ORIGINAL COMMUNICATION

Transcription

1 (2003) 57, & 2003 Nature Publishing Group All rights reserved /03 $ ORIGINAL COMMUNICATION Reduction of plasma homocysteine and serum methylmalonate concentrations in apparently healthy elderly subjects after treatment with folic acid, vitamin B 12 and vitamin B 6 : a randomised trial C Lewerin 1, H Nilsson-Ehle 1 *, M Matousek 2, G Lindstedt 3 and B Steen 2 1 Department of Haematology and Coagulation, Göteborg University, Göteborg, Sweden; 2 Department of Geriatric Medicine, Göteborg University, Göteborg, Sweden; and 3 Department of Clinical Chemistry and Transfusion Medicine, Göteborg University, Göteborg, Sweden Objectives: To investigate, in an elderly population: (1) the effects of oral B-vitamin therapy on P-tHcys, S-MMA and Hb/MCV, (2) the appropriate decision limit for high metabolite concentrations and (3) the estimated prevalence of vitamin B 12 /folate deficiency on the basis of different decision limits. Design: Double-blind placebo-controlled intervention study. Setting: Outpatient clinic. Subjects: A total of 209 community-dwelling subjects, median age 76 y (range 70 93) y. Intervention: Four months of oral daily supplementation with 0.5 mg cyanocobalamin, 0.8 mg folic acid and 3 mg vitamin B 6. Results: High P- thcys was found in 64% of men and 45% of women, high S-MMA in 11% of both. Vitamin B 12 deficiency was observed in 7.2% and folate deficiency in 11% of all subjects. Health-related upper reference limits for the metabolites at the start were higher than the laboratory s upper reference limits. The latter were, however, similar to those of the vitamin replete group. There was a significant decrease in P-tHcys (Po0.001) and S-MMA (P ¼ 0.009) after 4 months of vitamin treatment. In a multivariate analysis, the P-Hcys change correlated positively with baseline P-tHcys and inversely with baseline P-folate and transferrin saturation (Fe/TIBC ratio). The S-MMA change correlated with baseline S-MMA and inversely with baseline vitamin B 12 and age. Conclusions: Suboptimal vitamin status is an important cause of elevated P-tHcys and S-MMA in apparently healthy elderly subjects. Oral B-vitamin therapy is an effective and convenient way to normalise P-tHcys and S-MMA. Sponsorship: SupportFRecip AB. (2003) 57, doi: /sj.ejcn Keywords: methylmalonic acid; homocysteine; vitamin B 12 ; folic acid; aged; reference values *Correspondence: H Nilsson-Ehle, Department of Medicine, Haematology and Coagulation, Sahlgrenska University Hospital, Göteborg University, Göteborg, SE , Sweden. herman.nilsson-ehle@medic.gu.se Contributors: B Steen, H Nilsson-Ehle and M Matousek designed the study. G Lindstedt contributed expertise and supervision on the laboratory assays. C Lewerin, with H Nilsson-Ehle as her tutor, gathered data and was responsible for the statistical calculations and for the preliminary preparation of the manuscript. All authors contributed to revision of the final draft of the manuscript. Recip AB manufactured both the vitamin and placebo tablets. Received 6 June 2002; revised 25 November 2002; accepted 6 December 2002 Introduction Pernicious anaemia occurs in 0.3% of the total Swedish population and 2% over the age of 70 (Nilsson-Ehle et al, 1989). Neurological signs may be more common in this group (Lindenbaum et al, 1988; Healton et al, 1991). There is an age-related decline in serum cobalamin (S-B 12 ) concentrations (Nilsson-Ehle et al, 1991), and cobalamin (vitamin B 12 ) deficiency is more common in the elderly (Nilsson-Ehle, 1998). This might be due to an increased prevalence of atrophic gastritis (Streeter et al, 1976; Ihamäki et al, 1979;

2 Krasinski et al, 1986; Carmel et al, 1987; Jones et al, 1987) and certain medicaments (Adams et al, 1983; Belaiche et al, 1983; Kittang & Schjönsby, 1987; Lindstedt, 1999). The extent of vitamin B 12 malnutrition in the elderly is unknown but concern relates to the observation that microwave heating of food can cause degradation of hydroxo-cobalamin (Watanabe et al, 1998). In the elderly, megaloblastic anaemia due to folate deficiency is rare (Nilsson-Ehle, 1998). Poor folate status correlates to cognitive dysfunction (Fioravanti et al, 1997; Lehmann et al, 1999) and mental depression (Bottiglieri, 1996; Alpert & Fava, 1997; Fava et al, 1997). Particularly susceptible to folate deficiency are homozygotes (5 15% of the population) for a mutation in a 5-MTHFR gene (EC , 677C-T) (Jacques et al, 1996; Molloy et al, 1997; Bailey & Gregory, 1999). Folate deficiency is mainly caused by inadequate intake or excessive cooking of vegetables, certain medicaments (Zimmerman et al, 1987; Morgan et al, 1998; Apeland et al, 2001) and intestinal malabsorption.vitamin B 6 deficiency is not uncommon in the elderly (Haller et al, 1991), mostly due to reduced intake (Bates et al, 1999). Vitamin B 12, folate and B 6 are all involved in the methionine metabolism as cofactors or substrates for key enzymes. This might explain the considerable overlap in signs and symptoms of deficiency. Elevated concentrations of plasma total homocysteine (P-tHcys) is seen in B vitamin deficiency (B 12, folic acid, B 6 ), impaired renal function (Norlund et al, 1998; Bostom et al, 1999; Brattström & Wilcken, 2000) and hypothyroidism (Hussein et al, 1999). Certain lifestyle factors may also contribute (Nygård et al, 1998; Stolzenberg-Solomon et al, 1999; de Bree et al, 2001). Elevated concentrations of serum methylmalonate (S-MMA) is