Hyperhomocysteinemia may contribute to the development of osteoporosis. The relationship of Hcy and vitamin B12 with bone turnover markers, BUA, and fracture incidence was studied in 1267 subjects of the Longitudinal Aging Study Amsterdam. High Hcy and low vitamin B12 concentrations were significantly associated with low BUA, high markers of bone turnover, and increased fracture risk.
Introduction: Hyperhomocysteinemia may contribute to the development of osteoporosis. Vitamin B12 is closely correlated to homocysteine (Hcy). The main objective of our study was to examine the association of Hcy and vitamin B12 status and the combined effect of these two with broadband ultrasound attenuation (BUA), bone turnover markers, and fracture.
Materials and Methods: Subjects were 615 men and 652 women with a mean age of 76 ± 6.6 (SD) years of the Longitudinal Aging Study Amsterdam (LASA). At baseline (1995/1996), blood samples were taken after an overnight fast for dairy products. Plasma Hcy was measured with IMx, serum vitamin B12 with competitive immunoassay (IA) luminescence, serum osteocalcin (OC) with immunoradiometric assay (IRMA), and urinary excretion of deoxypyridinoline (DPD) with competitive IA and corrected for creatinine (Cr) concentration. CVs were 4%, 5%, 8%, and 5%, respectively. BUA was assessed in the heel bone twice in both the right and left calcaneus. Mean BUA value was calculated from these four measurements. CV was 3.4%. After baseline measurements in 1995, a 3-year prospective follow-up of fractures was carried out until 1998/1999. Subjects were grouped by using two different approaches on the basis of their vitamin B12 concentration, normal versus low (<200 pM) or lowest quartile (Q1) versus normal quartiles (Q2-Q4), and Hcy concentration, normal versus high (>15 μM) or highest quartile (Q4) versus normal quartiles (Q1-Q3). Analysis of covariance was performed to calculate mean values of BUA, OC, and DPD/Crurine based on the specified categories of Hcy and vitamin B12 and adjusted for several confounders (potential confounders were age, sex, body weight, body height, current smoking [yes/no], mobility, cognition). The relative risk (RR) of any fracture was assessed with Cox regression analysis. Quartiles were used when Hcy and vitamin B12 were separately studied in their relationship with fracture incidence.
Results: Fourteen percent of the men and 9% of the women had high Hcy (>15 μM) and low vitamin B12 (<200 pM) concentrations. Women with vitamin B12 levels <200 pM and Hcy concentrations >15 μM had lower BUA, higher DPD/Cr, and higher OC concentrations than their counterparts. In men, no differences were found between the different Hcy and vitamin B12 categories in adjusted means of BUA, OC, or DPD/Crurine. Twenty-eight men and 43 women sustained a fracture during the 3-year follow-up period. The adjusted RR for fractures (95% CI) for men with high Hcy and/or low vitamin B12 concentrations was 3.8 (1.2-11.6) compared with men with normal Hcy and vitamin B12 concentrations. Women with high Hcy and/or low vitamin B12 concentrations had an adjusted RR for fractures of 2.8 (1.3-5.7).
Conclusions: High Hcy and low vitamin B12 concentrations were significantly associated with low BUA, high markers of bone turnover, and increased fracture risk.
Osteoporosis is a multifactorial disease with an increasing incidence with aging. The burden of osteoporosis is caused by fractures of the distal radius, vertebrae, hip, and other bones. This burden on society is increasing because of the rapidly increasing number of elderly people. Fractures are associated with considerable morbidity and mortality, impaired quality of life, and high costs for society.(1–5)
A number of modifiable risk factors for osteoporosis and fractures have been identified, including nutritional factors such as vitamin D deficiency and low calcium intake.(6–8) It is well known that homocystinuria patients are often diagnosed with osteoporosis.(9) High homocysteine (Hcy) levels can be modified by dietary folate, vitamin B12, and vitamin B6. Both moderate hyperhomocysteinemia and vitamin B12 deficiency are highly prevalent in old age(10–12) and may play a role in diseases that are characteristic for old age. In vivo and in vitro studies support disturbance of cross-linking of collagen in bone by Hcy as a possible explanation.(13,14) Recently, a relationship was observed between Hcy levels and the incidence of fracture in three subcohorts from two independent prospective population-based studies of older men and women. A serum Hcy level in the highest quartile doubled the risk for fractures compared with the three lowest quartiles, and this increased risk seemed to be independent of age, sex, and other risk factors for fracture, such as smoking, recent falling, diabetes mellitus, and peripheral arterial disease.(15) Similar findings were observed in the study of McLean et al.(16)
In one of these prospective studies, the Longitudinal Aging Study Amsterdam (LASA), data on BMD, broadband ultrasound attenuation (BUA), and bone turnover markers were also collected. Van Meurs et al.(15) showed that Hcy was not associated with BMD. Our study aims to show that Hcy and vitamin B12 status are associated with bone strength. BUA can predict bone strength(17) and risk for hip fracture or any fracture.(18) The main objective of our study was to examine the association of Hcy and vitamin B12 status with BUA, bone turnover markers, and fracture incidence. The second objective was to provide evidence of the additive, combined effect of Hcy and vitamin B12 status on BUA, bone turnover markers, and fracture incidence.
