Eur J Nutr DOI 10.1007/s00394-013-0647-y
ORIGINAL CONTRIBUTION
Three-month B vitamin supplementation in pre-school children affects folate status and homocysteine, but not cognitive performance Astrid Rauh-Pfeiffer • Uschi Handel • Hans Demmelmair • Wolfgang Peissner Mareile Niesser • Diego Moretti • Vanessa Martens • Sheila Wiseman • Judith Weichert • Moritz Heene • Markus Bu¨hner • Berthold Koletzko
•
Received: 16 July 2013 / Accepted: 18 December 2013 Ó Springer-Verlag Berlin Heidelberg 2014
Abstract Background Suboptimal vitamin B status might affect cognitive performance in early childhood. We tested the hypothesis that short-term supplementation with folic acid and selected B vitamins improves cognitive function in healthy children in a population with relatively low folate status. Methods We screened 1,002 kindergarten children for suboptimal folate status by assessing the total urinary paraaminobenzoylglutamate excretion. Two hundred and fifty low ranking subjects were recruited into a double blind, randomized, controlled trial to receive daily a sachet containing 220 lg folic acid, 1.1 mg vitamin B2, 0.73 mg B6, 1.2 lg B12 and 130 mg calcium, or calcium only for 3 months. Primary outcomes were changes in verbal IQ, short-term memory and processing speed between baseline and study end. Secondary outcomes were urinary markers
Berthold Koletzko is the recipient of a Freedom to Discover Award of the Bristol-Myers Squibb Foundation, New York, NY. A. Rauh-Pfeiffer U. Handel H. Demmelmair W. Peissner M. Niesser B. Koletzko (&) Dr von Hauner Children’s Hospital, University of Munich Medical Centre, Lindwurmstraße 4, 80337 Munich, Germany e-mail: [email protected] D. Moretti V. Martens S. Wiseman Unilever R&D, 3133 AT Vlaardingen, The Netherlands D. Moretti Laboratory of Human Nutrition, Institute of Food Science and Nutrition, ETH Zu¨rich, 8092 Zurich, Switzerland J. Weichert M. Heene M. Bu¨hner Department of Psychology, University of Munich, 80802 Munich, Germany
of folate and vitamin B12 status, acetyl-para-aminobenzoylglutamate and methylmalonic acid, respectively, and, in a subgroup of 120 participants, blood folate and plasma homocysteine. Results Pre- and post-intervention cognitive measurements were completed by 115 children in the intervention and 122 in the control group. Compared to control, median blood folate increased by about 50 % (P for difference, P \ 0.0001). Homocysteine decreased by 1.1 lmol/L compared to baseline, no change was seen in the control group (P for difference P \ 0.0001) and acetyl-para-aminobenzoylglutamate was 4 nmol/mmol higher compared to control at the end of the intervention (P \ 0.0001). We found no relevant differences between the groups for the cognitive measures. Conclusion Short-term improvement of folate and homocysteine status in healthy children does not appear to affect cognitive performance. Keywords B vitamin supplementation Cognition Pre-school children
Introduction Severe nutritional deficiencies in the vitamins B2, B6, B12 and folate (in the course of the paper also called vitamin B) cause neurological and cognitive abnormalities in humans, and indications exist that these vitamins may be important for cognitive performance when intake and status are suboptimal [1]. Folate and vitamins B12, B6 and B2 share a metabolic pathway that may affect the central nervous system by modulating one-carbon metabolism, methylation and neurotransmitter synthesis [1, 2]. Evidence from animal studies [3, 4] suggests an effect of folate/folic acid
123
Eur J Nutr
supplementation on long-term memory [3] and spatial learning [4], while mice with vitamin B12 and folate deficiency showed increased hyperactivity [5]. Limited human data suggest a role for folate, vitamin B6 and vitamin B12 in cognitive performance at a young age. Associative evidence from a cross-sectional study in 6-year-old children reported positive correlations between folate status and total and verbal intelligence quotient (IQ) [6] and folate status was significantly associated with cognitive performance in Indian infants [7]. One-month supplementation with folic acid and vitamin B6 was reported to be associated with recall performance and word recall in young adults between 20 and 30 years while vitamin B12 supplementation was associated with specific measures of short and long-term memory performance [8]. Similar associations were reported in a study of Dutch adolescents who had followed a vegan diet in childhood where a correlation was found between vitamin B12 status and measures of abstract reasoning, spatial ability and short-term memory at the age of 10–18 years [9]. In a feeding intervention study among primary school children in rural Kenya, the increased intake of vitamins B2 and B12 showed significantly higher gains in digit span-forward test scores [10]. Folate, vitamin B12 and vitamin B6 support the metabolic availability of methyl groups and thus facilitate the remethylation of homocysteine to methionine. Poor dietary folate and vitamin B12 intake leads to increased serum homocysteine concentrations in adults [11, 12] and may affect homocysteine levels and folate status in children. Folate status is associated with serum homocysteine concentrations in children above 2 years of age [13], however, to our knowledge, limited direct evidence exists on the effect of B vitamin supplementation on homocysteine in this age group. Suboptimal folate intake has been reported in the paediatric population in Germany [14, 15]. Urinary folate catabolites, acetyl-p-aminobenzoyglutamate (ap-ABG) and p-aminobenzoylglutamate (p-ABG), have been proposed as non-invasive markers of folate intake in humans, and particularly ap-ABG was reported to reflect long-term folate pools [16, 17]. Total p-ABG (sum of ap-ABG and p-ABG) was used to identify subjects with low folate status and to select a population with low folate intake. We aimed to test the hypothesis that short-term (3 months) dietary supplementation with folic acid and vitamins B12, B6 and B2 improves folate status, decreases homocysteine and affects measures of cognitive performance in healthy pre-school children in a population not exposed to folic acid-fortified foods. To our knowledge, no controlled intervention study on the effects of a combination of B vitamins and folic acid on cognitive and biochemical outcomes has been performed in pre-school children to date.
123
Subjects and methods Study design and study population This study aimed to enrol 250 apparently healthy children aged 4–6 years, attending kindergartens in the Munich, Germany area into the placebo or treatment groups of this randomized controlled trial (RCT), based on the power calculation reported below. Children who agreed to participate, and whose parents provided written informed consent, were asked to provide a spot early morning urine sample for measurement of the urinary folate catabolites ap- and p-ABG. Urine from 1002 children was screened in order to enrol those children with the lowest p-ABG excretion into the study. Children who took vitamin B/folic acid supplements or who were not fluent in the German language were excluded from the study. Families were asked to complete a questionnaire on socio-economic and demographical descriptors, and a food frequency questionnaire regarding the nutritional habits of the child focused on the intake of dietary folate equivalents (DFE). A battery of cognitive tests was administered to all participating subjects before and after the intervention. The primary outcome measures were changes in short-term memory, verbal IQ and mental speed between baseline and study end. Secondary outcome measures were urinary markers of folate and vitamin B12 status ap-ABG and methylmalonic acid (MMA), respectively, and, in the subgroup in which the subjects had agreed to blood sampling, whole blood folate and plasma homocysteine. This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving human subjects were approved by the Ethical Committee of the Medical Faculty of the University of Munich. The study was explained to the parents and to the children in oral and written form, using age-appropriate pictographic leaflets for children. Written informed consent was obtained from all subjects’ parents. The study followed the recommendations of the CONSORT guidelines [18]. The study was registered at clinicaltrials.gov under NCT00811291. Intervention Two hundred and fifty subjects ranking low with regard to urinary total p-ABG within the screening group were enrolled into the intervention trial which was conducted between March and August 2009. Subjects were randomly allocated to receive either the intervention or the control product by means of a computer-generated list of randomization codes. Participants, parents, kindergarten teachers and all members of the study team were blinded during the whole trial. The intervention product was
Eur J Nutr
provided as a flavourless powder in sachets containing folic acid (220 lg), riboflavin (1.1 mg), pyridoxine (0.73 mg), cobalamin (1.2 lg) and calcium lactate pentahydrate (130 mg). The control matched the intervention product in taste and appearance but contained only 130 mg calcium. The additional intake of B vitamins as well as the estimated overall vitamin B intake of the children participating in this trial was below the upper limits set by the European Food Safety Authority (EFSA) for children aged 4–6 years [19]. The sachets were delivered to the subjects at the time of their first study appointment. Parents were instructed to mix the contents of one sachet per day into their child’s breakfast or other meal over a period of 90 days at home. Children’s compliance was supported by small gifts and incentives. Families were advised not to provide any other vitamin supplements or multivitamin juices to the child and not to change their usual dietary habits during the intervention period. Compliance with provision of test products was encouraged and monitored throughout the trial by repeated phone calls and letters to the families, and supplement usage was entered into diaries supposed to be completed daily by the families. Parents were asked to collect two early morning urine samples from their child, one before the start of the intervention period and one at the end of the trial. In a subgroup of children who were willing to undergo blood sampling and whose parents approved, blood samples were collected by finger prick before and after the intervention. Blood samples were taken by experienced staff in a child-friendly atmosphere with the use of a lancing device (FineTouch, Terumo, Eschborn, Germany).
Table 1 Overview cognitive function measures Test
Measured abilities
Verbal scale (HAWIVA-III) Information
Long-term memory and reproduction
The child answers questions of general knowledge, e.g. ‘What do people use for writing?’
Vocabulary
Long-term memory and language development
The child gives definitions for words that are read by the test assistant, e.g. ‘What does ‘polite’ mean?’
