Vitamin B12 deficiency in the aged: a population-based study

Transcription

1 Vitamin B12 deficiency in the aged Age and Ageing 2007; 36: The Author Published by Oxford University Press on behalf of the British Geriatrics Society. doi: /ageing/afl150 All rights reserved. For Permissions, please Published electronically 21 December 2006 Vitamin B12 deficiency in the aged: a population-based study SAILA LOIKAS 1,2,PERTTI KOSKINEN 1,2,KERTTU IRJALA 1,3,MINNA LÖPPÖNEN 4,5,RAIMO ISOAHO 4,6,SIRKKA-LIISA KIVELÄ 4,7,8,TARJA-TERTTU PELLINIEMI 1,2 1 Department of Clinical Chemistry, University of Turku, Turku, Finland 2 TYKSLAB, Hospital District of Southwest Finland, Turku, Finland 3 Mehiläinen Oy, Turku, Finland 4 Institute of Clinical Medicine, Family Medicine, University of Turku, Turku, Finland 5 Härkätie Health Centre, Lieto, Finland 6 Pori Health Centre, Pori, Finland 7 Unit of Family Medicine, Turku University Central Hospital, Turku, Finland 8 Satakunta Central Hospital, Pori, Finland Address correspondence to: S. Loikas. Fax: sailalo@utu.fi Abstract Background: vitamin B12 deficiency is common in the aged, but it is controversial whether only some risk groups should be investigated instead of screening the entire aged population. Objectives: to describe the prevalence of vitamin B12 deficiency in the Finnish aged, and to find out if the subjects especially prone to vitamin B12 deficiency could be identified by the risk factors or clinical correlates. Design: a cross-sectional, population-based study of 1048 aged subjects (age years) was carried out. Data on lifestyle factors and clinical conditions were collected, physical examinations were conducted and laboratory variables related to vitamin B12 were measured. Results: vitamin B12 deficiency had been previously diagnosed in 27 (2.6%) subjects, and a laboratory diagnosis (total vitamin B12 <150 pmol/l, or total vitamin B pmol/l and holotranscobalamin 37 pmol/l and homocysteine 15 µmol/l) was made for 97 (9.5%) subjects. Low serum total vitamin B12 (<150 pmol/l) was observed in 6.1% and borderline total vitamin B12 ( pmol/l) in 32% of the subjects. Male gender (OR 1.9, 95% CI ), age 75 (OR 2.2, 95% CI ) and refraining from milk products (OR 2.3, 95% CI ) increased the probability for vitamin B12 deficiency. Anaemia (OR 1.3, 95% CI ) or macrocytosis (OR 1.2, 95% CI ) did not predict vitamin B12 deficiency. Conclusion: undiagnosed vitamin B12 deficiency is remarkably common in the aged, but no specific risk group for screening can be identified. Thus, biochemical screening of unselected aged population is justified. General practitioners play a key role in diagnosing early vitamin B12 deficiency. Keywords: vitamin B12, holotranscobalamin, vitamin B12 deficiency, aged, screening, cohort study Introduction Previously, vitamin B12 deficiency was considered a rare disease, but easy to diagnose because of the dramatic findings of megaloblastic anaemia. At present, it is known that vitamin B12 deficiency is common and that its prevalence increases with age. It is estimated to affect 5 40% of the aged, depending on the diagnostic criteria used [1 5]. However, data about vitamin B12-related factors in unselected aged populations are scarce and many previous studies have been based on small and selective samples. 177