seen in vitamin B 12 deficiency, impaired renal function (Hvas et al, 2000) and dehydration (Stabler et al, 1988), low values after antibiotic therapy (Lindenbaum et al, 1990). Functional vitamin deficiency, defined as elevations of P- thcys or S-MMA combined with low plasma vitamin concentrations, is prevalent in the elderly. Vitamin B 12 deficiency occurs in 10 20% (Nilsson-Ehle, 1998), folate deficiency in 3 19% (Webster & Leeming, 1979; Joosten et al, 1993; Matthews, 1995). Vitamin B 6 might be more important for normal P-tHcys than previously appreciated (McKinley et al, 2001), but sometimes causes elevated P- thcys only after methionine loading (Ubbink et al, 1996). The reference intervals for P-tHcys are higher in elderly individuals and in men when compared with women (Rasmussen et al, 1996; Nilsson-Ehle, 1998; Selhub et al, 1999) and those for S-MMA are higher in older women than men (Rasmussen et al, 1996). Although most factors associated with increased P- thcys accumulate at high age (Herrmann et al, 1999; Schumann, 1999), it is not known whether P- thcys increases as a result of normal ageing. Whether elevation of S-MMA can be taken as definite proof of vitamin B 12 deficiency is still not settled (Chanarin & Metz, 1998; Hvas et al, 2001). The concept of health is complex, not least in the elderly. We have previously, using conventional methods, determined health-related reference intervals for, for example, S- B 12 in elderly populations (Solberg, 1983; Nilsson-Ehle et al, 1991). The correlation between functional and clinical vitamin deficiency has, however, yet to be determined in long-term randomised intervention trials with sufficient amounts of vitamin(s). The aims of this study were to investigate, in a population of community-dwelling elderly subjects, (1) the effects on haematological variables, P- thcys and S- MMA of 4 months oral substitution with vitamin B 12, folic acid and vitamin B 6, (2) to calculate health-related reference intervals for P- thcys and S- MMA and (3) to calculate the estimated prevalence of vitamin B 12 /folate deficiency on the basis of different decision limits. Methods Study design Participants in a previous study (Augustsson et al, 1994), living in a local urban area in Göteborg, were invited by letter to participate in this double-blind placebo-controlled study of 4 months duration and the first 209 subjects who accepted were enrolled and assigned to vitamin or placebo therapy according to a randomised parallel group design. Those who had taken any vitamin supplements during the last 3 months or pharmacological doses of vitamin B 12, folic acid and/or vitamin B 6 during the last 3 years were not allowed to enter the study. No further exclusion criteria for entering the study were applied but a number of subjects declined to participate for other reasons. Height, weight and blood pressure were measured and all probands were interviewed regarding allergy, drug consumption and smoking habits. During follow-up, any new symptoms possibly related to the treatment given were recorded. Using previous pilot data, the number of participants needed to reach a Po0.05 significance level for the difference before and after the planned vitamin therapy was found to be 180 (vitamin group n ¼ 120, placebo group n ¼ 60). Total study group The total study group comprised 209 men and women with a mean age of 76 y and 5 months; range y (women) and y (men). Of the total study group, 139 were randomised to active therapy (the vitamin group) and 70 to the placebo group (Table 1, Figure 1). Treatment in the vitamin group consisted of a daily tablet containing 3 mg pyridoxine hydrochloride, 800 mg folic acid and 500 mg cyanocobalamin. The tablet was identical in shape and composition to the placebo tablet apart from the vitamin content. In both groups, one tablet was taken daily in the morning for 4 months. To ensure compliance, all subjects received a specified blinded number of tablets, and at the end of the study, the number of remaining tablets was compared with 1427

3 1428 Table 1 Observations in the vitamin and placebo groups (mean values or numbers) Unit Vitamin group, n=139 Placebo, n=70 P Age y Body mass index kg/m Number of smokers 20/139 6/ Number of drop-outs 24/139 6/ Blood measurements Haemoglobin g/l Ery-MCV fl Leukocytes x10 9 /l Platelets x10 9 /l Folates nmol/l Plasma measurements Folates nmol/l thcys mmol/l Serum measurements B 12 pmol/l MMA mmol/l Creatinine mmol/l Iron mmol/l Iron-binding capacity mmol/l Ery-MCV=erythrocyte mean corpuscular volume. No sign response to vitamins: RS II = 66 for thcys, 115 for MMA Total study group n = 209 (of whom healthy n = 125 = RS I) Randomisation Before treatment Vitamin group n = 139 Placebo group n = 70 Treatment period Drop-outs n = 24 Drop-outs n = 6 After treatment Vitamin treatment complete n = 115 (of whom healthy n = 78 = RS III) Placebo treatment complete n = 64 Figure 1 Study design and reference sample groups. the initial number and planned intake during the study period. During the study, a total of 30 subjects (24 in the vitamin and six in the placebo group) were excluded for one or more of the following reasons: intervening illness (n ¼ 13, two deaths), poor compliance (subjects retaining an excess of 40 tablets more than expected, n ¼ 16), protocol violation (n ¼ 14), refusal (n ¼ 3), other (n ¼ 4) and were therefore not included in the calculations of any effects of the interventions.