Word reasoning
Reasoning and abilities to abstract
The child gives words to more and more detailed descriptions read by the test assistant, e.g. ‘It’s a room that people use to cook’
Processing speed quotient (HAWIVA-III) Symbol search
Sustained attention
The child scans a group of symbols for a match of one of them with a previously shown target symbol
Coding
Concentration, visualmotor co-ordination and speed of processing
Requires the child to draw the missing counterpart of nine pairs of symbols and geometric shapes from a given key
Sequential processing scale (K-ABC) Number recall
Short-term memory
The child repeats a string of numbers in the same order as read aloud by the test assistant before. Strings range from two to nine digits
Word order
Short-term memory
The child has to remember names of common objects read aloud by the test assistant. Then a series of pictures of these objects are shown to the child who has to touch them in the same order they were read out in
Cognitive tests Several subtests of two well-known intelligence test systems for pre-school children were used to assess cognitive performance. The ‘Hannover-Wechsler-Intelligenztest fu¨r das Vorschulalter-III’ (HAWIVA-III) [20], the German version of the Wechsler Pre-school and Primary Scale of Intelligence-III (WPPSI-III) [21], which is especially designed for children between the ages of 2 years 6 months and 6 years 11 months. A full-scale IQ is provided as well as several subtest scores in verbal and performance cognitive domains. In addition, subscales of the KaufmanAssessment Battery for Children (K-ABC) [22] in the German version [23] were chosen, as an instrument for assessing the cognitive development in children aged 2.5–12.5 years. It comprises four global test scores that include sequential processing scales, simultaneous processing scales, achievement scales and mental processing composite. For both instruments, it applies that quotient and composite scores have a mean of 100 and a standard deviation of 15, while subtest scaled scores have a mean of
Description
HAWIVA-III Hannover-Wechsler-Intelligenztest fu¨r das Vorschulalter-III, K-ABC Kaufman-Assessment Battery for Children
10 and a standard deviation of 3. By focusing only on the pre-/post-test change in the formerly described sensitive subtests in intervention and control groups, there was no need for a full-scale IQ measurement in our study sample. Therefore, only three scales (including seven subscales) were selected (Table 1). Nine cognitive function testers (all female students of psychology aged 20–43 years, three of them speaking
123
Eur J Nutr
German as their second language) participated in one common structured training session. This was followed by an individual training combined with a supervised testing session in a kindergarten setting that was not further involved in the study. All cognitive tests took place in rooms within the kindergartens which were familiar to the children. Each test session started by ensuring that the room offered a quiet and minimally disturbing environment with a comfortable sitting position for the child. Tests were scheduled according to room availability and took place from early morning to late afternoon. For each test, the child was introduced to the tester by its kindergarten teacher and was subsequently given approximately 15 min to play and become familiar with the tester. When the child was ready to start, the subtests were performed in the order described in Table 1 with a short break after the three verbal subtests. The complete test session took approximately 45 min, depending on the necessity of additional breaks and the child’s capabilities. Any irregularities in the testing situation or changes in the order of subtests were noted in the test protocols and considered in the evaluation process. All children were tested by different testers in preand post-tests. The children’s answers were recorded during the test session in protocols which were completed at a later date by calculating the raw scores as well as the equivalent standardized and age-adjusted scores according to the test manuals. These scores were then validated by a different tester in a second evaluation round and confirmed by a third party in case of discrepancies. Test data were transferred to the database by double entry. Biochemical markers Urine was collected at home and aliquots were sent in Urine MonovettesÒ (10 mL, Luer, Sarstedt, Germany) by mail to the Dr. von Hauner Children’s Hospital and centrifuged after arrival (4,000 g, 10 min, 4 °C). From the supernatant, creatinine was analysed by kinetic Jaffe´ reaction [24]. Aliquots were stored at -80 °C until analysis of MMA and folate catabolites. For the analysis of ap-ABG and p-ABG 80 lL, urine were mixed with 320 lL acetonitrile/methanol (50/50) containing [13C2D3]ap-ABG, [13C2]p-ABG as internal standards (50 nmol/L), shaken for 10 min and cooled for 1 h for protein precipitation before centrifugation. Eighty microlitre supernatant was transferred, dried and re-dissolved in 100 lL 3 N butanolic hydrogen chloride for derivatization (10 min, 60 °C). Solutions were dried again, then re-dissolved in 200 lL methanol/water (50/50) with 0.1 % formic acid and separated by reversedphase high-performance liquid chromatography (HPLC system, Agilent Technologies, Germany). Chromatography was performed isocratically on an Agilent Zorbax SB C18
123
column (2.1 9 50 mm, particle size 1.8 lm) using 0.1 % formic acid in water/methanol (39/61) as mobile phase. Di-butyl-esters of folate catabolites were quantified by tandem mass spectrometry (API 4000 QTrapÒ, AB Sciex, Germany) using mass transitions 379.2 ? 120.1, 381.2 ? 120.1, 421.2 ? 162.1, 426.2 ? 165.1, for derivatives of p-ABG, [13C2]p-ABG, ap-ABG and [13C2D3]ap-ABG, respectively. The LC–MS/MS method for measuring of p-ABG/ap-ABG in urine together with the specification for the accuracy and reproducibility of this method has been described by Niesser et al. [25]. Concentrations of folic acid catabolites in urine are expressed as nanomole per millimole creatinine. The sum of p-ABG and ap-ABG relative to creatinine is given as total p-ABG. In the screening part of the study, the urine samples were measured in two separate analytical batches. Within-batch z scores of total p-ABG were merged to construct the overall subject index used as the screening parameter. Urinary MMA was analysed by isotope-dilution gas chromatography–mass spectrometry as described by Matchar et al. [26] at the Institute for Clinical Chemistry, University of Munich Medical Centre. Deuterated MMA was added to 2 mL of urine, the sample was acidified, saturated with NaCl, and extracted with ethyl acetate. The dried extract was reacted with N-methyl-N(t-butyldimethylsilyl) trifluoracetamide and after gas chromatography silylated MMA was detected at masses 289 (MMA) and 292 ([D3]-MMA), respectively. Urinary MMA concentrations are expressed as milligram per gram creatinine. Blood sampling by finger prick was performed at the Dr. von Hauner Children’s Hospital. A 100 lL aliquot was collected into an EDTA-coated vial (Microvette 200, Kalium-EDTA, Sarstedt, Germany) for blood count, while the remaining volume (ca. 100 lL) was collected in a Liheparin tube (Microvette 300, Li-Heparin, Sarstedt, Germany). For folate measurement, a 20-lL aliquot of the heparinized blood was mixed with 180 lL 1 % ascorbic acid solution, incubated at room temperature in the dark for 30 min and stored at -80 °C until analysis. The remaining blood was centrifuged (1,500g, 5 min, 4 °C) and plasma was stored at -80 °C for later homocysteine analysis. Full blood folate was determined with a microbiological assay at TNO (Delft, The Netherlands). Samples (100 lL diluted full blood) were incubated for 42 h in well plates containing growth medium, Tween 80, chloramphenicol and ascorbic acid with Lactobacillus rhamnosus [27]. After incubation, bacterial growth was measured by reading the absorbance at 570 nm. Two aliquots were analysed from each sample and in a further aliquot of the diluted full blood haemoglobin concentrations were measured by reading the absorbance at 575 nm after combining 20 lL sample with 230 lL alkaline haematin detergent and 30 min of incubation [28]. Folate concentrations are given
Eur J Nutr
as picomole folate per gram of haemoglobin. Homocysteine analysis was performed as described by Hellmuth et al. [29]. Briefly, 20 lL of plasma, 20 lL of aqueous internal standard solution (5 lg/mL D8-homocysteine) and 20 lL of dithiothreitol (77 mg/mL) were mixed by shaking for 5 min and incubated at room temperature for further 15 min. Proteins were precipitated by adding 100 lL acetonitrile (0.1 % formic acid), mixing, incubating at 4 °C for 30 min and centrifugation for 10 min at 2,200g at room temperature. Clear supernatants were injected into the liquid chromatography–tandem mass spectrometry system and separated by aqueous normal-phase chromatography on a Cogent Diamond Hydride column (MicroSolv Technology Corporation, USA) using a flow rate of 0.5 mL/min and a binary gradient starting with 10 % formic acid (0.1 %) and 90 % acetonitrile (0.1 % formic acid), which increased linearly to 70 % formic acid within 4 min. Homocysteine was quantified by monitoring mass transitions 136.0 ? 90.0 (homocysteine) and 140.1 ? 94.0 (D4-homocysteine). Quality controls were included in all blood analyses. The acceptance criteria for these controls included a recovery of 85–115 %, these criteria were fulfilled in all analyses. Study size and power calculation The power calculation was based on a study reporting a significant effect on the Rey Auditory-Verbal Learning Test (RAVLT) in young women after 35 days supplementation with an effect size of 3.5 words and a standard deviation of about 9 words [8]. For our study, one hundred subjects per group were identified to be sufficient to detect this difference with a power of 80 % and a level of significance of 0.05 between the groups. Assuming a 20 % dropout rate, 250 subjects were planned to be included. However, during the preparation phase of the study it was decided not to include the RAVLT, as in a pilot study this test proved to be too difficult for children aged 4–6, as indicated by the inability of the children to understand the test instructions, resulting in test abortions due to a severe loss of test motivation. The necessary subgroup size for blood samples was calculated for an estimated standard deviation in homocysteine concentration in children aged 2–5 years of 1.1 lmol/L [30]. With a minimal estimated effect size of 0.5 lmol/L, a level of significance of 0.05, a power of 80 % and a dropout rate of 30 % approximately sixty-five participants per group were required. Statistical analysis Data management and statistical analyses were carried out with the software package SAS 9.2 for Windows (SAS
Institute Inc., Cary/NC, USA) and the software package R version 2.11 for Windows [31]. The results of the cognitive tests are presented as standardized and age-adjusted scores. Medians with interquartile ranges (IQR: 25th and 75th percentiles) are used for all outcome variables due to non-normal distribution. Description of baseline characteristics was performed on the full analysis set containing all allocated participants, whereas the safety analysis included all randomized children. Analyses of primary and secondary endpoints were initially performed according to the intention-totreat principle. In addition, per-protocol analyses were conducted to control the effect in the compliant subgroups. Significance was evaluated by the P value for the group effect. Primary analyses of the cognitive outcomes in scales and subscales were performed by analysis of covariance using the respective baseline test score as covariate, and in addition the baseline ap-ABG measure since the screening was based on this urinary marker. The secondary analyses of the biochemical markers whole blood folate, plasma homocysteine and urinary MMA (log transformed) were computed via ANCOVA adjusted for the number of intervention sachets consumed, age and respective baseline measure. For the outcome urinary ap-ABG a quantile (median) regression was chosen as the residuals were not normally distributed.
Results One thousand and two children with an average age of 5.2 years, whose parents had provided written informed consent, agreed to participate in the screening study and provided urine sample. Total urinary p-ABG ranged from 9.2 to 150.4 nmol/mmol creatinine in the screened sample (Fig. 1) and subjects ranking in the lower part of folate catabolite concentrations, resulting in values below 34 nmol/mmol creatinine, were invited to participate in the intervention trial. From the eligible population established by urinary screening, 250 children from seventy-seven kindergartens were enrolled into the main trial and randomized, and 243 started the intervention (Fig. 2). Post-intervention assessments of cognitive performance were completed by 115 children in the intervention group and 122 children in the control group. One child in the control group was excluded from primary analyses due to a critical delay between end of intervention and cognitive testing. Blood samples were obtained from sixty children from both the intervention and control groups pre- and post-intervention. The participating children had a mean age of 5.5 years when starting the intervention. There were no statistically relevant differences at baseline between the two groups regarding socio-economic
123
Eur J Nutr Fig. 1 Subject selection based on urinary folate catabolite measurements: concentrations of total p-ABG (nmol/L) in spot urine relative to creatinine concentration (mmol/L). Data points of subjects finally recruited for the intervention trial are marked with larger symbols. p-ABG p-aminobenzoyglutamate
Fig. 2 Randomization, allocation, follow-up and data analysis of study participants
characteristics, test scores, urinary and blood folate markers, and daily DFE intake (Table 2). Compliance with the intervention as assessed by 237 completed diaries was good, with 84 % of all study participants reporting to have consumed at least 90 % of the assigned sachets over at least 81 days (92 % of participants in the subgroup with blood samples). There were no differences in treatment compliance or loss to follow-up between active intervention and control groups. The cognitive testing variables
123
(tester’s age, tester’s first language and test time) were similar between the groups and were not considered as covariates in the main analyses. One or more adverse events (AE) were reported for ninety-one (36.4 %) children, including mostly diarrhoea, abdominal pain or vomiting (68 % of all AE in the intervention group, 71 % of all AE in the control group). All adverse events were rated by the study physician as mild to moderate. There were no group differences for total
Eur J Nutr
subsequent evaluations, neither multivariate analysis of covariance (MANCOVA) nor per-protocol analysis resulted in any relevant differences between the groups. Compared to control, at the end of the 3-month intervention median whole blood folate was about 50 % higher (P for difference, P \ 0.0001) and ap-ABG was 4 nmol/ mmol higher in the B vitamin and folate-receiving group (P \ 0.0001), respectively (Table 4). Homocysteine decreased by 1.1 lmol/L compared to baseline, while no
frequency of adverse events, frequency stratified for event type or in intensity; no relation was detected between active intervention and any adverse event. Analyses of covariance adjusted for the respective baseline measurements (test scores and ap-ABG) showed no statistically significant main effects of group assignment, neither on the cognitive scales nor on any of the seven cognitive subscales (Table 3). The mean group differences in cognitive scores were below one score point. In Table 2 Baseline characteristics of children by treatment group Intervention
Control
n (%)
Median (25 %, 75 %)
n (%)
Median (25 %, 75 %)
Male sex
64 (54.2)
–
73 (58.4)
–
Child with siblings
92 (78.0)
–
98 (78.4)
–
Low
12 (10.2)
–
17 (13.6)
–
Medium
48 (40.7)
–
47 (37.6)
–
High
58 (49.1)
–
61 (48.8)
–
Mother employed Age (months)
80 (67.8) 118
– 66.5 (59, 72)
80 (64.0) 125
– 66 (60, 72)
Dietary folate equivalent (lg/day)b
118
144 (109, 194)
125
153 (115, 214)
Maternal educationa
a Maternal education: low = no finished professional education, medium = at least a completed apprenticeship, high = at least a polytechnic degree b
1 Dietary folate equivalent = 1 lg food folate = 0.6 lg folic acid (from fortified food or supplement) consumed with food = 0.5 lg synthetic (supplemental) folic acid taken on an empty stomach
Table 3 Number of children, test scores for domains and subscales at baseline and after 3 months Intervention
P valuea
Control
Baseline
3 months
Baseline
3 months
No. of childrenb
118
115
125
122
Verbal scale (HAWIVA-III)
35 (31, 38)c
37 (32, 40)
35 (30, 39)
37 (32, 41)
0.32
Information
12 (10, 14)
13 (10, 15)
12 (10, 14)
12 (10, 15)
0.93
Vocabulary
12 (11, 13)
12 (11, 14)
11 (10, 13)
12 (11, 14)
0.23
Word reasoning
11 (10, 12)
12 (10, 13)
10 (9, 12)
11 (10, 13)
0.55
23 (20, 26)
24 (21, 27)
23 (20, 27)
24 (21, 27)
0.25
12 (10, 13)
12 (11, 14)
12 (10, 14)
12 (11, 14)
0.37
Processing speed quotient (HAWIVA-III) Symbol search Coding Sequential processing scale (K-ABC)d Word order Number recall
11 (9, 12)
11 (10, 13)
11 (10, 13)
12 (10, 13)
0.39
–
–
–
–
–
11 (9, 13) 10 (8, 12)
11 (9, 13) 11 (9, 13)
12 (10, 13) 10 (8, 12)
11 (10, 13) 11 (9, 13)
0.91 0.99
HAWIVA-III Hannover-Wechsler-Intelligenztest fu¨r das Vorschulalter-III, K-ABC Kaufman-Assessment Battery for Children, ap-ABG acetyl-paminobenzoyglutamate a
P value for group (intervention versus control) derived from ANCOVA adjusted for respective baseline value and baseline ap-ABG
b
Numbers vary slightly between measures
c
Median; 25th and 75th percentile in parentheses (all such values)
d
No composite score available for K-ABC
123
Eur J Nutr Table 4 Number of samples, urine measures and blood measures at baseline and after 3 months Intervention
No. of urine samplesa
Control
P value
Baseline
3 months
Baseline
3 months
118
115
125
122
Urinary ap-ABG (nmol/mmol Crea)
15.84 (12.61, 19.15)b
18.30 (15.26, 20.64)
15.05 (12.75, 21.58)
14.01 (11.76, 18.29) \0.0001c
Urinary MMA (mg/g Crea)
1.43 (1.01, 1.91)
1.19 (0.76, 1.60)
1.33 (0.96, 1.69)
1.17 (0.90,1.72)
No. of blood samples
a
0.1d
60
60
60
60
Full blood folate (pmol/g Hb)
1,413 (1,138, 1,676)
2,234 (1,877, 2,553)
1,367 (1,149, 1,715)
1,440 (1,207, 1,624) \0.0001d
Plasma homocysteine (lmol/L)
7.36 (6.31, 8.14)
6.23 (5.34, 6.98)
6.69 (5.62, 7.75)
6.67 (5.95, 8.42)
a
Numbers vary slightly between measures
b
Median; 25th and 75th percentile in parentheses (all such values)
\0.0001d
c
P value for group (intervention vs. control) derived from median regression adjusted for baseline value, age and number of intervention sachets consumed d
P value for group (intervention vs. control) derived from ANCOVA adjusted for respective baseline value, age and number of intervention sachets consumed
Fig. 3 Test score ‘HAWIVAIII-Coding’ in relation to whole blood folate in intervention and placebo group: changes in the standardized and age-adjusted scores for test scale ‘HAWIVAIII-Coding’ between post- and pre-tests against changes in whole blood folate (pmol/g Hb) between post- and preintervention measures. HAWIVA-III HannoverWechsler-Intelligenztest fu¨r das Vorschulalter-III
change was seen in the control group (P for difference P \ 0.0001). Urinary MMA in the intervention group significantly decreased only in the per-protocol analysis by 0.13 mg/g creatinine (0.003, 0.260) on log scale (data not shown). The pre-/post-intervention changes in folate markers were not associated with changes in cognitive performance, as Fig. 3 shows exemplarily using whole blood folate and HAWIVA-III-Coding test, i.e. also for responders with positive changes in folate markers, we found no significant effects regarding cognitive scores. In all analyses, no gender-specific impacts were found when adding gender as a covariate.