2 S. Loikas et al. Advanced stages of vitamin B12 deficiency seldom cause diagnostic problems, but early diagnosis is complicated by limitations of the laboratory methods and by subtle, non-specific symptoms [6, 7]. However, early recognition is essential for preventing irreversible damage. General practitioners are in the forefront in examining the aged with vague symptoms. Because typical signs and symptoms are frequently absent in early vitamin B12 deficiency [6], concern should be focused on persons with known risk factors [7]. Lifestyle factors, such as smoking, alcoholism and vegetarian diet, may predispose to vitamin B12 deficiency [5, 8]. Gastrointestinal diseases [9], gastric acid suppressive drugs [10] and metformin [11] may increase the probability of cobalamin malabsorption. Pernicious anaemia is known to associate with other autoimmune diseases [12]. In this study we report the prevalence of vitamin B12 deficiency in a representative sample of the Finnish aged population, and analyse known risk factors or possible clinical correlates of vitamin B12 deficiency in order to find out if a specific risk group among the aged can be identified for screening. Subjects and methods The lieto study subjects Lieto is a rural district in southwestern Finland. It has a population of about 13,000, with 11% aged 65 years. The Lieto study is a population-based health survey on unselected aged population living in this area [13]. The data were collected by the Department of Family Medicine of the University of Turku, the Department of Clinical Chemistry of the Turku University Central Hospital and the municipal health centre of Lieto during the period from March 1998 to December Ethical approval was obtained from the Ethics Committee of the Hospital District of Southwest Finland. All participants gave written informed consent. All Lieto residents born in 1933 or earlier (n = 1596) were invited to participate in the study. The study population consisted of 1048 aged persons (age years) (Figure 1). Interview, physical examination and biochemical analyses Two trained nurses interviewed all subjects. Demographic data, subjective health status, items on functional abilities, lifestyle factors, medication and vitamin supplementation were recorded. Mini-Mental State Examination (MMSE) and Zung Self-Rating Depression Scale were administered. Previous diseases were registered from the medical records. A 12-lead resting electrocardiogram (ECG) was recorded. Physical examination was performed by one of the two research physicians, who were both experienced general practitioners. Medical conditions were assessed during the examination. Venous blood samples were drawn after an overnight fast. Complete blood count, serum total vitamin B12, holotranscobalamin (holotc), plasma total homocysteine (thcy) and erythrocyte (ERC) folate were measured. For a description of the clinical outcome measures and biochemical analyses, see Appendix in the supplementary data (available online at Statistical analyses The differences between the groups were tested with the Kruskal Wallis nonparametric test for quantitative data and the chi-square test for categorical data. Spearman s rankorder correlation coefficients were calculated for measuring correlations. Logistic regression was used to estimate odds ratios (OR) and their 95% confidence interval (CI) for the associations between vitamin B12 deficiency and selected clinical outcome measures. P values <0.05 were considered as significant. The data were analysed using the SAS software version 8.2 (SAS Institute Inc., Cary, NC, USA). Results Prevalence of vitamin B12 deficiency in the finnish aged Previously diagnosed vitamin B12 deficiency In the study population, 27 (2.6%) subjects were receiving vitamin B12 substitution (hydroxycobalamin or cyanocobalamin injections). The diagnoses had mainly been based on low total vitamin B12 concentration and macrocytosis. Those with this condition were significantly older (median age 81 vs 73 years, P<0.001), and a larger proportion of them lived in a long-term institution (30% vs 6.0%, P<0.001) and were unmarried, divorced or widowed (63% vs 44% P = 0.045), compared to the subjects without vitamin B12 substitution. They showed more dependency in their capability for performing Activities of Daily Living (ADL) (32% vs 11%, P = 0.001), Instrumental Activities of Daily Living (IADL) (60% vs 39%, P = 0.035) and mobility scores (33% vs 14%, P = 0.005), and used more medications (6 vs 3, P = 0.001). There was no significant difference in multivitamin use, lifestyle factors, subjective health status, subjective life satisfaction or bodily pain. Adjustment for age and gender eliminated the significant difference only in IADL, mobility scores and marital status. Dementia was more common in the vitamin B12-substituted subjects (26% vs 8.2%, P = 0.001), but there were no significant differences in subjective assessments of memory impairment. Of the other medical conditions analysed, only the occurrence of gastritis showed a significant difference between the subjects with and without vitamin B12 substitution (19% vs 5.8%, P = 0.006). Previously undiagnosed vitamin B12 deficiency Subjects with previously diagnosed vitamin B12 deficiency were excluded from further analyses, leaving a study population of 1021 subjects. Their vitamin B12- related biochemical variables are presented in Table 1. Total 178

3 Vitamin B12 deficiency in the aged Figure 1. Study design. vitamin B12 was low (<150 pmol/l) in 6.1% (n = 62), borderline ( pmol/l) in 32% (n = 324) and normal (>250 pmol/l) in 62% (n = 635) of the subjects. thcy was increased ( 15 µmol/l) in 74% (n = 46) of the subjects with low total vitamin B12, in 41% (n = 132) of the subjects with borderline total vitamin B12 and in 26% (n = 164) of the subjects with normal total vitamin B12. The corresponding proportions of low holotc ( 37 pmol/l) were 77% (n = 48), 17% (n = 56) and 1.3% (n = 8). For the other variables, see Table A1 in the supplementary data. The total vitamin B12 concentration correlated positively with holotc (r = 0.78, P<0.001) and ERC 179

4 S. Loikas et al. Table 1. Age and biochemical variables in the study population a Total study group II Men Women n = 1021 n = 423 (41%) n = 598 (59%) P value b... Age 75 years 382 (37%) 134 (32%) 248 (41%) Serum total vitamin B12 (pmol/l) 281 ( ) 262 ( ) 293 ( ) <0.001 Borderline total vitamin B12 ( pmol/l) 324 (32%) 157 (37%) 167 (28%) <0.001 Low total vitamin B12 ( 150 pmol/l) 62 (6.1%) 34 (8.0%) 28 (4.7%) Serum HoloTC (pmol/l) 71 (<20 790) 64 (<20 614) 76 (<20 790) <0.001 Low holotc ( 37 pmol/l) 112 (11%) 62 (15%) 50 (8.4%) Plasma thcy (µmol/l) 13 (1.3 80) 13 (1.3 80) 12 (5.3 62) <0.001 Increased thcy ( 15 µmol/l) 342 (34%) 159 (38%) 183 (31%) Erythrocyte folate (nmol/l) 400 ( ) 396 ( ) 404 ( ) Low erythrocyte folate (<320 nmol/l) 213 (21%) 88 (21%) 125 (21%) Hemoglobin (g/l) 136 (91 182) 144 (92 182) 132 (91 181) <0.001 Anaemia c 137 (13%) 91 (22%) 46 (7.7%) <0.001 MCV (fl) 92 (68 115) 93 (68 111) 92 (68 115) <0.001 Macrocytosis (MCV>98 fl) 67 (7%) 36 (8.5%) 31 (5.2%) Macrocytic anaemia 14 (1.4%) 7 (1.7%) 7 (1.2%) a Median and range; number and percentage of pathological results. b Differences between men and women (Kruskal Wallis nonparametric test for quantitative data and chi-square test for categorical data). A P value <0.05 is considered significant. c Hemoglobin concentration <134 g/l for men and <117 g/l for women. holotc, holotranscobalamin; thcy, total homocysteine; MCV, mean erythrocyte volume. folate (r = 0.30, P<0.001) and negatively with thcy (r = 0.32, P<0.001). In addition, there was a weak but significant correlation between milk product consumption and total vitamin B12 (r = 0.21, P<0.001). thcy correlated more strongly with ERC folate (r = 0.39, P<0.001), age (r = 0.37, P<0.001) and holotc (r = 0.33, P<0.001) than with total vitamin B12, and weakly but significantly with MMSE score (r = 0.25, P<0.001). There were no correlations between hemoglobin or MCV and vitamin B12-related variables. The subjects with low total vitamin B12 (n = 62) or borderline total vitamin B12 and low holotc together with increased thcy (n = 35) were considered to have previously undiagnosed vitamin B12 deficiency (9.5%, n = 97) (Figure 1). Significantly