- Email: [email protected]
Intake of vitamin K1 and K2 and risk of hip fractures: The Hordaland Health Study
Bone 49 (2011) 990–995
Contents lists available at ScienceDirect
Bone j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / b o n e
Original Full Length Article
Intake of vitamin K1 and K2 and risk of hip fractures: The Hordaland Health Study Ellen M. Apalset a, b,⁎, Clara G. Gjesdal b, c, Geir E. Eide a, d, Grethe S. Tell a a
Department of Public Health and Primary Health Care, University of Bergen, Kalfarveien 31, N-5018 Bergen, Norway Department of Rheumatology, Haukeland University Hospital, N-5021 Bergen, Norway c Section of Rheumatology, Institute of Medicine, University of Bergen, Bergen, Norway d Centre for Clinical Research, Haukeland University Hospital, N-5021 Bergen, Norway b
a r t i c l e
i n f o
Article history: Received 26 April 2011 Revised 5 July 2011 Accepted 23 July 2011 Available online 2 August 2011 Edited by: Robert Recker Keywords: Hip fracture Osteoporosis Apolipoprotein E Diet Vitamin K1 Vitamin K2
a b s t r a c t Background: Evidence of the effect of vitamin K on bone health is conflicting. The aim was to investigate the association between intake of vitamins K1 and K2 and subsequent risk of hip fracture in a general population sample, as well as potential effect modification by apolipoprotein E gene (APOE) status by presence of the E4 allele. Methods: 1238 men and 1569 women 71–75 years of age were included in the community-based Hordaland Health Study 1997–1999 in Western Norway. Information on hip fracture was obtained from hospitalizations in the region from enrolment until 31 December 2009. Information on intake of vitamins K1 and K2 collected at baseline was used as potential predictors of hip fracture in Cox proportional hazards regression analyses. Results: Participants in the lowest compared to the highest quartile of vitamin K1 intake had increased risk of suffering a hip fracture (hazard ratio (HR)= 1.57 [95% CI 1.09, 2.26]). Vitamin K2 intake was not associated with hip fracture. Presence of APOE4-allele did not increase the risk of hip fracture, nor was there any effect modification with vitamin K1 in relation to risk of hip fracture. Conclusions: A low intake of vitamin K1, but not K2, was associated with an increased risk of hip fractures. © 2011 Elsevier Inc. All rights reserved.
Introduction Hip fractures have serious consequences, leading to impaired health, disability and increased mortality [1,2]. Possible preventive strategies include change in health behaviors and optimization of diet. Vitamin K is a nutrient which possibly influences bone mineral density (BMD) and the risk of fracture [3]. One method of action may be regulation of osteoblast function [4] and bone mineral maturation by enhancing carboxylation of osteocalcin [5,6], the most abundant non-collagen protein in bone. Another proposed mechanism is through vitamin K2 induced activation of the steroid and xenobiotic receptor, SXR, influencing gene expressions in osteoblasts [7]. Vitamin K2 has also been shown to increase bone mineralization by a carboxylation independent mechanism [8], possibly through the SXR. Naturally occurring vitamin K-variants are phylloquinone (vitamin K1) found mainly in green vegetables and some plant oils, and multiple forms of menaquinones (vitamin K2), found in e.g. fermented food, meat and milk products. Vitamin K2 is also produced by bacteria in the gut, and can partly satisfy the amount required although not a major source of vitamin K [9].