Discussion This randomized controlled trial did not find any indication for an effect on cognitive function of vitamin B
123
supplementation for 3 months in healthy pre-school children, despite a marked change in folic acid, homocysteine, p-ABG and in the per-protocol analysis in vitamin B12 status. Prior inclusion, children from a relatively affluent population were classified to have low folate intake on the basis of the folate status marker total p-ABG. In a similar, 5-week intervention study in young adults, some indications were reported on marginal effects on measures of memory performance [8]. However, measured on the totality of the tests performed, the marginal effects found in a limited number of cognitive assessments of memory seem to be of limited practical significance. Furthermore, no information on vitamin status was reported in the latter study. In a similar study, in young adult men, a 2-month multivitamin supplementation trial reported a correlation between baseline folate status and self-reported mood, but no indications for effects on cognitive outcomes [32]. A recent study in the UK investigated the effect of a multivitamin and mineral supplement on cognition and
Eur J Nutr
mood over a 3-month period [33]. In relation to the whole battery of tests applied, the supplementation group performed more accurately on two attention tasks, but worse on a picture error recognition task. However, similarly to previous studies reporting significant effects on cognitive performance [34] this latter study does also not enable conclusions to be drawn on any component of the nutrient mix. Recent longer-term studies in elderly subjects reported benefits of B vitamins on cognitive functions [35, 36]. The most likely mechanism by which folic acid and vitamin B12 supplementation may affect cognition in older adults, who often have a poor B vitamin status [37] appears to be homocysteine lowering with concomitant improvement of microvascular circulation [38]. If this mechanism would be alone responsible for the reported cognitive effects in the elderly population, this may be of little relevance in young children without significant vascular alterations. We assume that no memory effects arose from using the same scales twice because no scores in the total sample increased or decreased significantly from pre- to post-test. In the pre-test, the total sample scored slightly above average on all scales (mean scores for subscales reached from 10.79 to 12.54) and therefore, the chance to achieve even higher scores in the post-test might have been limited. An intervention period of 3 months was chosen based on the assumption that this should suffice to achieve adequate changes in status and tissue concentrations, and analysis from biochemical samples supports this assumption. In this study, we did not test the hypothesis that longterm B vitamin supplementation and sustained increase in B vitamin status would affect cognitive performance over long term. Bryan [8] had suggested that 1 month of vitamin B supplementation in young women tended to have a positive effect on memory performance, our aim was to specifically test the hypothesis that a short term increase in vitamin B status and concomitant decrease in homocysteine would have an impact on cognitive measures in kindergarten children. Our results clearly showed no tendency for cognitive effects after 3 months of intervention. Whole blood folate measurements from samples collected by finger prick are highly correlated with concentrations in venous blood samples [39], and this procedure is much less invasive for children than venipuncture. As the subjects who consented to blood sampling did not differ from the full-study population in any of the measures, the significant increase in whole blood folate relative to haemoglobin underscores the compliance of the subjects to the protocol and demonstrates bioavailability of the supplemented folic acid. The observed basal folate concentration of approximately 1,400 pmol per gram haemoglobin would appear to correspond to an erythrocyte folate concentration
of approximately 490 nmol/L [39]. This is similar to concentrations recently observed in British children aged 4–18 years not consuming folic acid-fortified cereals [40] and is only slightly higher than observed in children of this age group in the United States before the implementation of mandatory flour fortification with folic acid [41]. The relatively low basal red blood cell (RBC) folate indicates that our screening procedure successfully identified children with lower folate status. Because the children enrolled into our study were not considered folate deficient according to biochemical measures, the outcome might have been different in a folate deficient population. On the other hand, observational studies in populations covering the whole range of folate levels indicate that maternal folic acid supplement intake during pregnancy [42] and plasma folic acid in children [43] are related to behaviour and cognitive performance of the children, respectively. Thus, we considered it worthwhile to test the hypothesis that additional folic acid intake improves cognitive performance in children at the age of 4–5 years in a randomized trial. In addition, the German population is known for low folate intake and our approach was to investigate if cognition could be affected by folic acid in children with relatively low folate intake. In the intervention group, RBC folate increased to 776 nmol/L which is in the range of the 75th percentile of British children, who regularly consumed vitamin supplements [40]. The chosen dose of 220 lg folic acid per day in the current study seems therefore adequate to effectively increase folate status, while not causing excessively high values. In agreement with observational studies in children [30] and interventional studies in infants [44, 45] as well as interventional and observational studies in adults [46, 47], vitamin B supplementation in our study was associated with a significant decrease in plasma homocysteine. To our knowledge, this is the first study to report an effect of B vitamin supplementation on homocysteine in healthy preschool children. Urinary MMA was measured as an additional marker of vitamin B status. Similar to plasma homocysteine, age adequate values [40, 48] were observed before and after intervention, but in this case levels were not significantly decreased by intervention. As urinary MMA is a specific marker for vitamin B12 deficiency [49], this result may indicate that the vitamin B12 status of our subjects was adequate and only marginally improved with supplementation of 1.2 lg/day. For optimal functioning of one-carbon metabolism in addition to folate and vitamin B12 also the vitamins B2 and B6 are needed, whereupon their effect on homocysteine lowering has been shown to be synergistic [50]. All four vitamins were included in the supplement. Since folate and cobalamin status are the strongest determinants of
123
Eur J Nutr
homocysteine levels in children [13], we focused our evaluations on these two vitamins. Although for the screening phase total urinary p-ABG was used, ap-ABG seemed during the progress of the study to be the more reliable marker for folate status in an intervention study, as it is related to the long-term marker red cell folate, whereas p-ABG follows fast-changing serum folate [17]. Thus, ap-ABG was specifically applied in the intervention study to avoid confounding due to different time intervals between blood sampling and final supplement intake. To our knowledge, no studies exploring the effect of folic acid supplementation on urine catabolites in children have been published so far. Assuming an average creatinine excretion of 2.8 mmol/day for children between 4 and 6 years [51], we estimated a daily ap-ABG excretion of 36–60 nmol from the baseline excretion of 14.5 pmol apABG/mmol creatinine. This corresponds to approximately 50 % of the ap-ABG excretion reported for young women (82–110 nmol/day) [17]. Considering the lower weight and smaller body pool size in our subjects, this lower excretion seems plausible. In accordance with previous observations [16], ap-ABG/creatinine was responsive to increased folic acid intake, as shown by the significantly higher urine apABG excretion in the intervention group in comparison to the control group at study end, when whole blood folate concentrations were also significantly increased in the vitamin B group (Table 4). The dropout rate of children in this study was extremely small (2.5 %) and was not related to group allocation. Similarly, adverse events generally comprised common symptoms for this age group and were not related to treatment. The participating children lived mostly in above average conditions in terms of housing and parental income. The maternal education level was higher than the German average: 89 % of the mothers had an advanced school degree, which is 18 % higher than in all German women aged 20–45 years [52] and 49 % of mothers had an academic degree, 34 % above the national average [52]. Relatively favourable dietary habits could be expected in this population group, but according to the food frequency questionnaires, the median intake in the whole intervention population was only 149 lg DFE/d, indicating a significantly lower level of folate intake than the estimated average requirement for this age group (240 lg/day [14] ). In the total group included for the urinary screening procedure, the median DFE intake was 161 lg/day. These data are in line with the previously reported low folate and folic acid intake in German pre-school children [14, 15]. All cognitive testing was performed within the kindergarten setting in rooms familiar to the participating children. However, conditions of testing rooms varied between
123
kindergartens. It was not feasible to perform all pre- and post-tests at the same time of the day, but the statistical analysis confirmed no effects of testing time on cognitive performance. To achieve similar conditions for all children and to reduce potential bias, pre- and post-tests were conducted by different testers who were blinded for the group assignment as well as previous test results. Other setting specific confounders were minimized as far as possible by giving extensive instructions to all persons involved (kindergarten teachers, parents, children and testers) and suitable preparation of the testing rooms. Testers were employed for the complete testing phase. An elaborate training and instruction phase guaranteed both familiarity with the materials and the test situation and a standardized and smooth test execution. To ensure a general understanding of psychological assessment, all selected testers were students of psychology. In addition, the inclusion of ten pre-school children into a pre-test phase confirmed the children’s acceptance of the testing situation as well as applicability of test duration and order of subtests. Three of nine testers spoke German as their second language. The main criteria for selecting testers were a good command of the German language as well as a clear pronunciation. However, the results obtained showed that children tested by native speakers achieved slightly higher scores in the three verbal subscales than those tested by second-language speakers. The observed effect was nevertheless small, accounting for only 3 % of the explained variance in each of the three verbal subscales. In conclusion, despite significant changes in folate status and homocysteine, we found no indication of an effect of 3-month supplementation with B vitamins and folic acid on cognitive performance in healthy pre-school children with low but not insufficient folate intake. As folate status has been shown to correlate with verbal IQ in nourished Spanish 6-year olds [6], the short duration of this study may have contributed to the lack of observed effect. Acknowledgments We thank the participating families and kindergarten staff for their enthusiastic support of the project. We thank Sabine Eiselen (Dr von Hauner Children’s Hospital, University of Munich Medical Centre, Munich, Germany) for practical organization of the study, Martina Weber (Dr von Hauner Children’s Hospital, University of Munich Medical Centre, Munich, Germany) and Winfried Theis (Unilever R&D, Vlaardingen, Netherlands) for statistical advice, Christian Hellmuth (Dr von Hauner Children’s Hospital, University of Munich Medical Centre, Munich, Germany) for laboratory analyses, and Ingrid Pawellek (Dr von Hauner Children’s Hospital, University of Munich Medical Centre, Munich, Germany) for the evaluation of the food frequency questionnaires. MMA analyses (Institute for Clinical Chemistry, University of Munich Medical Centre, Munich, Germany) were supported by the Hans-FischerGesellschaft, Mu¨nchen. This work was supported financially in part by NUTRIMENTHE (Grant Agreement No. 212652). This project was also supported by Unilever R&D, Vlaardingen, The Netherlands.
Eur J Nutr Conflict of interest of interest.