4 Reference sample groups For the calculation of health-related reference intervals, three subgroups of the total study group (reference sample groups (RS) I III; Figure 1) were defined. RS I comprised subjects not meeting criteria indicating disease (Table 2) and was used for calculations of traditional baseline healthrelated reference intervals. RS II comprised subjects in the vitamin group not achieving a significant decline in P-tHcys or S-MMA (arbitrarily defined as 43s.d. of the change in the placebo group), thus presumably not vitamin deficient at baseline. RS III comprised subjects in RS I who received active vitamin therapy, thus forming a healthy and vitamin-replete subgroup, analysed at the end of the study. Ethical considerations Informed consent was obtained from all probands, and the Research Ethics Committee of the Medical Faculty of Göteborg University approved the protocol. Table 2 Exclusion criteria for the calculation of reference intervals for different haematological components Anamnestic exclusion criteria Laboratory exclusion criteria Previously recognised disease ESR >35 mm/h women Cancer >30 mm/h men Myocardial infarction Blood haemoglobin o120 g/l women Stroke o130 g/l men Pneumonia Leukocytes o /l Heart failure > /l Platelets o /l Ongoing medical treatment > /l Corticosteroids Serum creatinine >150 mmol/l Serum B 12 o130 pmol/l For whole blood folates also Whole blood folates o90 nmol/l women Anticonvulsant drugs o100 nmol/l men Ery-MCV o82 fl >102 fl For S-MMA and P-tHcys also Transferrin o16% saturation Histamine H 2 -receptor antagonists Proton pump inhibitors Oral biguanides Modified-release potassium preparations Systemic antibiotics For serum B 12 also Vitamin C For blood Hb Corticosteroids Histamine H 2 -receptor antagonists Proton pump inhibitors Oral biguanides Modified-release potassium preparations Systemic antibiotics Iron salts Ery-MCV=erythrocyte mean corpuscular volume. Blood sampling and laboratory methods Blood samples were collected at the start of the study and after 1 and 4 months. Samples were obtained with the subjects in a recumbent position, after an overnight fast. Vacuum tubes were used. Blood for determination of cell counts was collected in EDTA tubes, for plasma analyses in heparinised tubes and for serum analyses in tubes without anticoagulant. After venipuncture, serum and plasma samples were centrifuged and thereafter kept at room temperature for 2 h before analysis. Blood haemoglobin and cell counts were analysed using a Technicon H2 flow cytometer, serum iron (Fe) and serum total iron-binding capacity (TIBC) using a Hitachi 917 analyser with ferrozine ascorbic acid as chromogen. The concentrations of serum B 12 and of whole blood and plasma folates were determined by radioassays (Solid Phase No Boil Dualcount, Diagnostic Products Corp., Los Angeles, CA, USA). S-MMA was measured using capillarygas chromatography and mass spectrometry (Rasmussen, 1989). P-tHcys was measured by high-performance liquid chromatography with fluorescence detection (Bald & Sypniewski, 1994). The current health-related upper reference limits of the laboratory were for plasma thcys 16 mmol/l and for serum MMA 0.34 mmol/l. Reference intervals for plasma folates were 6 35 nmol/l, for whole blood folates nmol/l and for serum B pmol/l. Statistical methods The primary efficacy variables were changes in S-MMA and P-tHcys concentrations after vitamin treatment. Tests of these changes were done with pairwise tests, differences in mean change between groups were assessed with the permutation t-test. Tests of difference in proportions between groups were done with Fisher s exact test. Stepwise multivariate linear regression was used to find baseline variables predicting changes in P- thcys (DP-tHcys) and S- MMA (DS-MMA) in the vitamin group. Two-tailed tests were used and P-values o0.05 were considered significant. Reference intervals, comprising the central 0.95 interfractile intervals, were calculated according to IFCC (International Federation for Clinical Chemistry) recommendations using a nonparametric method (Solberg, 1983) after analysis and, if needed, transformation of the actual distributions. In subgroups containing o100 subjects, the upper reference limits were calculated as mean þ 1.97s.d. Differences between groups were analysed with the permutation t-test. The software used was part of a statistical program system developed at the Department of Geriatrics, Institute of Community Medicine, Göteborg University. Results There were no cases of macrocytic anaemia at the start of the study. No adverse effects related to the vitamin treatment given were noted during the study period. Two subjects in the placebo group discontinued the treatment due to 1429

5 1430 gastrointestinal symptoms; in the remaining cases of discontinuation there were other than symptomatic reasons. Reference intervals for the biochemical components Mean values and distributions for serum vitamin B 12, plasma and blood folates, P-tHcys and S-MMA are given in Table 3. At the start of the study, the vitamin concentrations were slightly higher in the healthy subgroup (RS I) than in the total study group. They were, as expected, substantially higher in the healthy and treated subgroup (RS III). The upper reference limits of P-tHcys and S-MMA concentrations at the start of the study were higher than those of the laboratory s reference interval, slightly lower in the healthy subgroup than in the total study group. In RS III, concentrations of both metabolites were lower than in untreated subjects, and the upper reference limits were close to those stated by the laboratory. In all groups, P-tHcys was significantly higher in men than in women, whereas no significant sex differences were noted for S-MMA. Elevated baseline P-tHcys and S-MMA concentrations in the total study group Using the 97.5 centile values of the healthy, untreated subgroups as decision limits, the percentage of high P-Hcys ranged from 7 to 18%. The corresponding figures using the healthy vitamin replete-subgroup (RS III) and the laboratory s upper reference limit were 40 64%. By these criteria, high P-Hcys was significantly more common among men than women. High S-MMA values occurred in 1 to11% and no sex differences were observed. Low vitamin and high metabolite concentrations at the start of the study The fractions of subjects with high metabolite concentrations (P-tHcys 416 or S-MMA mmol/l) using different decision limits for concentrations of S-B 12, P- and B-folates are shown in Table 4. Some of these decision limits given in addition to the 2.5 percentile values for RS I and the stated lower reference limits at the laboratory, were chosen from selected reports in the literature (Yao et al, 1992; Joosten et al, 1993; Lindenbaum et al, 1994; Naurath et al, 1995; Brouwer Table 3 Reference intervals at baseline for the total study group (TSG), for apparently healthy subjects (RS I) and for subjects not showing a significant decline in P-tHcys or S-MMA during vitamin therapy (RS II). Values for apparently healthy subjects after completion of vitamin treatment (RS III) n M s.d. Measurement TSG Serum B Baseline Plasma folates Whole blood folates P- thcys m f S-MMA RS I Serum B Baseline Plasma folates Whole blood folates P-tHcys m f S-MMA RS II P-tHcys m Baseline f S-MMA RS III Serum B After vitamin Plasma folates therapy Whole blood folates P-tHcys m f S-MMA Values were calculated as described in the text. In subgroups containing o100 subjects, the central 0.95 fractile intervals calculated as mean71.97s.d. M=mean; m=men; f=women.