Our hypothesis is that high serum Hcy, possibly in combination with low serum vitamin B12, is associated with low BUA, high concentrations of the bone formation marker osteocalcin (OC), high concentrations of the bone resorption marker deoxypyridinoline (DPD; which was corrected for creatinine [Cr]), and a higher incidence of fractures. This would indicate that the metabolite Hcy could influence bone strength and cause collagen disturbances in bone as has been suggested by Lubec et al.(13)
MATERIALS AND METHODS
Study sample and procedures
This study was performed within a subsample of LASA.(19) For LASA, a random sample of elderly men and women (55-85 years), stratified by age, sex, level of urbanization, and expected 5-year mortality, was drawn from the population registers of 11 municipalities in three areas of The Netherlands. Data collection took place during four cycles of data collection, namely in 1992/1993, 1995/1996, 1998/1999, and 2001/2002. For this study, 1995/1996 was considered as baseline. The data that were collected in 1995/1996 were used for the cross-sectional data analysis to examine the relation of Hcy status and vitamin B12 status with BUA and the bone turnover markers DPD/Cr and OC. The fracture data that were collected from 1995/1996 to 1998/1999 were used for the prospective cohort analysis to examine the relation of Hcy status and vitamin B12 status with fractures. The data collected in 1992/1993 and 2001/2002 were not used in this study.
In the second cycle in 1995/1996, medical data were obtained from 1509 participants who were 65 years or older on 1 January 1996 (Fig. 1). Respondents were subsequently invited to the VU University Medical Center (VUMC) or a health care center near their homes, where BUA measurements were performed and blood and urine samples were collected (n = 1321). Information on incident fractures and falls was obtained for 1289 respondents between the second cycle of data collection in 1995/1996 and the third cycle of data collection in 1998/1999. More sampling and data collection procedures have been described in detail elsewhere.(20,21)
Informed consent was obtained from all respondents. The study was approved by the Medical Ethics Committee (VUMC) and conducted according to the principles of the Helsinki declaration.
Quantitative ultrasound measurements
Quantitative ultrasound data were obtained using the CUBA Clinical instrument (McCue Ultrasonics, Winchester, UK). BUA (dB/MHz) was measured twice in both the right and left calcaneus. Mean BUA value was calculated from these four measurements.(18) The CV, calculated in 20 healthy volunteers measured on five occasions consecutively within 1 h, was 3.4%.(22)
Ascertainment of fractures
Data on fractures that occurred between the second examination (1995/1996) and the third examination in 1998/1999 were prospectively reported by the participant on a calendar. Eighty-two percent of all reported fractures were verified by a physician or by radiographs. To reach sufficient power, every osteoporotic fracture was noted as an outcome measure. Fractures caused by an (motor vehicle) accident (n = 10) and fractures of the head, hand, fingers, foot, toes, ankle, and vertebrae (n = 20) were excluded. Duration of follow-up was calculated as the time from the first examination in 1995/1996 to the first occurrence of a fracture in weeks.