they more often had low ERC folate and anaemia than the rest of the study population. The rates of macrocytosis and macrocytic anaemia did not differ significantly. Five subjects had low total vitamin B12 but no pathological results in any other vitamin B12-related variables. Prevalence of vitamin B12 deficiency Vitamin B12 deficiency had earlier been diagnosed in 27 of the 1048 subjects (2.7%), and a new laboratory diagnosis of vitamin B12 deficiency was made for 97 subjects. Taken together, the prevalence of vitamin B12 deficiency was 12% (n = 124) (Figure 1). Vitamin B12 deficiency was more common in men than in women (12% vs 7.4%, P = 0.018). In this aged population, 78% of the subjects with vitamin B12 deficiency were previously undiagnosed. The associations of risk factors and clinical correlates to vitamin B12 deficiency In order to find out if a specific risk group among the aged could be found, the occurrence of vitamin B12 deficiency was comparatively assessed for the subjects with and without possible predisposing diseases or lifestyle factors or consequent clinical correlates. For further risk assessment, logistic regression analyses adjusted for age and gender were conducted (Table 2). The probability for low vitamin B12 was twofold greater for men compared to women (OR 1.9, 95% CI ) and for subjects aged 75 years compared to younger subjects (OR 2.2, 95% CI ). No associations were found between vitamin B12 deficiency and functional abilities or lifestyle factors, other than milk product use. The subjects who did not use milk products were two times more likely to have vitamin B12 deficiency (OR 2.3, 95% CI ). Of the diseases analysed, only diverticulosis significantly increased the probability of vitamin B12 deficiency, and a slight effect remained after adjustment for age and gender (OR 1.8, 95% CI ). The probability for vitamin B12 deficiency was lower than expected in subjects with depression and CHD. The association was weak but statistically significant (OR 0.5, 95% CI , and OR 0.6, 95% CI , respectively). This finding does not obviate the need for screening for vitamin B12 deficiency in these subjects. Vitamin B12 deficiency was also more frequent in the subjects with anaemia (15% vs 8.6%, 0.013), but adjustment for age and gender removed the effect. The presence of macrocytosis or macrocytic anaemia did not increase the probability of vitamin B12 deficiency. 180

5 Table 2. Associations of selected variables and vitamin B12 deficiency adjusted for age and gender Vitamin B12 deficiency in the aged Variable Prevalence of variable n (%) OR (95% CI) P value f... Demographic data Male gender 423 (41) 1.9 ( ) a Age (37) 2.2 ( ) b <0.001 Unmarried, divorced or widowed 445 (44) 1.5 ( ) Being resident in an institution 60 (5.9) 0.7 ( ) Living alone 302 (30) 1.4 ( ) Life style factors Vegetarian diet 16 (1.6) 1.4 ( ) No milk product use 85 (8.3) 2.3 ( ) Smoking 76 (7.5) 0.5 ( ) Heavy alcohol drinking 36 (3.6) 1.1 ( ) No multivitamin use 830 (81) 1.7 ( ) Functional abilities Poor subjective health 141 (14) 0.6 ( ) Poor subjective life satisfaction 191 (19) 1.0 ( ) Bodily pain 301 (30) 1.0 ( ) Poor subjective memory 139 (14) 0.7 ( ) MMSE score (11) 0.7 ( ) Zung score (19) 0.8 ( ) Dependency in ADL 109 (11) 1.1 ( ) Dependency in IADL 394 (39) 0.9 ( ) Dependency in mobility 142 (14) 0.8 ( ) Predisposing diseases Gastritis 59 (5.8) 1.1 ( ) Diverticulosis 109 (11) 1.8 ( ) Hypothyroidism 100 (9.8) 0.8 ( ) COPD 84 (8.2) 1.4 ( ) Possible clinical correlates Anaemia c 137 (13) 1.3 ( ) Macrocytosis d 67 (6.6) 1.2 ( ) Macrocytic anaemia 14 (1.4) 1.2 ( ) Dementia 84 (8.2) 0.9 ( ) Depression 196 (19) 0.5 ( ) CHD 425 (42) 0.6 ( ) Stroke 98 (9.6) 0.7 ( ) Medications Estrogens e 103 (17) 0.5 ( ) Gastric acid suppressive drugs 70 (6.9) 1.5 ( ) Metformin 48 (4.7) 0.8 ( ) Antibiotics 45 (4.4) 0.6 ( ) a Adjusted for age. b