6 Table 4 Frequency (%) of elevated concentrations of S-MMA and P-tHcys in the total study group at baseline in relation to vitamin concentrations in all 209 subjects 1431 Fraction (%) of subjects with high metabolite concentration S-MMA P-tHcys Vitamin concentrations Source for decision limit Subjects number % >0.34 >16 S-B 12 o103 A /5 (40) 3/5 (60) S-B 12 o130 Lab.ref /8 (38) 5/8 (63) S-B 12 o148 B, RS I /12 (42) 8/12 (67) S-B 12 o221 C /52 (23) 36/52 (69) S-B 12 o258 B /74 (20) 48/74 (65) S-B 12 > /135 (7) 63/135 (47) P-folates o6.0 Lab.ref /3 (66) 3/3 (100) P-folates o8.6 RS I /16 (19) 13/16 (81) P-folates o10 D /28 (14) 21/28 (75) P-folates > /180 (12) 89/180 (49) B-folates o100 Lab.ref 0 0 B-folates o185 RS I /11 (27) 10/11 (91) B-folates > /198 (11) 103/198 (52) A, Joosten et al (1993) and Naurath et al (1995); B, Lindenbaum et al (1994); C, Yao et al (1992); D, Brouwer et al (1998). S-B 12 =serum B 12 concentration (pmol/l); P-folates=plasma folate concentration (nmol/l); B-folates=whole blood folate concentration (nmol/l). et al, 1998). The fractions of subjects with high metabolite values were larger in subgroups with low vitamin concentrations. The proportion of subjects with vitamin B 12 deficiency, defined as a high S-MMA concentration combined with low serum B 12, could be calculated to be from 1 to 7.2%, depending on the decision limit for serum B 12. Using the highest reported serum B 12 value compatible with deficiency (258 pmol/l (Lindenbaum et al, 1994)) and the reference limits for S-MMA of 0.38, 0.35 and 0.34 (corresponding to RS I, II and III), the proportion of vitamin B 12 deficient subjects could be calculated to 4.8, 6.2 and 7.2%, respectively. The proportion of subjects with folate deficiency, defined as a P-tHcys 416 mmol/l, with a low plasma folate value ranged from 1.4 to 10%, with a low whole blood folate value from 0 to 4.8%. The prevalence of folate deficiency, defined as plasma folates below 10 nmol/l (Brouwer et al, 1998) and P-tHcys values above those of RS I III, ranged from 1.9 to 11%. Significant changes in vitamin and metabolite concentrations after vitamin therapy Serum B 12, plasma folate and blood folate concentrations increased significantly in the vitamin as compared to the placebo group (Table 5). At the end of the study, the lowest values were, for serum B pmol/l, for plasma folates 25 nmol/l and, for blood folates 289 pmol/l. Mean P-tHcys values decreased by 32% and mean S-MMA by 14% in all vitamin-treated subjects. The corresponding figures for the healthy subsample (RS III) were 30 and 10%, respectively. In the placebo group, there were no significant decreases in metabolite concentrations, and the differences in posttreatment values between the vitamin and placebo groups were highly significant except for S-MMA in the healthy subgroup (Table 5). The distributions (cumulative frequencies) of P-tHcys and S-MMA before and after treatment are shown in Figures 2 and 3. At the end of the study, all but a few of the 115 vitamintreated subjects showed metabolite concentrations within health-related reference limits. Thirteen subjects had P-tHcys values 416 mmol/l. They all showed lower P-tHcys after treatment, the highest baseline value being 33 mmol/l. These 13 subjects had a mean serum creatinine concentration of 141 mmol/l, as compared to 97.1 mmol/l in the remaining 102 subjects. Four subjects had S-MMA concentrations mmol/l after vitamin treatment. One individual had a stated post-treatment concentration of 0.64 mmol/l, obviously a false value as the pretreatment concentration was 0.16 mmol/l. The serum creatinine concentration was 91 mmol/l. The other three had increased serum creatinine concentrations. Three of the four subjects with S-MMA mmol/l had lower P-tHcys after treatment, two of which had concentrations slightly above 16 mmol/l. There were no significant differences in blood haemoglobin concentration or mean erythrocyte volume either in the vitamin or in the placebo group during the study. Multivariate analysis of independent variables for predicting decrease in P-tHcys and S-MMA In multivariate analyses of the total study group, changes in vitamin concentrations, baseline vitamin and metabolite concentrations and vitamin treatment were included in the

7 1432 Table 5 Mean (s.d.) P-tHcys, S-MMA, serum B 12, plasma and blood folates in the total study group and in the healthy subgroups (RS III) before and after four months of treatment Vitamin group Placebo group Change in blood components, vitamin group vs placebo group Group n Before After Group n Before After P P-tHcys All (5.7) 11.9 (2.8) All (4.7) 17.1 (5.4) o0.001 Healthy RS III (5.0) 11.7 (2.6) Healthy (3.1) 16.2 (2.9) o0.001 S-MMA All (0.1) 0.19 (0.1) All (0.1) 0.22 (0.1) Healthy RS III (0.1) 0.18 (0.1) Healthy (0.2) 0.24 (0.1) S-B 12 All (124) 569 (199) All (165) 337 (134) o0.001 Healthy RS III (121) 555 (182) Healthy (196) 349 (154) o0.001 P-folates All (6.5) 56.2 (7.7) All (5.1) 16.5 (5.3) o0.001 Healthy RS III (5.6) 56.5 (7.4) Healthy (5.2) 15.5 (4.2) o0.001 B-folates All (133) 854 (175) All (104) 359 (102) o0.001 Healthy RS III (113) 844 (156) Healthy (95) 344 (100) o0.001 Figure 2 Cumulative distributions of S-MMA in the vitamin group (n ¼ 115) before and after 4 months of daily oral supplementation with 0.5 mg cyanocobalamin, 