Assessment of falls
Participants were asked to record fall events on a weekly “fall calendar” for 3 years. A fall was defined as “an unintentional change in position resulting in coming to rest at a lower level or on the ground.” A “recurrent faller” was defined as a subject who fell at least two times within 6 months during the 3-year fall follow-up.(23)
Blood was collected in 1995 and 1996 after an overnight fast, and EDTA plasma samples were analyzed for total Hcy (tHcy) with the Abbott IMx analyzer at the Laboratory of Clinical Chemistry (VUMC). The IMx method uses fluorescence polarization immunoassay (FPIA) technology. Serum levels of vitamin B12, OC, and estradiol were determined at the Endocrine Laboratory (VUMC) with a competitive immunoassay luminescence on the automated ACS 180 System (Bayer Diagnostics, Mijdrecht, The Netherlands). Serum OC was measured with an immunoradiometric assay (Biosource Diagnostics, Fleuris, Belgium). Estradiol concentrations were measured by radioimmunoassay (Diasorin Biomedica, Saluggia, Italy). CVs were 4%, 5%, 8%, and 10%, respectively.
Overnight urinary excretion of DPD was determined at the Endocrine Laboratory (VUMC). DPD was determined with a competitive immunoassay on an automated ACS 180 System (Chiron Diagnostics, Emeryville, CA, USA). The values were corrected for creatinine concentration (Cr) in the same urine sample. CV was 5%.
Levels of serum 25-hydroxyvitamin D were measured with a competitive protein-binding assay (Nichols Institute Diagnostics). CV was 10-15%.
Baseline information on age and sex was derived from the municipal registries. During the main interview of the second data collection, body weight, body height, current smoking (yes/no), and mobility were assessed in a face-to-face interview. Body weight was measured without clothes and shoes using a calibrated bathroom scale. Height was measured using a stadiometer. Body mass index (BMI) was calculated as body weight in kilograms divided by height in meters squared. The level of mobility was assessed with three physical performance tests,(24) including a walking test, a chair stands test, and a cardigan test. For each test, a score of one to four points was assigned corresponding to the quartile of the time needed. The more time that was needed, the lower the score. Participants who were not able to perform a test obtained a score of zero points. The scores of the three tests were summed up to a physical performance score (range, 0-12).
Cognition was measured using the Mini-Mental State Examination (MMSE), which is a questionnaire used as a screening test for general cognitive functioning. The highest possible score is 30; a score of ≤23 indicates the presence of a cognitive impairment.(25)
Because of the substantial differences in fractures, BUA, and bone markers between men and women, all analyses with osteoporosis measures as outcome were stratified by sex. The distributions of Hcy, OC, and DPD/Crurine were normalized by transformation to their natural logarithm.
Men and women were divided into different categories of Hcy and vitamin B12 concentrations. For these different categories, the adjusted means of the outcome measurements BUA, OC, and DPD/Crurine were calculated by performing analysis of covariance.
The risk of fracture with Hcy and vitamin B12 levels were evaluated with a quartile-based analysis and with predefined cut-off points for Hcy (>15 μM) and vitamin B12 (<200 pM) levels. The quartiles were defined in a sex- and age-specific manner for each of the 5-year categories. The relative risk (RR; 95% CI) of any fracture (except nonosteoporotic fractures and fractures caused by an accident) was assessed with Cox regression analysis for Hcy and vitamin B12 quartiles separately. For calculation of risk of fracture on the basis of cut-off points, the group with normal Hcy and vitamin B12 levels was used as the reference group; the other group therefore had high Hcy and/or low vitamin B12 levels.
Potential confounders were included in the multivariate models to control for confounding and to enhance precision. Because age is a very well-known determinant for markers of bone turnover and fracture risk, all analyses were adjusted for age.
The mean age of the LASA participants in this study was 75 years. Forty-six percent of the men and 29% of the women had a Hcy concentration >15 μM.(26,27) A low serum vitamin B12 level (<200 pM) occurred in 22% of the men and in 16% of the women. The combination of elevated Hcy and low vitamin B12 concentration occurred in 14% of the men and in 9% of the women. The median serum OC concentration was ∼2 nM and the median urinary DPD/Cr excretion was ∼5 nM for both men and women. Twenty-eight men and 43 women sustained an osteoporotic fracture during the 3-year follow-up period. Twenty-five percent of the subjects were recurrent fallers according to the definition (Table 1).