Adjusted for gender. c Hemoglobin concentration <134 g/l for men and <117 g/l for women. d Mean erythrocyte volume >98 fl. e Women only. f A P value <0.05 is considered significant. OR, odds ratio; CI, confidence limit; MMSE, Mini-Mental State Examination; ADL, Activities of Daily Living; IADL, Instrumental Activities of Daily Living; COPD, chronic obstructive pulmonary disease; CHD, coronary heart disease Discussion The prevalence of vitamin B12 deficiency was 12% in this representative sample of an aged population. This is in accordance with the previously reported prevalence rates in other countries [1, 4, 5]. The proportion of previously undiagnosed vitamin B12 deficiency was remarkably large. Male gender, age 75 years or more and refraining from milk products doubled the probability for vitamin B12 deficiency, but no clinically relevant risk group for screening could be defined. Anaemia or macrocytosis did not predict vitamin B12 deficiency. We used total vitamin B12 measurement as the first-line laboratory test because it is most commonly used by general practitioners in diagnosing vitamin B12 deficiency. It is inexpensive and widely available, but it has been considered not sensitive or specific enough to be used as the only laboratory test [2, 14]. The 181

6 S. Loikas et al. determination of thcy or holotc has been proposed as additional tests for subjects with borderline total vitamin B12 concentrations ( pmol/l) [7]. However, the prevalence of hyperhomocysteinemia is high in the aged and increases with age. The main determinants for plasma thcy concentrations are folate and vitamin B12 status as well as renal function [14]. The effect of impaired renal function and other diseases on thcy seems to increase with age, which reduces its value as a laboratory test for vitamin B12 deficiency [15]. To overcome the unspecificity of thcy, we also used low holotc in subjects with borderline total vitamin B12 for the laboratory diagnosis of vitamin B12 deficiency. False low total vitamin B12 concentrations have been found, but only a few reasons are known for them [16]. Recently, it has been shown that mild haptocorrin deficiency is more common than previously assumed and may explain as much as 15% of the low total vitamin B12 concentrations [16]. We are aware that our definition of vitamin B12 deficiency may have resulted in some overdiagnosis. Five subjects in this population could have been regarded as having false low total vitamin B12, because they had both normal thcy and normal holotc. We consider this justified since overdiagnosis of vitamin B12 deficiency is not as harmful as underdiagnosis, because no toxic effects have been reported from vitamin B12 supplementation. For this reason, we also consider additional holotc and thcy measurements efficient only in subjects with borderline total vitamin B12. Vitamin B12 deficiency seems to be markedly underdiagnosed. In this study cohort, 78% of the subjects with vitamin B12 deficiency were previously undiagnosed. In clinical practice, more awareness is needed to prevent irreversible damage due to delayed diagnosis. Therefore, screening for even asymptomatic aged people has been suggested [5, 17]. Another option is to recognise the early clinical manifestations or possible risk factors. However, only a few recent studies have analysed possible clinical associates other than cognitive impairment. According to them, the signs and symptoms traditionally linked to vitamin B12 deficiency are not related to the occurrence of the disease [6, 18, 19]. Smoking, alcoholism and vegetarian diet were not associated with vitamin B12 deficiency in this aged population. These factors have been linked with an increased risk of vitamin B12 deficiency in younger adults [5, 8], but, to our knowledge, these relationships have not been previously investigated in aged populations. Milk