0.8 mg folic acid and 3 mg pyridoxine. Figure 3 Cumulative distributions of P-tHcys in the vitamin group (n ¼ 115) before and after 4 months of daily oral supplementation with 0.5 mg cyanocobalamin, 0.8 mg folic acid and 3 mg pyridoxine. model. Increase in P-folate, baseline P-tHcy and vitamin treatment correlated independently to P-tHcy decline (cumulative r 2 ¼ 0.68). For S-MMA, increases in S-B 12 and B-folate, baseline S-B 12 and S-MMA were independently correlated to metabolite decline (cumulative r 2 ¼ 0.38). In the vitamin group, the mean decrease in P-tHcys concentration in baseline quintile order highest to lowest was 11.2, 6.3, 4.8, 3.6 and 1.9 mmol/l, respectively. The corresponding figures for the S-MMA quintiles were 0.12, 0.03, 0.02, 0.01 and mmol/l. In multivariate analyses, high baseline metabolite concentrations, low vitamin concentrations, low age (for S-MMA) and low transferrin saturation (for P-tHcys) were independently correlated to metabolite decline (Table 6). Discussion Oral supplementation with 0.5 mg cyanocobalamin, 0.8 mg folic acid and 3 mg pyridoxine produced significant declines in P-tHcys and S-MMA in this community-dwelling population, whose health status was probably better than a representative urban Swedish population sample. Needless to say, these findings cannot be generalised to other, nonhealthy elderly populations. The metabolite decline

8 Table 6 Multivariate analysis, in the vitamin group, of the correlation between change in S-MMA (DS-MMA) and P-tHcys (DP-tHcys), respectively, and variables at the start of study 1433 Measure n Significant explanatory variables Regression coefficient (b) P DS-MMA 115 S-MMA serum B Age DP-tHcys 115 P-tHcys Plasma folates Transferrin saturation Further variables included in the model: Sex, smoking habits, blood Hb, Ery - MCV, whole blood folates, serum creatinine and exclusion criteria according to Table 2. was clearly induced by the vitamin treatment, as illustrated by the significant differences between the vitamin and placebo groups and by multivariate analysis of the total study group. Data regarding confounders not associated with clinical disease (eg lifestyle factors and 5-MTHFR mutations) were not obtained, but presumably these variables were evenly distributed in the vitamin and placebo groups. The vitamin and placebo groups were well balanced (Table 1). The differences in mean baseline S-B 12 and P-tHcy values reached statistical significance. It seems, unlikely however, that these differences would indicate a true difference in vitamin status, since neither folate nor S-MMA concentrations differed. The vitamin doses were chosen in order also to treat pronounced deficiency of these vitamins. The vitamin B 12 dose of 0.5 mg is 250 times larger than the present US recommended daily allowance (RDA, (Russell & Suter, 1993)), and, as well as the dose of folic acid, judged to be adequate (Hathcock & Troendle, 1991; de Bree et al, 1997; Clarke, 1998; Wald et al, 2001). Compliance was good except in single subjects, who were therefore excluded from the calculations of effects of the vitamin treatment. The high percentages of elevated base-line P-tHcys (53%) and S-MMA (11%) using standard laboratory criteria are noteworthy. However, the definition of decision limits for high concentrations is essential. The percentages of elevated values were much lower (7 8%) using traditional health-related intervals, as calculated for the present study population. On the assumption that the criteria employed to select a healthy subsample might fail to detect a suboptimal vitamin status, we instead used another criterion, that is, subsequent biochemical response to vitamin supplementation. Hence, baseline health-related reference intervals for P- thcys were shifted to the left; the prevalence of high baseline values increased to 15 18%. S-MMA, on the other hand, was slightly higher in this group. The upper reference limits for healthy and vitamin-replete subjects for both P-tHcys and S-MMA were in accordance with laboratory standards and those commonly found in the literature. As in many reports, men showed higher P- thcys than the women even after vitamin supplementation. Reference intervals may differ between populations for both pre-analytical and analytical reasons (Rasmussen & Möller, 2000). The reference intervals for P-tHcys calculated by Selhub et al (1999) for vitamin-replete subjects in a large population study in the US were substantially lower than those of the present study. In subjects aged 60 y or older, reference intervals defined as the 5th and 95th percentiles (ie the 0.90 central interfractile interval) were mmol/l in men, mmol/l in women. An estimate of the 95th percentile values for thcys in vitamin-replete subjects in the present study would be 17.7 mmol/l in men, 15.0 mmol/l in women. Rasmussen et al (1996) found lower reference intervals in a vitamin-replete but somewhat younger population. Vitamin supplementation had no side effects and produced significant declines in both P-tHcys and S-MMA. In multivariate analyses of the vitamin group, serum B 12 and plasma folates correlated independently to rate of decline of S-MMA and P-tHcys, respectively. After vitamin treatment, single subjects with higher serum creatinine concentrations had metabolite concentrations