⁎ Corresponding author at: Department of Public Health and Primary Health Care, University of Bergen, Kalfarveien 31, N-5018 Bergen, Norway. Fax: + 47 55586130. E-mail addresses: [email protected] (E.M. Apalset), [email protected] (C.G. Gjesdal), [email protected] (G.E. Eide), [email protected] (G.S. Tell). 8756-3282/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.bone.2011.07.035
There is conflicting evidence on whether dietary or supplementary vitamin K has any preventive effect on osteoporosis and fracture risk. Some studies have suggested an association between low dietary vitamin K1 and fracture risk [10,11], whereas others have not [12]. Due to better absorption and longer half-life, dietary vitamin K2 has been suggested to contribute more to the total vitamin K effect than its 10–20% of the total vitamin K intake [13,14]. Most studies in favor of vitamin K as treatment for osteoporosis and prevention of fractures have investigated vitamin K in pharmacological doses [13], and published studies on the association of dietary (non-supplemental) vitamin K2 and fracture risk are, as far as we know, non-existing. Apolipoprotein E is a protein crucial for gastrointestinal uptake and transport of fat molecules and may also play a role in uptake and transportation of the fat soluble vitamin K to bone tissue [15]. Genetic polymorphisms for the apolipoprotein E gene (APOE) are most commonly expressed as the isoforms E2, E3 and E4. Some studies have found an association between presence of the E4 allele of APOE and low bone mineral density of the hip [16,17] or increased fracture risk [18], possibly through rapid clearance of fat and phylloquinone from the circulation. Other studies have not found any influence of APOE status on BMD of the hip or risk of fracture [10,19,20]. Thus, the importance of the E4 allele regarding BMD and fracture risk is ambiguous. The reported frequency of the E4 allele is higher in the Scandinavian countries than in the south of Europe [21]. The high incidence of osteoporosis and osteoporotic fractures in Norway compared to southern European
E.M. Apalset et al. / Bone 49 (2011) 990–995
countries is thought to be caused by genetic differences. In two Danish studies, no significant association between E4 status and fractures of hip and forearm or with vertebral fractures was found [22,23]. The role of APOE status as a risk factor for osteoporotic fractures in the Scandinavian countries has, however, not been fully investigated. As previously reported [24], dietary vitamins K1 and K2 are not major determinants of BMD in the Hordaland Health Study. The aim of the present study was to expand this previous investigation to the effect of these nutrients on the risk of hip fracture. We also wanted to examine the possible influence of APOE status and the potential effect modification of the E4 allele on vitamin K in relation to hip fractures. Material and methods Study population The community-based Hordaland Health Study (HUSK) was conducted in 1997–1999 in collaboration between the National Health Screening Service (now the Norwegian Institute of Public Health), the University of Bergen, and local health services. The study cohort included here consisted of people born from 1925–27 who had previously participated in the Hordaland Homocysteine study in 1992–93. Inclusion was from April 1998 to June 1999. Of 4338 persons invited, 3341 (77.0%) participated and of these, 3019 completed a food frequency questionnaire (FFQ). A total of 212 were excluded (65 because of treatment with warfarin, a vitamin K antagonist, and 147 for energy intake extremes (less than 2.5 or above 97.5 percentiles) leaving 2807 (65.0%) subjects for analysis. APOE was previously analyzed for participants residing in the city of Bergen (2392 subjects) [25], this constituted the APOE sub-study cohort. Some 1946 took part in the osteoporosis sub-study with BMD measurements. Both APOE status and BMD was known for 1684 subjects. The Regional Committee for Medical Research Ethics approved the study protocol, as did the Norwegian Data Inspectorate. All participants signed an informed consent. Hip fractures The computerized discharge diagnoses records of the hospitals serving Hordaland County were searched for hip fractures occurring between baseline and 31 December 2009. Only hip fractures confirmed by a concurrent code of an adequate surgical procedure were included, as previously described [26]. A hip fracture was defined as the first fracture of the proximal femur occurring during the observation period. The discharge diagnoses used to classify a hip fracture were according to the International Classification of Diseases, Ninth Revision (ICD-9): 820–820.9 and Tenth Revision (ICD-10): S720-S722. Surgical treatments for each hospital stay are coded according to the Norwegian Classification of Surgical Procedures (version 2, 1989 and 3, 1995) and later the NOMESCO (Nordic Medico-Statistical Committee) Classification of Surgical Procedures (version 1, 1999 and version 3, 2004). Information on time of death was obtained from the Norwegian Population Register. Food frequency questionnaire A FFQ was used to measure usual intake of 169 food items and vitamin and mineral supplements over the past year. The FFQ was developed at the Department of Nutrition at the University of Oslo and previously validated [27–29]. The FFQ offered alternatives for frequency, number of units eaten and portion sizes. Intakes of vitamin K and other nutrients were computed from a food database developed at the Department of Nutrition, University of Oslo, as described previously [24]. No distinctions were made between the different menaquinones, thus vitamin K2 has been studied as one entity. At the time of the study, the most commonly used supplements in Norway did not contain