The authors declare that they have no conflict
References 1. Bryan J, Osendarp S, Hughes D, Calvaresi E, Baghurst K, van Klinken JW (2004) Nutrients for cognitive development in school-aged children. Nutr Rev 62:295–306 2. Bailey LB, Gregory JF III (1999) Folate metabolism and requirements. J Nutr 129:779–782 3. Crowe SF, Ross CK (1997) Effect of folate deficiency and folate and B12 excess on memory functioning in young chicks. Pharmacol Biochem Behav 56:189–197 4. Troen AM, Shukitt-Hale B, Chao WH, Albuquerque B, Smith DE, Selhub J, Rosenberg J (2006) The cognitive impact of nutritional homocysteinemia in apolipoprotein-E deficient mice. J Alzheimers Dis 9:381–392 5. Lalonde R, Barraud H, Ravey J, Gueant JL, Bronowicki JP, Strazielle C (2008) Effects of a B-vitamin-deficient diet on exploratory activity, motor coordination, and spatial learning in young adult Balb/c mice. Brain Res 1188:122–131. doi:10.1016/j. brainres.2007.10.068 6. Arija V, Esparo´ G, Ferna´ndez-Ballart J, Murphy MM, Biarne´s E, Canals J (2006) Nutritional status and performance in test of verbal and non-verbal intelligence in 6 year old children. Intelligence 34:141–149. doi:10.1016/j.intell.2005.09.001 7. Strand TA, Taneja S, Ueland PM, Refsum H, Bahl R, Schneede J, Sommerfelt H, Bhandari N (2013) Cobalamin and folate status predicts mental development scores in North Indian children 12–18 mo of age. Am J Clin Nutr 97:310–317. doi:10.3945/ajcn. 111.032268 8. Bryan J, Calvaresi E, Hughes D (2002) Short-term folate, vitamin B-12 or vitamin B-6 supplementation slightly affects memory performance but not mood in women of various ages. J Nutr 132:1345–1356 9. Louwman MW, van Dusseldorp M, van de Vijver FJ, Thomas CM, Schneede J, Ueland PM, Refsum H, van Staveren WA (2000) Signs of impaired cognitive function in adolescents with marginal cobalamin status. Am J Clin Nutr 72:762–769 10. Gewa CA, Weiss RE, Bwibo NO, Whaley S, Sigman M, Murphy SP, Harrison G, Neumann CG (2009) Dietary micronutrients are associated with higher cognitive function gains among primary school children in rural Kenya. Br J Nutr 101:1378–1387. doi:10. 1017/S0007114508066804 11. Homocysteine Lowering Trialists’ Collaboration (1998) Lowering blood homocysteine with folic acid based supplements: metaanalysis of randomised trials. BMJ 316:894–898 12. Lonn E, Yusuf S, Arnold MJ, Sheridan P, Pogue J, Micks M, McQueen MJ, Probstfield J, Fodor G, Held C, Genest J Jr (2006) Homocysteine lowering with folic acid and B vitamins in vascular disease. N Engl J Med 354:1567–1577. doi:10.1056/ NEJMoa060900 13. Ueland PM, Monsen AL (2003) Hyperhomocysteinemia and B-vitamin deficiencies in infants and children. Clin Chem Lab Med 41:1418–1426. doi:10.1515/CCLM.2003.218 14. Sichert-Hellert W, Wenz G, Kersting M (2006) Vitamin intakes from supplements and fortified food in German children and adolescents: results from the DONALD study. J Nutr 136:1329–1333 15. Mensink GBE, Herseker H, Richter A, Stahl A, Vohmann C (2007) Erna¨hrungstudie als KIGGS-Modul (EsKiMo), German [KIGGS nutrition survey]. In: Robert Koch Institut and Universita¨t Paderborn, Berlin and Paderborn
16. Wolfe JM, Bailey LB, Herrlinger-Garcia K, Theriaque DW, Gregory JF III, Kauwell GP (2003) Folate catabolite excretion is responsive to changes in dietary folate intake in elderly women. Am J Clin Nutr 77:919–923 17. Kim HA, Choi JH, Lim HS (2007) Childbearing women of twenty and under are at greater risk than those of twenty-five and over for compromised folate status. Nutr Res Pract 1:254–259. doi:10.4162/nrp.2007.1.4.254 18. Moher D, Schulz KF, Altman D (2001) The CONSORT statement: revised recommendations for improving the quality of reports of parallel-group randomized trials. JAMA 285:1987–1991 19. Scientific Committee on Food & Scientific Panel on Dietetic Products Nutrition and Allergies (2006) Tolerable upper intake levels for vitamins and minerals. In: European Food Safety Authority 20. Ricken G, Fritz A, Schuck KD, Preuß U (eds) (2007) HAWIVAIII, Hannover-Wechsler-Intelligenztest fu¨r das Vorschulalter-III, German [HAWIVA-III, Hannover-Wechsler-Preschool Scale of Intelligence-III]. Huber, Go¨ttingen 21. Wechsler D (2002) Wechsler Preschool and Primary Scale of Intelligence-III (WPPSI-III). The Psychological Corporation, San Antonio, TX 22. Kaufman AS, Kaufman NL (1983) Kaufman assessment battery for children: K-ABC. American Guidance Service, Circle Pines, MN 23. Melchers P, Preuß U (2005) Kaufman assessment battery for children. Pearson Assessment, Frankfurt/Main 24. Bartels H, Bo¨hmer M, Heierli C (1972) Serum kreatininbestimmung ohne enteiweissen, German [Serum creatinine determination without protein precipitation]. Clin Chim Acta 37:193–197. doi:10.1016/0009-8981(72)90432-9 25. Niesser M, Harder U, Koletzko B, Peissner W (2013) Quantification of urinary folate catabolites using liquid chromatographytandem mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci 929:116–124. doi:10.1016/j.jchromb.2013.04. 008 26. Matchar DB, Feussner JR, Millington DS, Wilkinson RH Jr, Watson DJ, Gale D (1987) Isotope-dilution assay for urinary methylmalonic acid in the diagnosis of vitamin B12 deficiency. A prospective clinical evaluation. Ann Intern Med 106:707–710 27. O’Broin S, Kelleher B (1992) Microbiological assay on microtitre plates of folate in serum and red cells. J Clin Pathol 45:344–347 28. Heuck CC, Reinauer H, Wood WG (2008) The alkaline haematin detergent (AHD575) method for the determination of haemoglobin in blood–a candidate reference measurement procedure. Clin Lab 54:255–272 29. Hellmuth C, Koletzko B, Peissner W (2011) Aqueous normal phase chromatography improves quantification and qualification of homocysteine, cysteine and methionine by liquid chromatography-tandem mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci 879:83–89. doi:10.1016/j.jchromb.2010.11. 016 30. Huemer M, Vonblon K, Fodinger M, Krumpholz R, Hubmann M, Ulmer H, Simma B (2006) Total homocysteine, folate, and cobalamin, and their relation to genetic polymorphisms, lifestyle and body mass index in healthy children and adolescents. Pediatr Res 60:764–769. doi:10.1203/01.pdr.0000246099.39469.18 31. R Development Core Team (2010) R: A language and environment for statistical computing. In: R Foundation for Statistical Computing, Vienna, Austria 32. Heseker H, Kuebler W, Westenhoefer J, Pudel V (1990) Psychische Veraenderungen als Fruehzeichen einer suboptimalen Vitaminversorgung, German [Psychic changes as early symptoms
123
Eur J Nutr
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
of suboptimal vitamin supply]. 2. Ernaehrungs-Umschau 37: 87–94 Haskell CF, Scholey AB, Jackson PA, Elliott JM, Defeyter MA, Greer J, Robertson BC, Buchanan T, Tiplady B, Kennedy DO (2008) Cognitive and mood effects in healthy children during 12 weeks’ supplementation with multi-vitamin/minerals. Br J Nutr 100:1086–1096. doi:10.1017/S0007114508959213 Osendarp SJ, Baghurst KI, Bryan J, Calvaresi E, Hughes D, Hussaini M, Karyadi SJ, van Klinken BJ, van der Knaap HC, Lukito W, Mikarsa W, Transler C, Wilson C (2007) Effect of a 12-mo micronutrient intervention on learning and memory in well-nourished and marginally nourished school-aged children: 2 parallel, randomized, placebo-controlled studies in Australia and Indonesia. Am J Clin Nutr 86:1082–1093 Durga J, Verhoef P, Anteunis LJ, Schouten E, Kok FJ (2007) Effects of folic acid supplementation on hearing in older adults: a randomized, controlled trial. Ann Intern Med 146:1–9 Vogel T, Dali-Youcef N, Kaltenbach G, Andres E (2009) Homocysteine, vitamin B12, folate and cognitive functions: a systematic and critical review of the literature. Int J Clin Pract 63:1061–1067. doi:10.1111/j.1742-1241.2009.02026.x Fabian E, Bogner M, Kickinger A, Wagner KH, Elmadfa I (2011) Intake of medication and vitamin status in the elderly. Ann Nutr Metab 58:118–125. doi:10.1159/000327351 Mansoor MA, Hervig T, Stakkestad JA, Drablos PA, Apeland T, Wentzel-Larsen T, Bates CJ (2011) Serum folate is significantly correlated with plasma cysteine concentrations in healthy industry workers. Ann Nutr Metab 58:68–73. doi:10.1159/000325537 O’Broin SD, Kelleher BP, Davoren A, Gunter EW (1997) Fieldstudy screening of blood folate concentrations: specimen stability and finger-stick sampling. Am J Clin Nutr 66:1398–1405 Kerr MA, Livingstone B, Bates CJ, Bradbury I, Scott JM, Ward M, Pentieva K, Mansoor MA, McNulty H (2009) Folate, related B vitamins, and homocysteine in childhood and adolescence: potential implications for disease risk in later life. Pediatrics 123:627–635. doi:10.1542/peds.2008-1049 Pfeiffer CM, Johnson CL, Jain RB, Yetley EA, Picciano MF, Rader JI, Fisher KD, Mulinare J, Osterloh JD (2007) Trends in blood folate and vitamin B-12 concentrations in the United States, 1988 2004. Am J Clin Nutr 86:718–727 Roza SJ, van Batenburg-Eddes T, Steegers EA, Jaddoe VW, Mackenbach JP, Hofman A, Verhulst FC, Tiemeier H (2010) Maternal folic acid supplement use in early pregnancy and child
123
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
behavioural problems: The Generation R Study. Br J Nutr 103:445–452. doi:10.1017/S0007114509991954 Nguyen CT, Gracely EJ, Lee BK (2013) Serum folate but not vitamin B-12 concentrations are positively associated with cognitive test scores in children aged 6–16 years. J Nutr 143:500–504. doi:10.3945/jn.112.166165 Bjorke-Monsen AL, Torsvik I, Saetran H, Markestad T, Ueland PM (2008) Common metabolic profile in infants indicating impaired cobalamin status responds to cobalamin supplementation. Pediatrics 122:83–91. doi:10.1542/peds.2007-2716 Torsvik I, Ueland PM, Markestad T, Bjorke-Monsen AL (2013) Cobalamin supplementation improves motor development and regurgitations in infants: results from a randomized intervention study. Am J Clin Nutr 98:1233–1240. doi:10.3945/ajcn.113. 061549 Homocysteine Lowering Trialists’ Collaboration (2005) Dosedependent effects of folic acid on blood concentrations of homocysteine: a meta-analysis of the randomized trials. Am J Clin Nutr 82:806–812 Chew SC, Khor GL, Loh SP (2011) Association between dietary folate intake and blood status of folate and homocysteine in Malaysian adults. J Nutr Sci Vitaminol (Tokyo) 57:150–155 Erdogan E, Nelson GJ, Rockwood AL, Frank EL (2010) Evaluation of reference intervals for methylmalonic acid in plasma/ serum and urine. Clin Chim Acta 411:1827–1829. doi:10.1016/j. cca.2010.06.017 Bjorke Monsen AL, Ueland PM (2003) Homocysteine and methylmalonic acid in diagnosis and risk assessment from infancy to adolescence. Am J Clin Nutr 78:7–21 Stanger O, Herrmann W, Pietrzik K, Fowler B, Geisel J, Dierkes J, Weger M, e.V D-LH (2003) DACH-LIGA homocystein (german, austrian and swiss homocysteine society): consensus paper on the rational clinical use of homocysteine, folic acid and B-vitamins in cardiovascular and thrombotic diseases: guidelines and recommendations. Clin Chem Lab Med 41:1392–1403. doi:10.1515/CCLM.2003.214 Lentze MJ, Schaub J, Schulte FJ, Spranger J (2007) Pa¨diatrie: Grundlagen und Praxis, German [Pediatrics: Basics and practice]. In: Springer Medizin Verlag Heidelberg, Berlin, Heidelberg Statistisches Bundesamt (ed) (2011) Statistisches Jahrbuch 2011, German [statistical yearbook 2011]. Statistisches Bundesamt, Wiesbaden