above the reference limits of the laboratory. Somewhat unexpectedly (Norlund et al, 1998; Bostom et al, 1999; Brattström & Wilcken, 2000; Hvas et al, 2000), metabolite concentrations did not correlate to serum creatinine. This might call for a better marker for renal function in this elderly population. Vitamin B 12 deficiency occurred in 1 7.2%, depending on the decision limits for S-B 12 and S-MMA. The prevalence of folate deficiency, defined as elevated P-tHcys in combinations with low plasma or whole blood folates, was at most 10 11%. There were no signs of haematologically significant deficiency, either at baseline or as disclosed by vitamin supplementation. Thus, frank vitamin deficiency could hardly explain the large proportion of subjects with elevated P-tHcys and S-MMA. However, suboptimal vitamin status in spite of normal vitamin concentrations was considered a significant factor for high metabolite concentrations, based on three circumstances. First, mean vitamin concentrations in the apparently healthy subgroup (RS I) increased by further exclusion of subjects with high metabolite concentrations (data not shown) and the proportions of subjects with high metabolite values seemed larger in groups with low vitamin concentrations. Second, baseline serum B 12 and plasma folates correlated, in a multivariate analysis, independently

9 1434 with the metabolite decline induced by vitamin supplementation. Third, the metabolite distributions in the vitamin group after supplementation were similar to the laboratory (health-related) standard reference interval. The reasons for any mild vitamin deficiency in this community-dwelling and mostly healthy population might, as in any elderly population, be complex. Vitamin malabsorption due to age-related changes in the gastric and/or intestinal mucosa, on the other hand, is probably rather frequent. To our knowledge, severe malnutrition did not occur, or, at most, was very uncommon. The fact, however, that transferrin saturation correlated independently with P-tHcys decline might also indicate a suboptimal nutritional status. It seems unlikely that a low transferrin saturation in this material would indicate a nonhealthy state. In the total study group, only 5% had a transferrin saturation below 16% (data not shown). Suboptimal nutrition in a population with a health status above the average might be an argument for food fortification. In Sweden, food iron fortification was abolished in 1995 (Olsson et al, 1997), and folate fortification is now being considered, nota bene with respect to neural tube defects. However, since the start of food folate fortification programmes, for example in the US, there has been concern that this might have negative effects on elderly subjects with mild vitamin B 12 deficiency (Rothenberg, 1999). The introduction of any fortification requires a systematic approach and the safety issues are not entirely solved (Wharton & Booth, 2001). The Food and Drug Administration has stated that a folate intake of 1 mg/day is the upper safe limit for untreated vitamin B 12 deficiency (Berg, 1999). In the present study, both suboptimal folate and vitamin B 12 status were corrected. In conclusion, elevated concentrations of S-MMA and P-tHcys were common in this healthy community-dwelling population with a median age of 76 years. Four months of daily substitution with a combination tablet containing 0.5 mg vitamin B 12, 0.8 mg folic acid and 3 mg of vitamin B 6 produced significant declines in S-MMA and P-tHcys concentrations. Suboptimal vitamin status is an important factor behind abnormal metabolite concentrations in healthy elderly subjects. Acknowledgements We are grateful to Recip AB for support of the study and for providing the vitamin and placebo tablets. References Adams JF, Clark JS, Ireland JT, Kesson CM & Watson WS (1983): Malabsorption of vitamin B12 and intrinsic factor secretion during biguanide therapy. Diabetologia 24, Alpert JE & Fava M (1997): Nutrition and depression: the role of folate. Nutr. Rev. 55, Apeland T, Mansoor MA, Strandjord RE, Vefring H & Kristensen O (2001): Folate, homocysteine and methionine loading in patients on carbamazepine. Acta Neurol. Scand. 103, Augustsson O, Eriksson BG, Rosenhall U, Rothenberg E, Warne B & Steen B (1994): The Johanneberg studyfa sociomedical survey in an urban elderly population. I. General presentation of the study including an analysis of non-response and identification of risk groups. Scand. J. Soc. Med. 22, Bailey LB & Gregory JF (1999): Polymorphisms of methylenetetrahydrofolate reductase and other enzymes: metabolic significance, risks and impact on folate requirement. J. Nutr. 129, Bald E & Sypniewski S (1994): Ion-pair high-performance liquid chromatography of cysteine and metabolically related compounds in the form of their S-pyridinium derivates. J. Chromatogr. A 676, Bates CJ, Pentieva KD, Prentice A, Mansoor MA & Finch S (1999): Plasma pyridoxal phosphate and pyridoxic acid and their relationship to plasma homocysteine in a representative sample of British men and women aged 65 years and over. Br. J. Nutr. 81, Belaiche J, Zittoun J, Marquet J, Nurit Y & Yvart J (1983): Effet de la ranitidine sur la secretion de facteur intrinseque gastrique et sur l absorption de la vitamine B12. Gastroenterol. Clin. Biol. 7, Berg