991
vitamin K. Thus, vitamin K intake in the present report is from diet only. Vitamin D supplements were, however, common, as part of multivitamins or calcium supplements, or through intake of cod liver oil. Vitamin D and calcium intakes are calculated from dietary intake and supplements combined. Additional data Participants filled out a self-administered questionnaire providing information on health status, use of medications, sociodemographic data and lifestyle factors. Smokers were categorized as never smokers or ex/current smokers. For women estrogen use was categorized as current or no use. Height and weight were measured in light clothing and body mass index (BMI; in kg/m 2) calculated. Apolipoprotein E A subgroup consisting of residents in Bergen City was genotyped for APOE during the 1992–93 survey. Details of the collection of blood samples have been described previously [30]. Genotyping for APOE was carried out by a standard polymerase chain reaction method, and the APOE genotypes were in Hardy–Weinberg equilibrium [25]. The APOE4 allele was regarded as the risk allele and subjects with one or two E4 alleles were considered E4-carriers. In some studies, APOE2E4 carriers have been excluded [20,31], as opposing effects of the E2 and E4 allele have been reported for other conditions. We did the analyses both including and excluding APOE2E4 carriers, and results including these carriers are reported. Bone mineral density A stationary, dual X-ray densitometer (EXPERT-XL; Lunar Company Inc, Madison, Wis., software version 1.72 and 1.90) was used for BMD measurements during 1998–2000. The left hip was scanned except in those with a left hip prosthesis or former hip fracture. A right hip scan was used in 3.8%. Calibration of the scanner was performed every day through measurements against a standard calibration block. The coefficient of variation for total hip BMD measurements based on duplicate measurement in 27 individuals was 1.2%, and daily phantom measurements showed no drift. Several areas of the hip were measured. Here we report the findings for BMD in total hip. Statistics Follow-up time was from enrolment until first hip fracture, death or 31 December 2009. Cox proportional hazards regression model [32] was used to study the associations between intake of vitamin K and hip fracture. Intakes of vitamins K1 and K2 were evaluated as continuous variables and adjusted for additional covariates in different model specifications. Further, vitamins K1 and K2 intakes were categorized into quartiles separately for men and women and the results reported as hazard ratios (HR) and 95% confidence intervals (CI). The analyses were performed for women and men combined. A Kaplan–Meier plot was made for cumulative survival of men and women combined, stratified on quartiles of vitamin K1 intake. A possible trend between quartiles was tested by the log-rank test. The results of analyses with total hip BMD are reported, but all analyses were also performed with BMD of the femoral neck. Risk of hip fracture according to E4-allele status (E4 carriers vs. noncarriers) was investigated by Cox proportional hazards regression models for women and men combined. Analyses were done both including and excluding subjects with E2E4 gene status. Possible effect modification of APOE-status on the relation between vitamin K1 intake and risk of hip fracture was investigated in a semi-proportional hazards model [33].
992
E.M. Apalset et al. / Bone 49 (2011) 990–995
models, possibly because of lack of statistical power. Additional adjustment for BMD in addition to sex, total energy intake, BMI, smoking, calcium- and vitamin D intake, did not change the results materially. The results were similar whether BMD of the total hip or femoral neck was used. There was no significant relation between vitamin K2 intake and subsequent hip fractures (Table 2). Table 3 shows that the hazard for hip fractures was higher in the lowest compared to the highest quartile of vitamin K1 intake, also after adjusting for sex, total energy intake, BMI, smoking and calcium and vitamin D intakes. No significant association was found between vitamin K2 intake and risk of hip fracture. Cumulative survival according to baseline quartile intake of vitamin K1 is shown in Fig. 1, with highest occurrence of fractures in the lowest quartile, and a significant trend across quartiles.
SPSS version 15.0 (SPSS inc., Chicago, Illinois) was used for all statistical analyses. Results Subject characteristics Baseline characteristics of the study population are shown in Table 1. The fracture group had lower BMI and lower BMD compared to the non-fracture group of the same sex. The proportion of smokers was larger in the fracture group in both sexes. A higher proportion of women who later sustained a hip fracture had regular physical activity than those not sustaining a fracture. The level of physical activity was quite similar between fracture and non-fracture groups among men. The total number of hip fractures was 225. The proportions of subjects who were heterozygote and homozygote for APOE4 were similar among women and men. During the study period 306 women and 413 men died. Personyears of follow-up were 27,108 for the whole cohort. 61.6% of the study population had a total daily intake of vitamin K equal to or above the recommended intake in Norway of 1 μg/kg body weight.