ORIGINAL CONTRIBUTION
Three-month B vitamin supplementation in pre-school children affects folate status and homocysteine, but not cognitive performance Astrid Rauh-Pfeiffer • Uschi Handel • Hans Demmelmair • Wolfgang Peissner Mareile Niesser • Diego Moretti • Vanessa Martens • Sheila Wiseman • Judith Weichert • Moritz Heene • Markus Bu¨hner • Berthold Koletzko
•
Received: 16 July 2013 / Accepted: 18 December 2013 Ó Springer-Verlag Berlin Heidelberg 2014
Abstract Background Suboptimal vitamin B status might affect cognitive performance in early childhood. We tested the hypothesis that short-term supplementation with folic acid and selected B vitamins improves cognitive function in healthy children in a population with relatively low folate status. Methods We screened 1,002 kindergarten children for suboptimal folate status by assessing the total urinary paraaminobenzoylglutamate excretion. Two hundred and fifty low ranking subjects were recruited into a double blind, randomized, controlled trial to receive daily a sachet containing 220 lg folic acid, 1.1 mg vitamin B2, 0.73 mg B6, 1.2 lg B12 and 130 mg calcium, or calcium only for 3 months. Primary outcomes were changes in verbal IQ, short-term memory and processing speed between baseline and study end. Secondary outcomes were urinary markers
Berthold Koletzko is the recipient of a Freedom to Discover Award of the Bristol-Myers Squibb Foundation, New York, NY. A. Rauh-Pfeiffer U. Handel H. Demmelmair W. Peissner M. Niesser B. Koletzko (&) Dr von Hauner Children’s Hospital, University of Munich Medical Centre, Lindwurmstraße 4, 80337 Munich, Germany e-mail: [email protected] D. Moretti V. Martens S. Wiseman Unilever R&D, 3133 AT Vlaardingen, The Netherlands D. Moretti Laboratory of Human Nutrition, Institute of Food Science and Nutrition, ETH Zu¨rich, 8092 Zurich, Switzerland J. Weichert M. Heene M. Bu¨hner Department of Psychology, University of Munich, 80802 Munich, Germany
of folate and vitamin B12 status, acetyl-para-aminobenzoylglutamate and methylmalonic acid, respectively, and, in a subgroup of 120 participants, blood folate and plasma homocysteine. Results Pre- and post-intervention cognitive measurements were completed by 115 children in the intervention and 122 in the control group. Compared to control, median blood folate increased by about 50 % (P for difference, P \ 0.0001). Homocysteine decreased by 1.1 lmol/L compared to baseline, no change was seen in the control group (P for difference P \ 0.0001) and acetyl-para-aminobenzoylglutamate was 4 nmol/mmol higher compared to control at the end of the intervention (P \ 0.0001). We found no relevant differences between the groups for the cognitive measures. Conclusion Short-term improvement of folate and homocysteine status in healthy children does not appear to affect cognitive performance. Keywords B vitamin supplementation Cognition Pre-school children
Introduction Severe nutritional deficiencies in the vitamins B2, B6, B12 and folate (in the course of the paper also called vitamin B) cause neurological and cognitive abnormalities in humans, and indications exist that these vitamins may be important for cognitive performance when intake and status are suboptimal [1]. Folate and vitamins B12, B6 and B2 share a metabolic pathway that may affect the central nervous system by modulating one-carbon metabolism, methylation and neurotransmitter synthesis [1, 2]. Evidence from animal studies [3, 4] suggests an effect of folate/folic acid
123
Eur J Nutr
supplementation on long-term memory [3] and spatial learning [4], while mice with vitamin B12 and folate deficiency showed increased hyperactivity [5]. Limited human data suggest a role for folate, vitamin B6 and vitamin B12 in cognitive performance at a young age. Associative evidence from a cross-sectional study in 6-year-old children reported positive correlations between folate status and total and verbal intelligence quotient (IQ) [6] and folate status was significantly associated with cognitive performance in Indian infants [7]. One-month supplementation with folic acid and vitamin B6 was reported to be associated with recall performance and word recall in young adults between 20 and 30 years while vitamin B12 supplementation was associated with specific measures of short and long-term memory performance [8]. Similar associations were reported in a study of Dutch adolescents who had followed a vegan diet in childhood where a correlation was found between vitamin B12 status and measures of abstract reasoning, spatial ability and short-term memory at the age of 10–18 years [9]. In a feeding intervention study among primary school children in rural Kenya, the increased intake of vitamins B2 and B12 showed significantly higher gains in digit span-forward test scores [10]. Folate, vitamin B12 and vitamin B6 support the metabolic availability of methyl groups and thus facilitate the remethylation of homocysteine to methionine. Poor dietary folate and vitamin B12 intake leads to increased serum homocysteine concentrations in adults [11, 12] and may affect homocysteine levels and folate status in children. Folate status is associated with serum homocysteine concentrations in children above 2 years of age [13], however, to our knowledge, limited direct evidence exists on the effect of B vitamin supplementation on homocysteine in this age group. Suboptimal folate intake has been reported in the paediatric population in Germany [14, 15]. Urinary folate catabolites, acetyl-p-aminobenzoyglutamate (ap-ABG) and p-aminobenzoylglutamate (p-ABG), have been proposed as non-invasive markers of folate intake in humans, and particularly ap-ABG was reported to reflect long-term folate pools [16, 17]. Total p-ABG (sum of ap-ABG and p-ABG) was used to identify subjects with low folate status and to select a population with low folate intake. We aimed to test the hypothesis that short-term (3 months) dietary supplementation with folic acid and vitamins B12, B6 and B2 improves folate status, decreases homocysteine and affects measures of cognitive performance in healthy pre-school children in a population not exposed to folic acid-fortified foods. To our knowledge, no controlled intervention study on the effects of a combination of B vitamins and folic acid on cognitive and biochemical outcomes has been performed in pre-school children to date.
123
Subjects and methods Study design and study population This study aimed to enrol 250 apparently healthy children aged 4–6 years, attending kindergartens in the Munich, Germany area into the placebo or treatment groups of this randomized controlled trial (RCT), based on the power calculation reported below. Children who agreed to participate, and whose parents provided written informed consent, were asked to provide a spot early morning urine sample for measurement of the urinary folate catabolites ap- and p-ABG. Urine from 1002 children was screened in order to enrol those children with the lowest p-ABG excretion into the study. Children who took vitamin B/folic acid supplements or who were not fluent in the German language were excluded from the study. Families were asked to complete a questionnaire on socio-economic and demographical descriptors, and a food frequency questionnaire regarding the nutritional habits of the child focused on the intake of dietary folate equivalents (DFE). A battery of cognitive tests was administered to all participating subjects before and after the intervention. The primary outcome measures were changes in short-term memory, verbal IQ and mental speed between baseline and study end. Secondary outcome measures were urinary markers of folate and vitamin B12 status ap-ABG and methylmalonic acid (MMA), respectively, and, in the subgroup in which the subjects had agreed to blood sampling, whole blood folate and plasma homocysteine. This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving human subjects were approved by the Ethical Committee of the Medical Faculty of the University of Munich. The study was explained to the parents and to the children in oral and written form, using age-appropriate pictographic leaflets for children. Written informed consent was obtained from all subjects’ parents. The study followed the recommendations of the CONSORT guidelines [18]. The study was registered at clinicaltrials.gov under NCT00811291. Intervention Two hundred and fifty subjects ranking low with regard to urinary total p-ABG within the screening group were enrolled into the intervention trial which was conducted between March and August 2009. Subjects were randomly allocated to receive either the intervention or the control product by means of a computer-generated list of randomization codes. Participants, parents, kindergarten teachers and all members of the study team were blinded during the whole trial. The intervention product was
Eur J Nutr
provided as a flavourless powder in sachets containing folic acid (220 lg), riboflavin (1.1 mg), pyridoxine (0.73 mg), cobalamin (1.2 lg) and calcium lactate pentahydrate (130 mg). The control matched the intervention product in taste and appearance but contained only 130 mg calcium. The additional intake of B vitamins as well as the estimated overall vitamin B intake of the children participating in this trial was below the upper limits set by the European Food Safety Authority (EFSA) for children aged 4–6 years [19]. The sachets were delivered to the subjects at the time of their first study appointment. Parents were instructed to mix the contents of one sachet per day into their child’s breakfast or other meal over a period of 90 days at home. Children’s compliance was supported by small gifts and incentives. Families were advised not to provide any other vitamin supplements or multivitamin juices to the child and not to change their usual dietary habits during the intervention period. Compliance with provision of test products was encouraged and monitored throughout the trial by repeated phone calls and letters to the families, and supplement usage was entered into diaries supposed to be completed daily by the families. Parents were asked to collect two early morning urine samples from their child, one before the start of the intervention period and one at the end of the trial. In a subgroup of children who were willing to undergo blood sampling and whose parents approved, blood samples were collected by finger prick before and after the intervention. Blood samples were taken by experienced staff in a child-friendly atmosphere with the use of a lancing device (FineTouch, Terumo, Eschborn, Germany).
Table 1 Overview cognitive function measures Test
Measured abilities
Verbal scale (HAWIVA-III) Information
Long-term memory and reproduction
The child answers questions of general knowledge, e.g. ‘What do people use for writing?’
Vocabulary
Long-term memory and language development
The child gives definitions for words that are read by the test assistant, e.g. ‘What does ‘polite’ mean?’