Table Table 1.. Characteristics of 630 Men and 668 Women From the LASA Study
Association of Hcy status and vitamin B12 status to fracture risk (prospective cohort study)
High plasma Hcy levels were associated with an increased fracture risk. After adjustment for age, the overall RR for fracture for each increment of 1 SD in Hcy level was 1.4 (95% CI, 1.0-2.0) for men and 1.1 (95% CI, 0.8-1.6) for women. After multivariate adjusting for several confounders, the RRs increased slightly. In Table 2, RRs are shown for the upper quartile (age- and sex-specific) of plasma Hcy concentration compared with the combined lower three quartiles (as the reference group). Men with a Hcy concentration in the upper quartile had a >2-fold increased risk to suffer a fracture after adjusting for age and after multivariate adjusting. In women, the increased RRs were not significant.
Table Table 2.. Multivariate Cox Proportional Hazards Regression Models for the Risk of Fractures by Hcy and Vitamin B12 Status in 1252 Men and Women From the LASA Study
Similar analyses were performed to study the RR of low serum vitamin B12 concentration for fracture. Women with a serum vitamin B12 concentration in the highest three quartiles were protected from a fracture compared with women in the lowest quartile (Table 2). After adjusting for age, women in the lowest vitamin B12 quartile had a RR for fractures of 2.2 (95% CI, 1.1-4.3) compared with those in the highest three quartiles. No significant relationship between vitamin B12 concentration and fracture risk was observed in men. RRs for fracture incidence for each decrease of 1 SD in vitamin B12 level were similar as to the RRs after dichotomizing the vitamin B12 concentration (data not shown).
After multivariate adjusting, men with both a high plasma Hcy concentration (>15 μM) and/or a low serum vitamin B12 concentration (<200 pM) had a 3.8 (95% CI, 1.2-11.6) times higher risk to suffer a fracture than men with both a normal plasma Hcy concentration and serum vitamin B12 concentration. In women, the RR for a fracture was 2.8 (95% CI, 1.3-5.7) in those women with a high Hcy concentration and/or a low vitamin B12 concentration after multivariate adjusting (Table 2).
Association of Hcy status and vitamin B12 status to BUA (cross-sectional study)
Figure 2A shows the mean BUA for different Hcy and vitamin B12 groups adjusted for age, BMI, and creatinine. Women with a low serum vitamin B12 concentration and high plasma Hcy concentration had a significantly lower adjusted mean BUA value compared with women with normal vitamin B12 and Hcy concentrations (p = 0.004) and also compared with women with a low vitamin B12 and normal Hcy concentration (p = 0.04). In men, no significant differences were observed in adjusted BUA levels in the different categories for vitamin B12 and Hcy concentrations.
Association of Hcy status and vitamin B12 status to bone turnover markers
After adjusting for age and BMI, median urinary DPD/Cr excretion was significantly higher in women with a low serum vitamin B12 and high plasma Hcy concentration compared with women with a normal vitamin B12 and Hcy concentration (p = 0.0003) and also compared with women with a normal vitamin B12 and high Hcy concentration (p = 0.04) (Fig. 2B). In men, again, no significant differences were observed in urinary DPD/Cr excretion for different Hcy and vitamin B12 groups.
Figure 2C shows that, after adjusting for age, BMI, and serum Cr concentration, women with a normal vitamin B12 and Hcy concentration had a significantly lower mean OC concentration than women with a low vitamin B12 and high Hcy concentration (p = 0.015) and women with a low vitamin B12 and normal Hcy concentration (p = 0.01).
Association of Hcy status and vitamin B12 status to recurrent falling
Because fractures are usually caused by a fall, we evaluated whether Hcy or vitamin B12 status was associated with falling. Neither Hcy status nor vitamin B12 status was related to recurrent falling (odds ratio was for both Hcy and vitamin B12 1.0 [0.9-1.1]).
This study showed that the presence of a low vitamin B12 concentration, high Hcy concentration, or having both conditions was associated with a three times higher risk for fractures in men and women. After inclusion of potential confounders, the RRs changed but remained significant. This implies that the increased RRs caused by a high Hcy concentration and/or low vitamin B12 concentration were independent of other common risk factors for fractures. The risk to suffer from a fracture was more than two times higher in men with a Hcy level in the highest quartile, whereas, surprisingly, no significant higher RRs were observed in women. A test for trend in increased risk for fracture caused by increased Hcy quartiles was also not observed in women (p = 0.53). Even though a higher number of men had vitamin B12 deficiency, vitamin B12 deficiency alone was not associated with a higher RR for fractures. A test for trend in increased risk for fracture caused by decreased vitamin B12 quartiles was not observed in men (p = 0.24). The association between vitamin B12 status and fractures was, however, high in women. In this study, the impact of a low vitamin B12 status is apparently more severe in women than in men, whereas a high Hcy status is more severely associated with fractures in men than in women.