MJ (1999): The importance of folic acid. J. Gend. Specif. Med. 2, Bostom AG, Bausserman L, Jacques PF, Liaugaudas G, Selhub J & Rosenberg IH (1999): Cystatin c as a determinant of fasting plasma total homocysteine levels in coronary artery disease patients with normal serum creatinine. Arterioscler. Thromb. Vasc. Biol. 19, Bottiglieri T (1996): Folate, vitamin B-12, and neuropsychiatric disorders. Nutr. Rev. 54, Brattström L & Wilcken DE (2000): Homocysteine and cardiovascular disease: cause or effect? Am. J. Clin. Nutr. 72, Brouwer DAJ, Welten HTME, Reijngoud DJ, Vandoormaal JJ & Muskiet FAJ (1998): Plasma folic acid cutoff value, derived from its relationship with homocyst(e)ine. Clin. Chem. 44, Carmel R, Sinow RM & Karnaze DS (1987): Atypical cobalamin deficiency. Subtle biochemical evidence of deficiency is commonly demonstrable in patients without megaloblastic anemia and is often associated with protein-bound cobalamin malabsorption. J. Lab. Clin. Med. 109, Chanarin I & Metz J (1998): Diagnosis of cobalamin deficiency: a reply. Br. J. Haematol. 100, Clarke R (1998): Lowering blood homocysteine with folic acid based supplements: meta-analysis of randomised trials. Homocysteine Lowering Trialists Collaboration. BMJ 316, de Bree A, van Dusseldorp M, Brouwer IA, van het Hof KH & Steegers- Theunissen RP (1997): Folate intake in europe: recommended, actual and desired intake. Eur. J. Clin. Nutr. 51, de Bree A, Verschuren WM, Blom HJ & Kromhout D (2001): Lifestyle factors and plasma homocysteine concentrations in a general population sample. Am. J. Epidemiol. 154, Fava M, Borus JS, Alpert JE, Nierenberg AA, Rosenbaum JF & Bottiglieri T (1997): Folate, vitamin B12, and homocysteine in major depressive disorder. Am. J. Psychiatry 154, Fioravanti M, Ferrario E, Massaia M, Cappa G, Rivolta G, Grossi E & Buckley AE (1997): Low folate levels in the cognitive decline of elderly patients and the efficacy of folate as a treatment for improving memory deficits. Arch. Gerontol. Geriatr. 26, Haller J, Lowik MR, Ferry M & Ferro Luzzi A (1991): Nutritional status: blood vitamins A, E, B6, B12, folic acid and carotene. Euronut SENECA investigators. Eur. J. Clin. Nutr. 45(Suppl 3), Hathcock JN & Troendle GJ (1991): Oral cobalamin for treatment of pernicious anemia? JAMA 265, Healton EB, Savage DG, Brust JC, Garrett TJ & Lindenbaum J (1991): Neurologic aspects of cobalamin deficiency. Medicine (Baltimore) 70,

10 Herrmann W, Quast S, Ullrich M, Schultze H, Bodis M & Geisel J (1999): Hyperhomocysteinemia in high-aged subjects: relation of b-vitamins, folic acid, renal function and the methylenetetrahydrofolate reductase mutation. Atherosclerosis 144, Hussein WI, Green R, Jacobsen DW & Faiman C (1999): Normalization of hyperhomocysteinemia with l-thyroxine in hypothyroidism. Ann. Intern. Med. 131, Hvas AM, Ellegaard J & Nexo E (2001): Vitamin B12 treatment normalizes metabolic markers but has limited clinical effect: a randomized placebo-controlled study. Clin. Chem. 47, Hvas AM, Juul S, Gerdes LU & Nexo E (2000): The marker of cobalamin deficiency, plasma methylmalonic acid, correlates to plasma creatinine. J. Intern. Med. 247, Ihamäki T, Varis K & Siurala M (1979): Morphological, functional and immunological state of the gastric mucosa in gastric carcinoma families. Comparison with a computer-matched family sample. Scand. J. Gastroenterol. 11, Jacques PF, Bostom AG, Williams RR, Ellison RC, Eckfeldt JH, Rosenberg IH, Selhub J & Rozen R (1996): Relation between folate status, a common mutation in methylenetetrahydrofolate reductase, and plasma homocysteine concentrations. Circulation 93, 7 9. Jones BP, Broomhead AF, Kwan YL & Grace CS (1987): Incidence and clinical significance of protein-bound vitamin B12 malabsorption. Eur. J. Haematol. 38, Joosten E, van den Berg A, Riezler R, Naurath HJ, Lindenbaum J, Stabler SP & Allen RH (1993): Metabolic evidence that deficiencies of vitamin B-12 (cobalamin), folate, and vitamin B-6 occur commonly in elderly people. Am. J. Clin. Nutr. 58, Kittang E & Schjönsby H (1987): Effect of gastric anacidity on the release of cobalamins from food and their subsequent binding to R-protein. Scand. J. Gastroenterol. 22, Krasinski SD, Russell RM, Samloff IM, Jacob RA, Dallal GE, McGandy RB & Hartz SC (1986): Fundic atrophic gastritis in an elderly population. Effect on hemoglobin and several serum nutritional indicators. J. Am. Geriatr. Soc. 34, Lehmann M, Gottfries CG & Regland B (1999): Identification of cognitive impairment in the elderly: homocysteine is an early marker. Dementia Geriatr. Cognitive Disorders 10, Lindenbaum J, Healton EB, Savage DG, Brust JCM, Garrett TJ, Podell ER, Marcell PD, Stabler SP & Allen RH (1988): Neuropsychiatric disorders caused by cobalamin deficiency in the absence of anemia or macrocytosis. N. Engl. J. Med. 318, Lindenbaum J, Rosenberg IH, Wilson PW, Stabler SP & Allen RH (1994): Prevalence of cobalamin deficiency in the Framingham elderly population. Am. J. Clin. Nutr. 60, Lindenbaum J, Savage DG, Stabler SP & Allen RH (1990): Diagnosis of cobalamin deficiency II: relative sensitivities of serum cobalamin, methylmalonic acid, and total homocysteine concentrations. Am. J. Hematol. 34, Lindstedt G (1999): Nitrous oxide may cause cobalamin deficiency syndrome; vitamin B12 a simple and inexpensive remedy. Lakartidningen 96, Matthews JH (1995): Cobalamin and folate deficiency in the elderly. Baillieres. Clin. Haematol. 