Word reasoning
Reasoning and abilities to abstract
The child gives words to more and more detailed descriptions read by the test assistant, e.g. ‘It’s a room that people use to cook’
Processing speed quotient (HAWIVA-III) Symbol search
Sustained attention
The child scans a group of symbols for a match of one of them with a previously shown target symbol
Coding
Concentration, visualmotor co-ordination and speed of processing
Requires the child to draw the missing counterpart of nine pairs of symbols and geometric shapes from a given key
Sequential processing scale (K-ABC) Number recall
Short-term memory
The child repeats a string of numbers in the same order as read aloud by the test assistant before. Strings range from two to nine digits
Word order
Short-term memory
The child has to remember names of common objects read aloud by the test assistant. Then a series of pictures of these objects are shown to the child who has to touch them in the same order they were read out in
Cognitive tests Several subtests of two well-known intelligence test systems for pre-school children were used to assess cognitive performance. The ‘Hannover-Wechsler-Intelligenztest fu¨r das Vorschulalter-III’ (HAWIVA-III) [20], the German version of the Wechsler Pre-school and Primary Scale of Intelligence-III (WPPSI-III) [21], which is especially designed for children between the ages of 2 years 6 months and 6 years 11 months. A full-scale IQ is provided as well as several subtest scores in verbal and performance cognitive domains. In addition, subscales of the KaufmanAssessment Battery for Children (K-ABC) [22] in the German version [23] were chosen, as an instrument for assessing the cognitive development in children aged 2.5–12.5 years. It comprises four global test scores that include sequential processing scales, simultaneous processing scales, achievement scales and mental processing composite. For both instruments, it applies that quotient and composite scores have a mean of 100 and a standard deviation of 15, while subtest scaled scores have a mean of
Description
HAWIVA-III Hannover-Wechsler-Intelligenztest fu¨r das Vorschulalter-III, K-ABC Kaufman-Assessment Battery for Children
10 and a standard deviation of 3. By focusing only on the pre-/post-test change in the formerly described sensitive subtests in intervention and control groups, there was no need for a full-scale IQ measurement in our study sample. Therefore, only three scales (including seven subscales) were selected (Table 1). Nine cognitive function testers (all female students of psychology aged 20–43 years, three of them speaking
123
Eur J Nutr
German as their second language) participated in one common structured training session. This was followed by an individual training combined with a supervised testing session in a kindergarten setting that was not further involved in the study. All cognitive tests took place in rooms within the kindergartens which were familiar to the children. Each test session started by ensuring that the room offered a quiet and minimally disturbing environment with a comfortable sitting position for the child. Tests were scheduled according to room availability and took place from early morning to late afternoon. For each test, the child was introduced to the tester by its kindergarten teacher and was subsequently given approximately 15 min to play and become familiar with the tester. When the child was ready to start, the subtests were performed in the order described in Table 1 with a short break after the three verbal subtests. The complete test session took approximately 45 min, depending on the necessity of additional breaks and the child’s capabilities. Any irregularities in the testing situation or changes in the order of subtests were noted in the test protocols and considered in the evaluation process. All children were tested by different testers in preand post-tests. The children’s answers were recorded during the test session in protocols which were completed at a later date by calculating the raw scores as well as the equivalent standardized and age-adjusted scores according to the test manuals. These scores were then validated by a different tester in a second evaluation round and confirmed by a third party in case of discrepancies. Test data were transferred to the database by double entry. Biochemical markers Urine was collected at home and aliquots were sent in Urine MonovettesÒ (10 mL, Luer, Sarstedt, Germany) by mail to the Dr. von Hauner Children’s Hospital and centrifuged after arrival (4,000 g, 10 min, 4 °C). From the supernatant, creatinine was analysed by kinetic Jaffe´ reaction [24]. Aliquots were stored at -80 °C until analysis of MMA and folate catabolites. For the analysis of ap-ABG and p-ABG 80 lL, urine were mixed with 320 lL acetonitrile/methanol (50/50) containing [13C2D3]ap-ABG, [13C2]p-ABG as internal standards (50 nmol/L), shaken for 10 min and cooled for 1 h for protein precipitation before centrifugation. Eighty microlitre supernatant was transferred, dried and re-dissolved in 100 lL 3 N butanolic hydrogen chloride for derivatization (10 min, 60 °C). Solutions were dried again, then re-dissolved in 200 lL methanol/water (50/50) with 0.1 % formic acid and separated by reversedphase high-performance liquid chromatography (HPLC system, Agilent Technologies, Germany). Chromatography was performed isocratically on an Agilent Zorbax SB C18
123
column (2.1 9 50 mm, particle size 1.8 lm) using 0.1 % formic acid in water/methanol (39/61) as mobile phase. Di-butyl-esters of folate catabolites were quantified by tandem mass spectrometry (API 4000 QTrapÒ, AB Sciex, Germany) using mass transitions 379.2 ? 120.1, 381.2 ? 120.1, 421.2 ? 162.1, 426.2 ? 165.1, for derivatives of p-ABG, [13C2]p-ABG, ap-ABG and [13C2D3]ap-ABG, respectively. The LC–MS/MS method for measuring of p-ABG/ap-ABG in urine together with the specification for the accuracy and reproducibility of this method has been described by Niesser et al. [25]. Concentrations of folic acid catabolites in urine are expressed as nanomole per millimole creatinine. The sum of p-ABG and ap-ABG relative to creatinine is given as total p-ABG. In the screening part of the study, the urine samples were measured in two separate analytical batches. Within-batch z scores of total p-ABG were merged to construct the overall subject index used as the screening parameter. Urinary MMA was analysed by isotope-dilution gas chromatography–mass spectrometry as described by Matchar et al. [26] at the Institute for Clinical Chemistry, University of Munich Medical Centre. Deuterated MMA was added to 2 mL of urine, the sample was acidified, saturated with NaCl, and extracted with ethyl acetate. The dried extract was reacted with N-methyl-N(t-butyldimethylsilyl) trifluoracetamide and after gas chromatography silylated MMA was detected at masses 289 (MMA) and 292 ([D3]-MMA), respectively. Urinary MMA concentrations are expressed as milligram per gram creatinine. Blood sampling by finger prick was performed at the Dr. von Hauner Children’s Hospital. A 100 lL aliquot was collected into an EDTA-coated vial (Microvette 200, Kalium-EDTA, Sarstedt, Germany) for blood count, while the remaining volume (ca. 100 lL) was collected in a Liheparin tube (Microvette 300, Li-Heparin, Sarstedt, Germany). For folate measurement, a 20-lL aliquot of the heparinized blood was mixed with 180 lL 1 % ascorbic acid solution, incubated at room temperature in the dark for 30 min and stored at -80 °C until analysis. The remaining blood was centrifuged (1,500g, 5 min, 4 °C) and plasma was stored at -80 °C for later homocysteine analysis. Full blood folate was determined with a microbiological assay at TNO (Delft, The Netherlands). Samples (100 lL diluted full blood) were incubated for 42 h in well plates containing growth medium, Tween 80, chloramphenicol and ascorbic acid with Lactobacillus rhamnosus [27]. After incubation, bacterial growth was measured by reading the absorbance at 570 nm. Two aliquots were analysed from each sample and in a further aliquot of the diluted full blood haemoglobin concentrations were measured by reading the absorbance at 575 nm after combining 20 lL sample with 230 lL alkaline haematin detergent and 30 min of incubation [28]. Folate concentrations are given
Eur J Nutr
as picomole folate per gram of haemoglobin. Homocysteine analysis was performed as described by Hellmuth et al. [29]. Briefly, 20 lL of plasma, 20 lL of aqueous internal standard solution (5 lg/mL D8-homocysteine) and 20 lL of dithiothreitol (77 mg/mL) were mixed by shaking for 5 min and incubated at room temperature for further 15 min. Proteins were precipitated by adding 100 lL acetonitrile (0.1 % formic acid), mixing, incubating at 4 °C for 30 min and centrifugation for 10 min at 2,200g at room temperature. Clear supernatants were injected into the liquid chromatography–tandem mass spectrometry system and separated by aqueous normal-phase chromatography on a Cogent Diamond Hydride column (MicroSolv Technology Corporation, USA) using a flow rate of 0.5 mL/min and a binary gradient starting with 10 % formic acid (0.1 %) and 90 % acetonitrile (0.1 % formic acid), which increased linearly to 70 % formic acid within 4 min. Homocysteine was quantified by monitoring mass transitions 136.0 ? 90.0 (homocysteine) and 140.1 ? 94.0 (D4-homocysteine). Quality controls were included in all blood analyses. The acceptance criteria for these controls included a recovery of 85–115 %, these criteria were fulfilled in all analyses. Study size and power calculation The power calculation was based on a study reporting a significant effect on the Rey Auditory-Verbal Learning Test (RAVLT) in young women after 35 days supplementation with an effect size of 3.5 words and a standard deviation of about 9 words [8]. For our study, one hundred subjects per group were identified to be sufficient to detect this difference with a power of 80 % and a level of significance of 0.05 between the groups. Assuming a 20 % dropout rate, 250 subjects were planned to be included. However, during the preparation phase of the study it was decided not to include the RAVLT, as in a pilot study this test proved to be too difficult for children aged 4–6, as indicated by the inability of the children to understand the test instructions, resulting in test abortions due to a severe loss of test motivation. The necessary subgroup size for blood samples was calculated for an estimated standard deviation in homocysteine concentration in children aged 2–5 years of 1.1 lmol/L [30]. With a minimal estimated effect size of 0.5 lmol/L, a level of significance of 0.05, a power of 80 % and a dropout rate of 30 % approximately sixty-five participants per group were required. Statistical analysis Data management and statistical analyses were carried out with the software package SAS 9.2 for Windows (SAS
Institute Inc., Cary/NC, USA) and the software package R version 2.11 for Windows [31]. The results of the cognitive tests are presented as standardized and age-adjusted scores. Medians with interquartile ranges (IQR: 25th and 75th percentiles) are used for all outcome variables due to non-normal distribution. Description of baseline characteristics was performed on the full analysis set containing all allocated participants, whereas the safety analysis included all randomized children. Analyses of primary and secondary endpoints were initially performed according to the intention-totreat principle. In addition, per-protocol analyses were conducted to control the effect in the compliant subgroups. Significance was evaluated by the P value for the group effect. Primary analyses of the cognitive outcomes in scales and subscales were performed by analysis of covariance using the respective baseline test score as covariate, and in addition the baseline ap-ABG measure since the screening was based on this urinary marker. The secondary analyses of the biochemical markers whole blood folate, plasma homocysteine and urinary MMA (log transformed) were computed via ANCOVA adjusted for the number of intervention sachets consumed, age and respective baseline measure. For the outcome urinary ap-ABG a quantile (median) regression was chosen as the residuals were not normally distributed.
Results One thousand and two children with an average age of 5.2 years, whose parents had provided written informed consent, agreed to participate in the screening study and provided urine sample. Total urinary p-ABG ranged from 9.2 to 150.4 nmol/mmol creatinine in the screened sample (Fig. 1) and subjects ranking in the lower part of folate catabolite concentrations, resulting in values below 34 nmol/mmol creatinine, were invited to participate in the intervention trial. From the eligible population established by urinary screening, 250 children from seventy-seven kindergartens were enrolled into the main trial and randomized, and 243 started the intervention (Fig. 2). Post-intervention assessments of cognitive performance were completed by 115 children in the intervention group and 122 children in the control group. One child in the control group was excluded from primary analyses due to a critical delay between end of intervention and cognitive testing. Blood samples were obtained from sixty children from both the intervention and control groups pre- and post-intervention. The participating children had a mean age of 5.5 years when starting the intervention. There were no statistically relevant differences at baseline between the two groups regarding socio-economic
123
Eur J Nutr Fig. 1 Subject selection based on urinary folate catabolite measurements: concentrations of total p-ABG (nmol/L) in spot urine relative to creatinine concentration (mmol/L). Data points of subjects finally recruited for the intervention trial are marked with larger symbols. p-ABG p-aminobenzoyglutamate
Fig. 2 Randomization, allocation, follow-up and data analysis of study participants
characteristics, test scores, urinary and blood folate markers, and daily DFE intake (Table 2). Compliance with the intervention as assessed by 237 completed diaries was good, with 84 % of all study participants reporting to have consumed at least 90 % of the assigned sachets over at least 81 days (92 % of participants in the subgroup with blood samples). There were no differences in treatment compliance or loss to follow-up between active intervention and control groups. The cognitive testing variables
123
(tester’s age, tester’s first language and test time) were similar between the groups and were not considered as covariates in the main analyses. One or more adverse events (AE) were reported for ninety-one (36.4 %) children, including mostly diarrhoea, abdominal pain or vomiting (68 % of all AE in the intervention group, 71 % of all AE in the control group). All adverse events were rated by the study physician as mild to moderate. There were no group differences for total
Eur J Nutr
subsequent evaluations, neither multivariate analysis of covariance (MANCOVA) nor per-protocol analysis resulted in any relevant differences between the groups. Compared to control, at the end of the 3-month intervention median whole blood folate was about 50 % higher (P for difference, P \ 0.0001) and ap-ABG was 4 nmol/ mmol higher in the B vitamin and folate-receiving group (P \ 0.0001), respectively (Table 4). Homocysteine decreased by 1.1 lmol/L compared to baseline, while no
frequency of adverse events, frequency stratified for event type or in intensity; no relation was detected between active intervention and any adverse event. Analyses of covariance adjusted for the respective baseline measurements (test scores and ap-ABG) showed no statistically significant main effects of group assignment, neither on the cognitive scales nor on any of the seven cognitive subscales (Table 3). The mean group differences in cognitive scores were below one score point. In Table 2 Baseline characteristics of children by treatment group Intervention
Control
n (%)
Median (25 %, 75 %)
n (%)
Median (25 %, 75 %)
Male sex
64 (54.2)
–
73 (58.4)
–
Child with siblings
92 (78.0)
–
98 (78.4)
–
Low
12 (10.2)
–
17 (13.6)
–
Medium
48 (40.7)
–
47 (37.6)
–
High
58 (49.1)
–
61 (48.8)
–
Mother employed Age (months)
80 (67.8) 118
– 66.5 (59, 72)
80 (64.0) 125
– 66 (60, 72)
Dietary folate equivalent (lg/day)b
118
144 (109, 194)
125
153 (115, 214)
Maternal educationa
a Maternal education: low = no finished professional education, medium = at least a completed apprenticeship, high = at least a polytechnic degree b
1 Dietary folate equivalent = 1 lg food folate = 0.6 lg folic acid (from fortified food or supplement) consumed with food = 0.5 lg synthetic (supplemental) folic acid taken on an empty stomach
Table 3 Number of children, test scores for domains and subscales at baseline and after 3 months Intervention
P valuea
Control
Baseline
3 months
Baseline
3 months
No. of childrenb
118
115
125
122
Verbal scale (HAWIVA-III)
35 (31, 38)c
37 (32, 40)
35 (30, 39)
37 (32, 41)
0.32
Information
12 (10, 14)
13 (10, 15)
12 (10, 14)
12 (10, 15)
0.93
Vocabulary
12 (11, 13)
12 (11, 14)
11 (10, 13)
12 (11, 14)
0.23
Word reasoning
11 (10, 12)
12 (10, 13)
10 (9, 12)
11 (10, 13)
0.55
23 (20, 26)
24 (21, 27)
23 (20, 27)
24 (21, 27)
0.25
12 (10, 13)
12 (11, 14)
12 (10, 14)
12 (11, 14)
0.37
Processing speed quotient (HAWIVA-III) Symbol search Coding Sequential processing scale (K-ABC)d Word order Number recall
11 (9, 12)
11 (10, 13)
11 (10, 13)
12 (10, 13)
0.39
–
–
–
–
–
11 (9, 13) 10 (8, 12)
11 (9, 13) 11 (9, 13)
12 (10, 13) 10 (8, 12)
11 (10, 13) 11 (9, 13)
0.91 0.99
HAWIVA-III Hannover-Wechsler-Intelligenztest fu¨r das Vorschulalter-III, K-ABC Kaufman-Assessment Battery for Children, ap-ABG acetyl-paminobenzoyglutamate a
P value for group (intervention versus control) derived from ANCOVA adjusted for respective baseline value and baseline ap-ABG
b
Numbers vary slightly between measures
c
Median; 25th and 75th percentile in parentheses (all such values)
d
No composite score available for K-ABC
123
Eur J Nutr Table 4 Number of samples, urine measures and blood measures at baseline and after 3 months Intervention
No. of urine samplesa
Control
P value
Baseline
3 months
Baseline
3 months
118
115
125
122
Urinary ap-ABG (nmol/mmol Crea)
15.84 (12.61, 19.15)b
18.30 (15.26, 20.64)
15.05 (12.75, 21.58)
14.01 (11.76, 18.29) \0.0001c
Urinary MMA (mg/g Crea)
1.43 (1.01, 1.91)
1.19 (0.76, 1.60)
1.33 (0.96, 1.69)
1.17 (0.90,1.72)
No. of blood samples
a
0.1d
60
60
60
60
Full blood folate (pmol/g Hb)
1,413 (1,138, 1,676)
2,234 (1,877, 2,553)
1,367 (1,149, 1,715)
1,440 (1,207, 1,624) \0.0001d
Plasma homocysteine (lmol/L)
7.36 (6.31, 8.14)
6.23 (5.34, 6.98)
6.69 (5.62, 7.75)
6.67 (5.95, 8.42)
a
Numbers vary slightly between measures
b
Median; 25th and 75th percentile in parentheses (all such values)
\0.0001d
c
P value for group (intervention vs. control) derived from median regression adjusted for baseline value, age and number of intervention sachets consumed d
P value for group (intervention vs. control) derived from ANCOVA adjusted for respective baseline value, age and number of intervention sachets consumed
Fig. 3 Test score ‘HAWIVAIII-Coding’ in relation to whole blood folate in intervention and placebo group: changes in the standardized and age-adjusted scores for test scale ‘HAWIVAIII-Coding’ between post- and pre-tests against changes in whole blood folate (pmol/g Hb) between post- and preintervention measures. HAWIVA-III HannoverWechsler-Intelligenztest fu¨r das Vorschulalter-III
change was seen in the control group (P for difference P \ 0.0001). Urinary MMA in the intervention group significantly decreased only in the per-protocol analysis by 0.13 mg/g creatinine (0.003, 0.260) on log scale (data not shown). The pre-/post-intervention changes in folate markers were not associated with changes in cognitive performance, as Fig. 3 shows exemplarily using whole blood folate and HAWIVA-III-Coding test, i.e. also for responders with positive changes in folate markers, we found no significant effects regarding cognitive scores. In all analyses, no gender-specific impacts were found when adding gender as a covariate.