product use is the rule rather than an exception among the Finnish aged, and it correlated positively with total vitamin B12 concentrations. Refraining from milk products doubled the probability for vitamin B12 deficiency. Many previous studies have analysed the associations between cognitive function and vitamin B12 status [6, 20, 21], and the results have varied with the markers used. We used subjective memory impairment, MMSE and dementia diagnosis, which are rather insensitive measures for slight cognitive impairments. As in many other studies, we found no correlation between the MMSE score and total vitamin B12 or holotc, but thcy had an inverse correlation. In addition, vitamin B12 deficiency was not more frequent in subjects with poor subjective memory or low MMSE score. This is in contrast to a recent large study of people aged 75 years or more in the United Kingdom [21], in which the symptoms of memory impairment and low MMSE score were associated with all the measures of vitamin B12 status used in the study. The risk for cognitive impairment was greater in subjects with total vitamin B12 or holotc concentrations in the lowest quartile and thcy or methylmalonic acid concentrations in the highest quartile. In the same study, no association was found between any measures of vitamin B12 status and depression, but two previous population-based studies have shown depression to be a risk factor for vitamin B12 deficiency [22, 23]. Surprisingly, in our study, the probability for vitamin B12 deficiency was slightly lower than expected in subjects with depression and CHD. The result is most likely random error due to a large amount of variables analysed and does not obviate the need for screening for vitamin B12 deficiency in these subjects. Symptoms and signs of polyneuropathy were not recorded in this study. Bodily pain may be used as a poor indicator of polyneuropathy. Previously, bodily pain has been found to be more common in subjects with low vitamin B12 levels [24], but we found no association. In this population, anaemia or macrocytosis did not predict vitamin B12 deficiency, and if the subjects with merely macrocytic anaemia had been suspected for vitamin B12 deficiency, only two of the 97 subjects would have been diagnosed. Thus, we question the traditional hematological approach to vitamin B12 deficiency. We conclude that undiagnosed vitamin B12 deficiency is remarkably common. This suggests that, in the current clinical practice, only overt signs and symptoms trigger laboratory testing for vitamin B12 deficiency. Evidently, routine screening would result in earlier diagnosis and reduce the related disability. However, long-term trials are required to judge the relevance of treatment of asymptomatic individuals. No specific risk groups among the aged can be defined, but aging itself increases the probability of vitamin B12 deficiency. Therefore, possible routine screening should be done at least among persons aged 75 years or more. We propose the measurement of total vitamin B12 as the first-line test, with a cut-off limit of 150 pmol/l to confirm and 250 pmol/l to exclude vitamin B12 deficiency. Additional measurements of thcy and holotc are suggested for subjects with borderline total vitamin B12 concentrations ( pmol/l). However, there is variation in the local reference limits for the tests and therefore no uniform decision limits can be given. Key points Undiagnosed vitamin B12 deficiency is remarkably common in the aged. 182