The results of the second part (the cross-sectional study) showed that, in women, the combination of low vitamin B12 and high Hcy status was associated with low BUA and also with elevated bone turnover markers DPD/Cr and OC.
Fractures are most often caused by a fall. Hcy and vitamin B12 status was not associated with recurrent falling. Therefore, falling cannot be an intermediate step in the association of Hcy or vitamin B12 status with fracture incidence. In other words, falling is not a consequence of low vitamin B12 status or high Hcy status, but the Hcy and vitamin B12 status is directly related to fractures.
The major strength of our study is the use of two types of data: prospective data and observational data. Hcy status in this study was similarly associated with fractures in longer prospective studies such as two cohorts of the Rotterdam Elderly Study with a follow-up time of 6 and 8 years(15) and the Framingham Heart Study with a follow-up time of 11.7 years.(16) Our cross-sectional results can show the association, but they cannot ascertain a causal relationship. A prospective study would have been more suitable to also explore the association between Hcy and B12 with bone turnover markers.
As has been suggested by McKusick,(14) the mechanism underlying the association of Hcy with fracture risk, BUA, and bone turnover markers might involve interference of Hcy with collagen cross-linking. Later, Lubec et al.(13) indeed showed lower amounts of collagen cross-links in serum of homocystinuria patients compared with those in normal controls. However, homocystinuria is often accompanied by several other medical conditions, such as clinical manifestations involving the eyes, high blood pressure, stomach, the vasculature, and the central nervous system,(28) which could influence collagen cross-links as well. This study gives further evidence for the statement of McKusick, because women with elevated Hcy concentrations had higher DPD/Cr concentrations than women with a normal Hcy concentration. Interference in cross-link formation would result in an altered bone matrix, resulting in more fragile bone.
BUA and DPD/Cr are good predictors of BMD and fracture risk.(18,29,30) Our observation that high Hcy status and low vitamin B12 status are associated with low BUA in women support the proposed mechanism that vitamin B12 deficiency and (as a consequence) elevated Hcy levels may play an important role in the cross-linking of collagen.
Of interest is the finding that OC, a noncollagenous protein,(31) is not related to collagen cross-linking. Nevertheless, we found an association of Hcy status and vitamin B12 status with OC. This suggests that bone formation may also be affected by elevated Hcy levels and low vitamin B12 status. We are not aware of other studies that examined the association of Hcy status and vitamin B12 status with OC. Therefore, more studies are needed to confirm and further explore this finding. OC influences mineralization,(32) and serum OC has been associated with BMD and bone loss.(33) In this study, we did not find an association of Hcy status,(15) vitamin B12 status, or the combination of Hcy and vitamin B12 status (data not shown) with BMD of spine and hip sites. In another population, we have shown a higher prevalence of osteoporosis (defined by BMD T score < −2.5) in vitamin B12-deficient frail elderly women than in their counterparts with normal vitamin B12 status.(34) Stone et al.(35) recently showed that low vitamin B12 levels were associated with increased hip bone loss in older women, but unfortunately, no information was available on Hcy and methylmalonic acid levels. The explanation for the association between vitamin B12 concentration and osteoporosis is less clear than for the association between Hcy and osteoporosis. In vitro experiments showed that alkaline phosphatase activity increased when vitamin B12 was added in a concentration-dependent manner to osteoblastic cells.(36) In a clinical study, alkaline phosphatase and OC concentrations rose significantly in vitamin B12-deficient patients after treatment with vitamin B12, whereas they remained unchanged in the control group.(37)
From our results, it cannot be concluded whether the elevation of Hcy levels or the vitamin B12 deficiency is causally responsible for the low BUA and increased fracture risk. Elevated Hcy levels are highly correlated with low folic acid levels and low vitamin B12 levels.(38,39) Unfortunately, we do not have data on folic acid levels in our participants. Therefore, our study does not provide conclusive evidence on the dominance of Hcy status or vitamin B12 status.