Discussion This randomized controlled trial did not find any indication for an effect on cognitive function of vitamin B
123
supplementation for 3 months in healthy pre-school children, despite a marked change in folic acid, homocysteine, p-ABG and in the per-protocol analysis in vitamin B12 status. Prior inclusion, children from a relatively affluent population were classified to have low folate intake on the basis of the folate status marker total p-ABG. In a similar, 5-week intervention study in young adults, some indications were reported on marginal effects on measures of memory performance [8]. However, measured on the totality of the tests performed, the marginal effects found in a limited number of cognitive assessments of memory seem to be of limited practical significance. Furthermore, no information on vitamin status was reported in the latter study. In a similar study, in young adult men, a 2-month multivitamin supplementation trial reported a correlation between baseline folate status and self-reported mood, but no indications for effects on cognitive outcomes [32]. A recent study in the UK investigated the effect of a multivitamin and mineral supplement on cognition and
Eur J Nutr
mood over a 3-month period [33]. In relation to the whole battery of tests applied, the supplementation group performed more accurately on two attention tasks, but worse on a picture error recognition task. However, similarly to previous studies reporting significant effects on cognitive performance [34] this latter study does also not enable conclusions to be drawn on any component of the nutrient mix. Recent longer-term studies in elderly subjects reported benefits of B vitamins on cognitive functions [35, 36]. The most likely mechanism by which folic acid and vitamin B12 supplementation may affect cognition in older adults, who often have a poor B vitamin status [37] appears to be homocysteine lowering with concomitant improvement of microvascular circulation [38]. If this mechanism would be alone responsible for the reported cognitive effects in the elderly population, this may be of little relevance in young children without significant vascular alterations. We assume that no memory effects arose from using the same scales twice because no scores in the total sample increased or decreased significantly from pre- to post-test. In the pre-test, the total sample scored slightly above average on all scales (mean scores for subscales reached from 10.79 to 12.54) and therefore, the chance to achieve even higher scores in the post-test might have been limited. An intervention period of 3 months was chosen based on the assumption that this should suffice to achieve adequate changes in status and tissue concentrations, and analysis from biochemical samples supports this assumption. In this study, we did not test the hypothesis that longterm B vitamin supplementation and sustained increase in B vitamin status would affect cognitive performance over long term. Bryan [8] had suggested that 1 month of vitamin B supplementation in young women tended to have a positive effect on memory performance, our aim was to specifically test the hypothesis that a short term increase in vitamin B status and concomitant decrease in homocysteine would have an impact on cognitive measures in kindergarten children. Our results clearly showed no tendency for cognitive effects after 3 months of intervention. Whole blood folate measurements from samples collected by finger prick are highly correlated with concentrations in venous blood samples [39], and this procedure is much less invasive for children than venipuncture. As the subjects who consented to blood sampling did not differ from the full-study population in any of the measures, the significant increase in whole blood folate relative to haemoglobin underscores the compliance of the subjects to the protocol and demonstrates bioavailability of the supplemented folic acid. The observed basal folate concentration of approximately 1,400 pmol per gram haemoglobin would appear to correspond to an erythrocyte folate concentration
of approximately 490 nmol/L [39]. This is similar to concentrations recently observed in British children aged 4–18 years not consuming folic acid-fortified cereals [40] and is only slightly higher than observed in children of this age group in the United States before the implementation of mandatory flour fortification with folic acid [41]. The relatively low basal red blood cell (RBC) folate indicates that our screening procedure successfully identified children with lower folate status. Because the children enrolled into our study were not considered folate deficient according to biochemical measures, the outcome might have been different in a folate deficient population. On the other hand, observational studies in populations covering the whole range of folate levels indicate that maternal folic acid supplement intake during pregnancy [42] and plasma folic acid in children [43] are related to behaviour and cognitive performance of the children, respectively. Thus, we considered it worthwhile to test the hypothesis that additional folic acid intake improves cognitive performance in children at the age of 4–5 years in a randomized trial. In addition, the German population is known for low folate intake and our approach was to investigate if cognition could be affected by folic acid in children with relatively low folate intake. In the intervention group, RBC folate increased to 776 nmol/L which is in the range of the 75th percentile of British children, who regularly consumed vitamin supplements [40]. The chosen dose of 220 lg folic acid per day in the current study seems therefore adequate to effectively increase folate status, while not causing excessively high values. In agreement with observational studies in children [30] and interventional studies in infants [44, 45] as well as interventional and observational studies in adults [46, 47], vitamin B supplementation in our study was associated with a significant decrease in plasma homocysteine. To our knowledge, this is the first study to report an effect of B vitamin supplementation on homocysteine in healthy preschool children. Urinary MMA was measured as an additional marker of vitamin B status. Similar to plasma homocysteine, age adequate values [40, 48] were observed before and after intervention, but in this case levels were not significantly decreased by intervention. As urinary MMA is a specific marker for vitamin B12 deficiency [49], this result may indicate that the vitamin B12 status of our subjects was adequate and only marginally improved with supplementation of 1.2 lg/day. For optimal functioning of one-carbon metabolism in addition to folate and vitamin B12 also the vitamins B2 and B6 are needed, whereupon their effect on homocysteine lowering has been shown to be synergistic [50]. All four vitamins were included in the supplement. Since folate and cobalamin status are the strongest determinants of
123
Eur J Nutr
homocysteine levels in children [13], we focused our evaluations on these two vitamins. Although for the screening phase total urinary p-ABG was used, ap-ABG seemed during the progress of the study to be the more reliable marker for folate status in an intervention study, as it is related to the long-term marker red cell folate, whereas p-ABG follows fast-changing serum folate [17]. Thus, ap-ABG was specifically applied in the intervention study to avoid confounding due to different time intervals between blood sampling and final supplement intake. To our knowledge, no studies exploring the effect of folic acid supplementation on urine catabolites in children have been published so far. Assuming an average creatinine excretion of 2.8 mmol/day for children between 4 and 6 years [51], we estimated a daily ap-ABG excretion of 36–60 nmol from the baseline excretion of 14.5 pmol apABG/mmol creatinine. This corresponds to approximately 50 % of the ap-ABG excretion reported for young women (82–110 nmol/day) [17]. Considering the lower weight and smaller body pool size in our subjects, this lower excretion seems plausible. In accordance with previous observations [16], ap-ABG/creatinine was responsive to increased folic acid intake, as shown by the significantly higher urine apABG excretion in the intervention group in comparison to the control group at study end, when whole blood folate concentrations were also significantly increased in the vitamin B group (Table 4). The dropout rate of children in this study was extremely small (2.5 %) and was not related to group allocation. Similarly, adverse events generally comprised common symptoms for this age group and were not related to treatment. The participating children lived mostly in above average conditions in terms of housing and parental income. The maternal education level was higher than the German average: 89 % of the mothers had an advanced school degree, which is 18 % higher than in all German women aged 20–45 years [52] and 49 % of mothers had an academic degree, 34 % above the national average [52]. Relatively favourable dietary habits could be expected in this population group, but according to the food frequency questionnaires, the median intake in the whole intervention population was only 149 lg DFE/d, indicating a significantly lower level of folate intake than the estimated average requirement for this age group (240 lg/day [14] ). In the total group included for the urinary screening procedure, the median DFE intake was 161 lg/day. These data are in line with the previously reported low folate and folic acid intake in German pre-school children [14, 15]. All cognitive testing was performed within the kindergarten setting in rooms familiar to the participating children. However, conditions of testing rooms varied between
123
kindergartens. It was not feasible to perform all pre- and post-tests at the same time of the day, but the statistical analysis confirmed no effects of testing time on cognitive performance. To achieve similar conditions for all children and to reduce potential bias, pre- and post-tests were conducted by different testers who were blinded for the group assignment as well as previous test results. Other setting specific confounders were minimized as far as possible by giving extensive instructions to all persons involved (kindergarten teachers, parents, children and testers) and suitable preparation of the testing rooms. Testers were employed for the complete testing phase. An elaborate training and instruction phase guaranteed both familiarity with the materials and the test situation and a standardized and smooth test execution. To ensure a general understanding of psychological assessment, all selected testers were students of psychology. In addition, the inclusion of ten pre-school children into a pre-test phase confirmed the children’s acceptance of the testing situation as well as applicability of test duration and order of subtests. Three of nine testers spoke German as their second language. The main criteria for selecting testers were a good command of the German language as well as a clear pronunciation. However, the results obtained showed that children tested by native speakers achieved slightly higher scores in the three verbal subscales than those tested by second-language speakers. The observed effect was nevertheless small, accounting for only 3 % of the explained variance in each of the three verbal subscales. In conclusion, despite significant changes in folate status and homocysteine, we found no indication of an effect of 3-month supplementation with B vitamins and folic acid on cognitive performance in healthy pre-school children with low but not insufficient folate intake. As folate status has been shown to correlate with verbal IQ in nourished Spanish 6-year olds [6], the short duration of this study may have contributed to the lack of observed effect. Acknowledgments We thank the participating families and kindergarten staff for their enthusiastic support of the project. We thank Sabine Eiselen (Dr von Hauner Children’s Hospital, University of Munich Medical Centre, Munich, Germany) for practical organization of the study, Martina Weber (Dr von Hauner Children’s Hospital, University of Munich Medical Centre, Munich, Germany) and Winfried Theis (Unilever R&D, Vlaardingen, Netherlands) for statistical advice, Christian Hellmuth (Dr von Hauner Children’s Hospital, University of Munich Medical Centre, Munich, Germany) for laboratory analyses, and Ingrid Pawellek (Dr von Hauner Children’s Hospital, University of Munich Medical Centre, Munich, Germany) for the evaluation of the food frequency questionnaires. MMA analyses (Institute for Clinical Chemistry, University of Munich Medical Centre, Munich, Germany) were supported by the Hans-FischerGesellschaft, Mu¨nchen. This work was supported financially in part by NUTRIMENTHE (Grant Agreement No. 212652). This project was also supported by Unilever R&D, Vlaardingen, The Netherlands.
Eur J Nutr Conflict of interest of interest.