7 Vitamin B12 deficiency in the aged No specific risk group for vitamin B12 deficiency can be defined among the aged. Routine screening should be directed at least to persons aged 75 years or more. Biochemical markers instead of macrocytosis should be used for screening. The role of general practitioners in diagnosing early vitamin B12 deficiency is emphasised. Acknowledgements We thank Mrs Paula Prinssi and Mrs Pirjo Itkonen for elaborate interviews, and the skilled laboratory personnel in the Turku University Central Hospital and the health centre of Lieto. We are indebted to Mr Jukka Saukkoriipi for managing the database. Conflicts of interest There are no conflicts of interests. Funding The reagents for total vitamin B12 and ERC folate measurements were provided by PerkinElmer Wallac and HoloTC RIA reagents were provided by Axis Shield ASA. This study was supported by the EVO-grant of the Turku University Central Hospital. References 1. Baik HW, Russell RM. Vitamin B12 deficiency in the elderly. Annu Rev Nutr 1999; 19: Carmel R. Current concepts in cobalamin deficiency. Annu Rev Med 2000; 51: Bjorkegren K, Svardsudd K. Serum cobalamin, folate, methylmalonic acid and total homocysteine as vitamin B12 and folate tissue deficiency markers amongst elderly Swedes a population-based study. J Intern Med 2001; 249: Clarke R, Grimley Evans J, Schneede J et al. Vitamin B12 and folate deficiency in later life. Age Ageing 2004; 33: Wolters M, Strohle A, Hahn A. Cobalamin: a critical vitamin in the elderly. Prev Med 2004; 39: Bjorkegren K, Svardsudd K. Reported symptoms and clinical findings in relation to serum cobalamin, folate, methylmalonic acid and total homocysteine among elderly swedes: a population-based study. J Intern Med 2003; 254: Schneede J, Ueland PM. Novel and established markers of cobalamin deficiency: complementary or exclusive diagnostic strategies. Semin Vasc Med 2005; 5: Herrmann W, Geisel J. Vegetarian lifestyle and monitoring of vitamin B-12 status. Clin Chim Acta 2002; 326: Gregg CR. Enteric bacterial flora and bacterial overgrowth syndrome. Semin Gastrointest Dis 2002; 13: Valuck RJ, Ruscin JM. A case-control study on adverse effects: H2 blocker or proton pump inhibitor use and risk of vitamin B12 deficiency in older adults. J Clin Epidemiol 2004; 57: Bauman WA, Shaw S, Jayatilleke E, Spungen AM, Herbert V. Increased intake of calcium reverses vitamin B12 malabsorption induced by metformin. Diabetes Care 2000; 23: Toh BH, van Driel IR, Gleeson PA. Pernicious anemia. N Engl J Med 1997; 337: The Lieto Study. html. Institute of Clinical Medicine, Family Medicine, University of Turku. Accessed March 15, Refsum H, Smith AD, Ueland PM et al. Facts and recommendations about total homocysteine determinations: an expert opinion. Clin Chem 2004; 50: Selhub J, Jacques PF, Rosenberg IH et al. 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 1999; 131: Carmel R. Mild transcobalamin I (haptocorrin) deficiency and low serum cobalamin concentrations. Clin Chem 2003; 49: Clarke R, Refsum H, Birks J et al. Screening for vitamin B-12 and folate deficiency in older persons. Am J Clin Nutr 2003; 77: Hvas AM, Ellegaard J, Nexo E. Increased plasma methylmalonic acid level does not predict clinical manifestations of vitamin B12 deficiency. Arch Intern Med 2001; 161: Hvas AM, Nexo E. Holotranscobalamin-a first choice assay for diagnosing early vitamin B deficiency? J Intern Med 2005; 257: Calvaresi E, Bryan J. B vitamins, cognition, and aging: a review. J Gerontol B Psychol Sci Soc Sci 2001; 56: Hin H, Clarke R, Sherliker P et al. Clinical relevance of low serum vitamin B12 concentrations in older people: the Banbury B12 study. Age Ageing 2006; 35: TiemeierH, vantuijl HR, Hofman Aet al. Vitamin B12, folate, and homocysteine in depression: the Rotterdam Study. Am J Psychiatry 2002; 159: Penninx BW, Guralnik JM, Ferrucci L et al. Vitamin B(12) deficiency and depression in physically disabled older women: epidemiologic evidence from the Women s Health and Aging Study. Am J Psychiatry 2000; 157: Bernard MA, Nakonezny PA, Kashner TM. The effect of vitamin B12 deficiency on older veterans and its relationship to health. J Am Geriatr Soc 1998; 46: Received 13 April 2006; accepted in revised form 31 October