8, McKinley MC, McNulty H, McPartlin J, Strain JJ, Pentieva K, Ward M, Weir DG & Scott JM (2001): Low-dose vitamin B-6 effectively lowers fasting plasma homocysteine in healthy elderly persons who are folate and riboflavin replete. Am. J. Clin. Nutr. 73, Molloy AM, Daly S, Mills JL, Kirke PN, Whitehead AS, Ramsbottom D, Conley MR, Weir DG & Scott JM (1997): Thermolabile variant of 5,10-methylenetetrahydrofolate reductase associated with low red-cell folates: implications for folate intake recommendations. Lancet 349, Morgan SL, Baggott JE, Lee JY & Alarcon GS (1998): Folic acid supplementation prevents deficient blood folate levels and hyperhomocysteinemia during longterm, low dose methotrexate therapy for rheumatoid arthritis: implications for cardiovascular disease prevention. J. Rheumatol. 25, Naurath HJ, Joosten E, Riezler R, Stabler SP, Allen RH & Lindenbaum J (1995): Effects of vitamin B12, folate, and vitamin B6 supplements in elderly people with normal serum vitamin concentrations. Lancet 346, Nilsson-Ehle H (1998): Age-related changes in cobalamin (vitamin B12) handling. Implications for therapy. Drugs Aging 12, Nilsson-Ehle H, Jagenburg R, Landahl S, Lindstedt S, Svanborg A & Westin J (1991): Serum cobalamins in the elderly: a longitudinal study of a representative population sample from age 70 to 81. Eur. J. Haematol. 47, Nilsson-Ehle H, Landahl S, Lindstedt G, Netterblad L, Stockbruegger R, Westin J & Åhren C (1989): Low serum cobalamin levels in a population study of 70- and 75-year-old subjects. Gastrointestinal causes and hematological effects. Dig. Dis. Sci. 34, Norlund L, Grubb A, Fex G, Leksell H, Nilsson JE, Schenck H & Hultberg B (1998): The increase of plasma homocysteine concentrations with age is partly due to the deterioration of renal function as determined by plasma cystatin C. Clin. Chem. Lab. Med. 36, Nygård O, Refsum H, Ueland PM & Vollset SE (1998): Major lifestyle determinants of plasma total homocysteine distribution: the Hordaland homocysteine study. Am. J. Clin. Nutr. 67, Olsson KS, Vaisanen M, Konar J & Bruce A (1997): The effect of withdrawal of food iron fortification in Sweden as studied with phlebotomy in subjects with genetic hemochromatosis. Eur. J. Clin. Nutr. 51, Rasmussen K (1989): Solid-phase sample extraction for rapid determination of methylmalonic acid in serum and urine by a stable-isotope-dilution method. Clin. Chem. 35, Rasmussen K & Möller J (2000): Total homocysteine measurement in clinical practice. Ann. Clin. Biochem. 37, Rasmussen K, Moller J, Lyngbak M, Pedersen AM & Dybkjaer L (1996): Age- and gender-specific reference intervals for total homocysteine and methylmalonic acid in plasma before and after vitamin supplementation [see comments]. Clin. Chem. 42, Rothenberg SP (1999): Increasing the dietary intake of folate: Pros and cons. Semin. Hematol. 36, Russell RM & Suter PM (1993): Vitamin requirements of elderly people: an update. Am. J. Clin. Nutr. 58, Schumann K (1999): Interactions between drugs and vitamins at advanced age. Int. J. Vitam. Nutr Res 69, Selhub J, Jacques PF, Rosenberg IH, Rogers G, Bowman BA, Gunter EW, Wright JD & Johnson CL (1999): Serum total homocysteine concentrations in the third national health and nutrition examination survey ( ): population reference ranges and contribution of vitamin status to high serum concentrations. Ann. Intern. Med. 131, Solberg HE (1983): The theory of reference values Part 5. Statistical treatment of collected reference values. Determination of reference limits. J. Clin. Chem. Clin. Biochem. 21, Stabler SP, Marcell PD, Podell ER, Allen RH, Savage DG & Lindenbaum J (1988): Elevation of total homocysteine in the serum of patients with cobalamin or folate deficiency detected by capillary gas chromatography mass spectrometry. J. Clin. Invest. 81, Stolzenberg-Solomon RZ, Miller ER, Maguire MG, Selhub J & Appel LJ (1999): Association of dietary protein intake and coffee consumption with serum homocysteine concentrations in an older population. Am. J. Clin. Nutr. 69, Streeter AM, Shum H-Y, Duncombe VM, Hewson JW & Thorpe MEC (1976): Vitamin B12 malabsorption associated with a normal Schilling test result. Med. J. Aust. 1, Ubbink JB, van der Merwe A, Delport R, Allen RH, Stabler SP, Riezler R & Vermaak WJ (1996): The effect of a subnormal vitamin B-6 status on homocysteine metabolism. J. Clin. Invest. 98,

11 1436 Wald DS, Bishop L, Wald NJ, Law M, Hennessy E, Weir D, McPartlin J & Scott J (2001): Randomized trial of folic acid supplementation and serum homocysteine levels. Arch. Intern. Med. 161, Watanabe F, Abe K, Fujita T, Goto M, Hiemori M & Nakano Y (1998): Effects of microwave heating on the loss of vitamin B-12 in foods. J. Agric. Food Chem. 46, Webster SGP & Leeming JT (1979): Erythrocyte folate levels in young and old. J. Am. Geriatr. Soc. 27, Wharton B & Booth I (2001): Fortification of flour with folic acid. BMJ 323, Yao Y, Yao SL, Yao SS, Yao G & Lou W (1992): Prevalence of vitamin B12 deficiency among geriatric outpatients. J. Fam. Pract. 35, Zimmerman J, Selhub J & Rosenberg IH (1987): Competitive inhibition of folate absorption by dihydrofolate reductase inhibitors, trimethoprim and pyrimethamine. Am. J. Clin. Nutr. 46,