The authors declare that they have no conflict
References 1. Bryan J, Osendarp S, Hughes D, Calvaresi E, Baghurst K, van Klinken JW (2004) Nutrients for cognitive development in school-aged children. Nutr Rev 62:295–306 2. Bailey LB, Gregory JF III (1999) Folate metabolism and requirements. J Nutr 129:779–782 3. Crowe SF, Ross CK (1997) Effect of folate deficiency and folate and B12 excess on memory functioning in young chicks. Pharmacol Biochem Behav 56:189–197 4. Troen AM, Shukitt-Hale B, Chao WH, Albuquerque B, Smith DE, Selhub J, Rosenberg J (2006) The cognitive impact of nutritional homocysteinemia in apolipoprotein-E deficient mice. J Alzheimers Dis 9:381–392 5. Lalonde R, Barraud H, Ravey J, Gueant JL, Bronowicki JP, Strazielle C (2008) Effects of a B-vitamin-deficient diet on exploratory activity, motor coordination, and spatial learning in young adult Balb/c mice. Brain Res 1188:122–131. doi:10.1016/j. brainres.2007.10.068 6. Arija V, Esparo´ G, Ferna´ndez-Ballart J, Murphy MM, Biarne´s E, Canals J (2006) Nutritional status and performance in test of verbal and non-verbal intelligence in 6 year old children. Intelligence 34:141–149. doi:10.1016/j.intell.2005.09.001 7. Strand TA, Taneja S, Ueland PM, Refsum H, Bahl R, Schneede J, Sommerfelt H, Bhandari N (2013) Cobalamin and folate status predicts mental development scores in North Indian children 12–18 mo of age. Am J Clin Nutr 97:310–317. doi:10.3945/ajcn. 111.032268 8. Bryan J, Calvaresi E, Hughes D (2002) Short-term folate, vitamin B-12 or vitamin B-6 supplementation slightly affects memory performance but not mood in women of various ages. J Nutr 132:1345–1356 9. Louwman MW, van Dusseldorp M, van de Vijver FJ, Thomas CM, Schneede J, Ueland PM, Refsum H, van Staveren WA (2000) Signs of impaired cognitive function in adolescents with marginal cobalamin status. Am J Clin Nutr 72:762–769 10. Gewa CA, Weiss RE, Bwibo NO, Whaley S, Sigman M, Murphy SP, Harrison G, Neumann CG (2009) Dietary micronutrients are associated with higher cognitive function gains among primary school children in rural Kenya. Br J Nutr 101:1378–1387. doi:10. 1017/S0007114508066804 11. Homocysteine Lowering Trialists’ Collaboration (1998) Lowering blood homocysteine with folic acid based supplements: metaanalysis of randomised trials. BMJ 316:894–898 12. Lonn E, Yusuf S, Arnold MJ, Sheridan P, Pogue J, Micks M, McQueen MJ, Probstfield J, Fodor G, Held C, Genest J Jr (2006) Homocysteine lowering with folic acid and B vitamins in vascular disease. N Engl J Med 354:1567–1577. doi:10.1056/ NEJMoa060900 13. Ueland PM, Monsen AL (2003) Hyperhomocysteinemia and B-vitamin deficiencies in infants and children. Clin Chem Lab Med 41:1418–1426. doi:10.1515/CCLM.2003.218 14. Sichert-Hellert W, Wenz G, Kersting M (2006) Vitamin intakes from supplements and fortified food in German children and adolescents: results from the DONALD study. J Nutr 136:1329–1333 15. Mensink GBE, Herseker H, Richter A, Stahl A, Vohmann C (2007) Erna¨hrungstudie als KIGGS-Modul (EsKiMo), German [KIGGS nutrition survey]. In: Robert Koch Institut and Universita¨t Paderborn, Berlin and Paderborn
16. Wolfe JM, Bailey LB, Herrlinger-Garcia K, Theriaque DW, Gregory JF III, Kauwell GP (2003) Folate catabolite excretion is responsive to changes in dietary folate intake in elderly women. Am J Clin Nutr 77:919–923 17. Kim HA, Choi JH, Lim HS (2007) Childbearing women of twenty and under are at greater risk than those of twenty-five and over for compromised folate status. Nutr Res Pract 1:254–259. doi:10.4162/nrp.2007.1.4.254 18. Moher D, Schulz KF, Altman D (2001) The CONSORT statement: revised recommendations for improving the quality of reports of parallel-group randomized trials. JAMA 285:1987–1991 19. Scientific Committee on Food & Scientific Panel on Dietetic Products Nutrition and Allergies (2006) Tolerable upper intake levels for vitamins and minerals. In: European Food Safety Authority 20. Ricken G, Fritz A, Schuck KD, Preuß U (eds) (2007) HAWIVAIII, Hannover-Wechsler-Intelligenztest fu¨r das Vorschulalter-III, German [HAWIVA-III, Hannover-Wechsler-Preschool Scale of Intelligence-III]. Huber, Go¨ttingen 21. Wechsler D (2002) Wechsler Preschool and Primary Scale of Intelligence-III (WPPSI-III). The Psychological Corporation, San Antonio, TX 22. Kaufman AS, Kaufman NL (1983) Kaufman assessment battery for children: K-ABC. American Guidance Service, Circle Pines, MN 23. Melchers P, Preuß U (2005) Kaufman assessment battery for children. Pearson Assessment, Frankfurt/Main 24. Bartels H, Bo¨hmer M, Heierli C (1972) Serum kreatininbestimmung ohne enteiweissen, German [Serum creatinine determination without protein precipitation]. Clin Chim Acta 37:193–197. doi:10.1016/0009-8981(72)90432-9 25. Niesser M, Harder U, Koletzko B, Peissner W (2013) Quantification of urinary folate catabolites using liquid chromatographytandem mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci 929:116–124. doi:10.1016/j.jchromb.2013.04. 008 26. Matchar DB, Feussner JR, Millington DS, Wilkinson RH Jr, Watson DJ, Gale D (1987) Isotope-dilution assay for urinary methylmalonic acid in the diagnosis of vitamin B12 deficiency. A prospective clinical evaluation. Ann Intern Med 106:707–710 27. O’Broin S, Kelleher B (1992) Microbiological assay on microtitre plates of folate in serum and red cells. J Clin Pathol 45:344–347 28. Heuck CC, Reinauer H, Wood WG (2008) The alkaline haematin detergent (AHD575) method for the determination of haemoglobin in blood–a candidate reference measurement procedure. Clin Lab 54:255–272 29. Hellmuth C, Koletzko B, Peissner W (2011) Aqueous normal phase chromatography improves quantification and qualification of homocysteine, cysteine and methionine by liquid chromatography-tandem mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci 879:83–89. doi:10.1016/j.jchromb.2010.11. 016 30. Huemer M, Vonblon K, Fodinger M, Krumpholz R, Hubmann M, Ulmer H, Simma B (2006) Total homocysteine, folate, and cobalamin, and their relation to genetic polymorphisms, lifestyle and body mass index in healthy children and adolescents. Pediatr Res 60:764–769. doi:10.1203/01.pdr.0000246099.39469.18 31. R Development Core Team (2010) R: A language and environment for statistical computing. In: R Foundation for Statistical Computing, Vienna, Austria 32. Heseker H, Kuebler W, Westenhoefer J, Pudel V (1990) Psychische Veraenderungen als Fruehzeichen einer suboptimalen Vitaminversorgung, German [Psychic changes as early symptoms
123
Eur J Nutr
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
of suboptimal vitamin supply]. 2. Ernaehrungs-Umschau 37: 87–94 Haskell CF, Scholey AB, Jackson PA, Elliott JM, Defeyter MA, Greer J, Robertson BC, Buchanan T, Tiplady B, Kennedy DO (2008) Cognitive and mood effects in healthy children during 12 weeks’ supplementation with multi-vitamin/minerals. Br J Nutr 100:1086–1096. doi:10.1017/S0007114508959213 Osendarp SJ, Baghurst KI, Bryan J, Calvaresi E, Hughes D, Hussaini M, Karyadi SJ, van Klinken BJ, van der Knaap HC, Lukito W, Mikarsa W, Transler C, Wilson C (2007) Effect of a 12-mo micronutrient intervention on learning and memory in well-nourished and marginally nourished school-aged children: 2 parallel, randomized, placebo-controlled studies in Australia and Indonesia. Am J Clin Nutr 86:1082–1093 Durga J, Verhoef P, Anteunis LJ, Schouten E, Kok FJ (2007) Effects of folic acid supplementation on hearing in older adults: a randomized, controlled trial. Ann Intern Med 146:1–9 Vogel T, Dali-Youcef N, Kaltenbach G, Andres E (2009) Homocysteine, vitamin B12, folate and cognitive functions: a systematic and critical review of the literature. Int J Clin Pract 63:1061–1067. doi:10.1111/j.1742-1241.2009.02026.x Fabian E, Bogner M, Kickinger A, Wagner KH, Elmadfa I (2011) Intake of medication and vitamin status in the elderly. Ann Nutr Metab 58:118–125. doi:10.1159/000327351 Mansoor MA, Hervig T, Stakkestad JA, Drablos PA, Apeland T, Wentzel-Larsen T, Bates CJ (2011) Serum folate is significantly correlated with plasma cysteine concentrations in healthy industry workers. Ann Nutr Metab 58:68–73. doi:10.1159/000325537 O’Broin SD, Kelleher BP, Davoren A, Gunter EW (1997) Fieldstudy screening of blood folate concentrations: specimen stability and finger-stick sampling. Am J Clin Nutr 66:1398–1405 Kerr MA, Livingstone B, Bates CJ, Bradbury I, Scott JM, Ward M, Pentieva K, Mansoor MA, McNulty H (2009) Folate, related B vitamins, and homocysteine in childhood and adolescence: potential implications for disease risk in later life. Pediatrics 123:627–635. doi:10.1542/peds.2008-1049 Pfeiffer CM, Johnson CL, Jain RB, Yetley EA, Picciano MF, Rader JI, Fisher KD, Mulinare J, Osterloh JD (2007) Trends in blood folate and vitamin B-12 concentrations in the United States, 1988 2004. Am J Clin Nutr 86:718–727 Roza SJ, van Batenburg-Eddes T, Steegers EA, Jaddoe VW, Mackenbach JP, Hofman A, Verhulst FC, Tiemeier H (2010) Maternal folic acid supplement use in early pregnancy and child
123
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
behavioural problems: The Generation R Study. Br J Nutr 103:445–452. doi:10.1017/S0007114509991954 Nguyen CT, Gracely EJ, Lee BK (2013) Serum folate but not vitamin B-12 concentrations are positively associated with cognitive test scores in children aged 6–16 years. J Nutr 143:500–504. doi:10.3945/jn.112.166165 Bjorke-Monsen AL, Torsvik I, Saetran H, Markestad T, Ueland PM (2008) Common metabolic profile in infants indicating impaired cobalamin status responds to cobalamin supplementation. Pediatrics 122:83–91. doi:10.1542/peds.2007-2716 Torsvik I, Ueland PM, Markestad T, Bjorke-Monsen AL (2013) Cobalamin supplementation improves motor development and regurgitations in infants: results from a randomized intervention study. Am J Clin Nutr 98:1233–1240. doi:10.3945/ajcn.113. 061549 Homocysteine Lowering Trialists’ Collaboration (2005) Dosedependent effects of folic acid on blood concentrations of homocysteine: a meta-analysis of the randomized trials. Am J Clin Nutr 82:806–812 Chew SC, Khor GL, Loh SP (2011) Association between dietary folate intake and blood status of folate and homocysteine in Malaysian adults. J Nutr Sci Vitaminol (Tokyo) 57:150–155 Erdogan E, Nelson GJ, Rockwood AL, Frank EL (2010) Evaluation of reference intervals for methylmalonic acid in plasma/ serum and urine. Clin Chim Acta 411:1827–1829. doi:10.1016/j. cca.2010.06.017 Bjorke Monsen AL, Ueland PM (2003) Homocysteine and methylmalonic acid in diagnosis and risk assessment from infancy to adolescence. Am J Clin Nutr 78:7–21 Stanger O, Herrmann W, Pietrzik K, Fowler B, Geisel J, Dierkes J, Weger M, e.V D-LH (2003) DACH-LIGA homocystein (german, austrian and swiss homocysteine society): consensus paper on the rational clinical use of homocysteine, folic acid and B-vitamins in cardiovascular and thrombotic diseases: guidelines and recommendations. Clin Chem Lab Med 41:1392–1403. doi:10.1515/CCLM.2003.214 Lentze MJ, Schaub J, Schulte FJ, Spranger J (2007) Pa¨diatrie: Grundlagen und Praxis, German [Pediatrics: Basics and practice]. In: Springer Medizin Verlag Heidelberg, Berlin, Heidelberg Statistisches Bundesamt (ed) (2011) Statistisches Jahrbuch 2011, German [statistical yearbook 2011]. Statistisches Bundesamt, Wiesbaden