Studies on gene polymorphisms that have only an effect on Hcy status(40) might give an indication on the role of Hcy with regard to osteoporosis outcome measures. Abrahamsen et al.(41) showed a relation between the polymorphism of the methylenetetrahydrofolate reductase (MTHFR) gene and fracture incidence in the Danish Osteoporosis Prevention Study. Postmenopausal women with the TT genotype had a higher hazard ratio for fractures than those with the CC or CT genotype. Similar odds ratios were found in the Danish twin study.(42) Unfortunately the association between Hcy concentrations and fracture risk was not assessed in this study, although data were available. In a Danish case-control study, contradictory results were found: increased odds ratios of fracture for postmenopausal women with the wildtype C allele.(43) A lower BMD was found in Japanese(44) and Danish postmenopausal women(40) with the MTHFR TT genotype than women with the CC or CT genotype. McLean et al(45) found contradictory results in the association of CC, CT, and TT individuals with BUA and BMD after dividing the individuals into groups with low and normal folate status.
What could explain the difference in results found between men and women? We found an association between Hcy status and fracture risk in men but not in women, whereas we found an association between vitamin B12 status and fractures in women but not in men. It is known that hormone replacement therapy suppresses Hcy levels in women.(46–49) We evaluated whether serum estradiol levels confounded the association between Hcy and fractures. Because estradiol levels were not correlated to Hcy or vitamin B12, the levels were not associated with fractures, and they did not confound the association of Hcy status or vitamin B12 status with fractures (data not shown).
Our study has some limitations. First, the respondents of this study are the healthier older people of a population sample, because the frailest respondents of the LASA study could not visit the hospital or health care center for blood sampling and ultrasound measurement. Figlin et al.(50) showed a trend toward a higher incidence of folate deficiency and vitamin B12 deficiency in dependent persons compared with persons who were living independently. It is likely that the prevalence of vitamin B12 deficiency was also higher in these nonresponders. Therefore, underestimation of the associations might have occurred. Second, although the sample size was relatively large, the power to detect significant differences was still limited for most outcome measures. We cannot exclude the possibility that there was an association present between Hcy status or vitamin B12 status and BMD, because BMD was only measured in a part of this cohort study (n = 515). Third, cut-off points for vitamin B12 deficiency are still not generally defined and accepted. Several studies have used 150 pM as a cut-off point (see review of Baik and Russell(51)), although now, more studies choose a higher cut-off point for defining mild vitamin B12 deficiency in combination with elevated methylmalonic acid (MMA) concentrations. For defining vitamin B12 deficiency, 258 pM was used in the Framingham Heart Study,(52) and 260 pM was used in combination with 0.32 μM for MMA in the Dutch part of the SENECA study.(10) In the Dutch study, a prevalence of 24% was found for vitamin B12 deficiency. Therefore, we decided to use 200 pM as a cut-off point for vitamin B12 deficiency, which was found in about 20% of the participants in our study. Fourth, we cannot exclude that the elevated risks for suffering from a fracture, caused by high Hcy or low vitamin B12 concentrations, are because of a low nutritional score and not because of the disturbed Hcy or vitamin B12 concentrations; nutritional data on vitamin B12 intake were not available. However, we note that the problem of vitamin B12 deficiency in elderly people is often a problem of malabsorption of vitamin B12 rather than a low dietary intake of vitamin B12.(10)
In conclusion, the results of this large community-based study show that high plasma levels of Hcy and low serum levels of vitamin B12 and especially the combination of these two parameters are related to high fracture risk, low BUA, and increased bone turnover markers. Large-scale trials of folic acid and vitamin B12 supplements have shown an effective reduction in Hcy levels.(53) A causal relation between Hcy status and collagen cross-links could be further explored by measuring the urinary DPD/Cr excretion or serum cross-links concentration after folic acid and vitamin B12 supplementation. In addition, long-term intervention trials with vitamin B supplementation should be performed to evaluate fracture risk and its relation to Hcy and vitamin B12 status. Furthermore, in vitro studies should be performed to increase complementary understanding of the underlying mechanism.
This study was based on data collected from the LASA, which is largely funded by the Ministry of Health, Welfare, and Sports of The Netherlands. It was partly supported by the Praeventiefonds, The Hague, The Netherlands (Grant 28-25510) and by the Dutch Dairy Association, Zoetermeer, The Netherlands.