that copper may interact with the body's metabolism of homocysteine, B vitamins and blood lipids. For example, it is known that copper deficiency can impair certain enzymes involved in homocysteine metabolism (namely, s-adenosylhomocysteine hydrolase and methionine synthase) and can disrupt phospholipid metabolism by impairing choline phosphotransferase. Furthermore, the symptoms of clinical copper deficiency are similar to those of the congenital disease homocystinuria, which is characterised by extremely high homocysteine levels. Therefore, it may be that the risk of developing heart disease is influenced by the levels of copper in the body.

Rationale And Objectives

The CuFol project aim was to identify possible links between copper status, B vitamin metabolism, lipid metabolism and homocysteine, a risk factor for cardiovascular disease.

In a previous study (Opticue, NO5019), 38 apparently healthy men aged 18 - 45 years completed a placebo-controlled programme of 30 mg/day zinc supplementation for 14 weeks, with the intention of inducing marginal copper deficiency by impairing dietary copper absorption with zinc, a well-known copper antagonist. Following this was a period of copper supplementation (3 mg/day for 8 weeks) intended to correct any copper depletion. Samples taken at baseline, 2, 14, 16, 18 and 22 weeks were analysed to identify indicators of copper depletion and repletion.

The CuFol study used frozen plasma samples taken during the Opticue study. Whilst the Opticue study showed no evidence of copper depletion, there remained the possibility that none of the indices measured were sufficiently sensitive to reveal slight depletion. Therefore it was necessary to draw a definitive conclusion regarding the issue of copper depletion. To this end, the samples were first analysed for activity of the copper-containing enzyme semicarbazide-sensitive benzylamine-oxidase (BAO), which has recently been implicated as a possible sensitive marker of copper status. Analyses were then performed to examine the effect of copper status on total plasma homocysteine (tHcy), the B-vitamin folate, choline and the related nutrient, betaine.

Because levels of the vitamins B6 and B12 are themselves considered likely to affect levels of homocysteine, folate and/or choline, it was necessary to measure vitamin B6 and B12. This would enable their effect to be distinguished from any effects caused by copper status. It was also necessary to consider the subjects¿ genetic make-up; specifically, the gene which codes for the enzyme methylene tetrahydrofolate reductase (MTHFR). People with the TT genotype have a genetic tendency toward having high levels of homocysteine in their blood, which could make it harder to recognise any effects related to copper.

In order to distinguish effects of copper depletion from any effects caused by an improvement in zinc status, samples were analysed for plasma tHcy, folate, phospholipids, choline and betaine, after 2 weeks of zinc supplementation. If subjects began the intervention with sub-optimal zinc status, any effects of zinc repletion should be observed by week 2, whereas copper depletion would only become evident with longer-term zinc supplementation.

In the event that no effect of 22 weeks of copper and zinc supplementation was found, an alternative investigative strategy was to consider the information from all the subjects together, in order to investigate the underlying relationships between copper, homocysteine, choline-containing phospholipids, folate, choline and betaine.

All subjects were found to have adequate status of vitamins B6 and B12 (over 18.6 nanomol per litres and 150 nanograms per litre respectively). One individual, identified as being folate deficient (plasma folate less than 2.7 nanograms per litre), and five subjects identified as homozygous for the 677C~>T MTHFR polymorphism were excluded from analyses of the effects of copper supplementation followed by zinc supplementation on plasma tHcy, choline-containing phospholipids, choline and betaine. To exclude the effects of potential zinc repletion, two-way ANOVA was used to compare subjects' baseline data with values after two weeks of zinc supplementation. No effects of zinc repletion were indicated.

Repeated-measures ANOVA was used to examine the effects of 22 weeks of copper and zinc supplementation on plasma BAO activity, choline-containing phospholipids, tHcy, free choline or betaine. No significant effects of treatment were found. However, in the population as a whole, significant differences were found between the different time-points, both in BAO activity and in concentrations of choline-containing phospholipids. Mean BAO activity was significantly higher at week 16 than at week 2, 18 or 22 (P = 0.001, 0.000 and 0.000 respectively). Choline-containing phospholipids were significantly higher at week 18 than at week 2 (P = 0.001). Concentrations of tHcy were significantly higher in the placebo group than in the supplemented group at baseline, week 2 and week 16 (P = 0.030, 0.013 and 0.009 respectively).

Because no effects of copper supplementation or zinc supplementation were revealed, either in the preceding analyses or in any of the analyses performed on these samples in the previous Opticue study, it was concluded that there had been no effect of treatment. Data from all 38 subjects, therefore, were used together in an investigation into the correlations between copper and zinc status, and levels of homocysteine, folate, phospholipids, free choline and betaine. Pearson correlations indicated significant inverse relationships between plasma tHcy and indicators of folate, vitamin B12, vitamin B6 and betaine status. Concentrations of tHcy did not correlate significantly with BAO or with the other putative markers of copper/zinc status previously measured in the Opticue study, namely extracellular superoxide dismutase, serum copper, serum zinc and alkaline phosphatase.

Multiple linear regression analysis (MLRs) of the observed variation in plasma tHcy levels indicated that significant variation in measured tHcy was contributed by plasma folate, betaine, vitamin B6 and choline-containing phospholipids. MLRs failed to identify any of the putative copper/zinc status markers as significant determinants of plasma tHcy.

In TT individuals, plasma concentrations of tHcy, folate, phospholipids, choline and betaine were generally different from the other subjects. Mean tHcy for the non-TT population was significantly lower than the TT mean (P = 0.003). Conversely, the non-TT population mean was higher than the TT mean for folate (P = 0.000), vitamin B12 (P = 0.003), vitamin B6 (P = 0.000), choline and betaine (logged data, P = 0.008 and 0.039 respectively).

Observed differences between BAO activity at different time-points were unrelated to treatment. Little information is available, in the published literature, concerning normal fluctuations in plasma BAO activity of healthy adults. One preliminary study, comparing the BAO activity detectable in two plasma samples taken six weeks apart, reported no significant difference (Boomsma et al., 2003). However, as only 12 subjects were involved, and on only two occasions, it would be difficult to use the Boomsma study to make inferences regarding the longer-term variability of plasma BAO activity. The current 22-week investigation, covering a period from November to April, is the first to document BAO activity over such a time-span. The fluctuations observed during this period may warrant further investigation.

Although zinc supplementation followed by copper supplementation was demonstrated to have no significant effect on tHcy, this may in part be related to the low-normal pre-trial plasma tHcy concentrations in the treatment group. Generally, the higher the tHcy, the greater the potential for lowering by intervention; but these subjects were not recruited on the basis of their tHcy. The finding of significantly higher tHcy levels in the placebo group than in the supplemented group (average values over the study period were 10.12 and 8.24 respectively) was probably not a result of the supplementation treatment. As the difference was evident at baseline and remained present throughout the intervention, it was likely an artefact of the subject randomisation prior to intervention. It should be noted that this randomisation was undertaken to test the original hypothesis, unconnected with tHcy and B vitamin metabolism. Therefore, no stratification with respect to tHcy values was performed.

The inverse relationships between plasma tHcy concentrations and status of folate, vitamin B12 and vitamin B6 are consistent with previous findings, which have identified these B vitamins as determinants of plasma tHcy. The finding of betaine as a significant determinant of variation in tHcy is also in keeping with current understanding of homocysteine metabolism: oral betaine treatment is used clinically to decrease plasma tHcy in patients with homocystinuria. The lack of a correlation between choline and tHcy was perhaps owing to the low-normal levels of plasma tHcy in the current study population (median 8.60, IQR 7.51 - 10.48). Abnormally elevated plasma homocysteine has previously been linked with choline deficiency (Varela-Moreiras et al., 1995), and has been ameliorated by choline supplementation (Dudman et al., 1993), but when tHcy is within the lower range of normal, it is likely to be less responsive to dietary choline intake.

In this preliminary investigation of plasma concentrations of choline and betaine with respect to MTHFR genotype, the frequency of homozygosity for the 677CT MTHFR polymorphism in Northern Ireland, between 10 and 15%, has, as expected, yielded too few homozygotes (n = 5) to support a meaningful statistical analysis. Nonetheless, the observed differences between TT and non-TT individuals provide clear evidence of differing homocysteine metabolism in these two groups. There are, to our knowledge, no existing published data on plasma choline concentrations in TT individuals, but the observed low folate status in TTs is in keeping with existing literature on this group (de Bree A, 2003); and mouse models of this genotype are reported to have lower than normal liver levels of betaine (Schwahn et al., 2003). Power calculations based on the measured difference in plasma betaine indicate that, to attain statistical significance, a difference of this magnitude would have to be observed in 80 TT individuals. These preliminary findings suggest that further investigation of this novel field may prove fruitful.

Contact: Ms Mamta Singh
Tel: 020 7276 8919
Email: mamta.singh@foodstandards.gsi.gov.uk