APOE genotypes APOE genotypes were classified into two categories: APOE4 noncarriers (E2E2, E2E3 and E3E3, n = 1652) and APOE4 carriers (E2E4, E3E4, E4E4, n = 740). E4 allele frequency (the proportion with presence of one or two E4 alleles) was 16.9%. There were no major differences between the groups regarding BMD, vitamin K1 or K2 intake. Adjusting for presence of E4 did not change the estimates for vitamin K1 in Table 2 materially, although the statistical significance of the associations was weakened, possibly due to loss of statistical power. Of the 1652 participants with no E4, 119 (7.2%) sustained a hip fracture. Of the 740 who were hetero- or homozygotes for E4, 67 (9.1%) had a hip fracture during follow-up. The hazard ratio for hip fractures among E4 carriers vs. non-carriers was 1.30 (95% CI: 0.96 to 1.75). Excluding subjects with E2E4-allele in analyses including the APOE gene status did not change the results materially. Analyzing the association of dietary vitamin K1 with risk of hip fracture by adjusting for the presence of the E4-allele either in a proportional hazards model [32] or by stratifying on E4 in a semi-
Vitamin K intake and hip fracture As shown in Table 2, there was a negative association between baseline vitamin K1 intake and subsequent risk of hip fracture. HR was 0.97 (95% CI: 0.95 to 0.99) for each 10 μg/d increase in vitamin K1 intake with no major change after additional adjustment for sex, total energy intake, BMI, smoking and intakes of vitamin D and calcium. Analyses of the association between total vitamin K intake and hip fractures showed similar results as for vitamin K1 with HR 0.97 (95% CI: 0.95 to 0.99). The association between vitamin K1 intake and risk of hip fracture was similar for the BMD sub-cohort as for the whole cohort (Table 2), but the results were not significant in the adjusted
Table 1 Baseline characteristics of the Hordaland Health Study cohort (n = 2807). Characteristics
Age at inclusion in years, mean (SD) Follow-up time in years, median (range) BMD total hip in g/cm2, mean (SD) BMI in kg/m2, mean (SD) APOE, E4-allele, n (%) No E4 Heterozygote Homozygote Smoking status, n (%) Never Current/former Alcohol consumption in g/week, median (IQR) Physical activity Regular No/irregular Estrogen treatment, n (%) Current user Non user Dietary intake Vitamin K1 (μg/d), median (IQR) Vitamin K2 (μg/d), median (IQR) Vitamin D (μg/d), median (IQR) Calcium (mg/d), median (IQR) Total energy (kJ/d), median (IQR) Fat (g/d), median (IQR)
Women (n = 1569)
Men (n = 1238)
No hip fracture
Hip fracture
No hip fracture
Hip fracture
n
n
n
n
1417 1417 947 1414 1186
72.4 (0.85) 10.9 (11.5) 0.81 (0.12) 26.4 (4.4)
152 152 93 126
854 (72.0) 299 (25.2) 33 (2.8) 1417
1117 1302
72.5 (0.89) 7.0 (11.3) 0.74 (0.11) 24.8 (4.0) 84 (66.7) 37 (29.4) 5 (4.0)
152 872 (61.5) 545 (38.5) 0.0 (1.4)
152 140
572 (43.9) 730 (56.1) 1152
1165 1165 855 1163 967
1165 86 (56.6) 66 (43.3) 0.0 (1.1)
1165 1130
72.5 6.4 0.87 25.5
(0.90) (10.8) (0.12) (3.3)
35 (61.4) 19 (33.3) 3 (5.3) 73
279 (23.9) 886 (76.1) 2.0 (7.6)
73 69
14 (19.2) 59 (80.8) 1.4 (5.2) 21 (30.4) 48 (69.6) NA
20 (16.4) 102 (83.6) 152
67.0 (66.6) 10.8 (7.4) 5.5 (7.1) 663 (360) 6361 (2548) 51.5 (29.1)
73 73 51 73 57
336 (29.7) 794 (70.3) NA
122
1417
(0.86) (11.6) (0.15) (3.2)
679 (70.2) 260 (22.3) 28 (2.9)
69 (49.3) 71 (50.7)
165 (11.6) 987 (85.7)
72.4 10.8 0.97 26.0
1165 57.9 (64.3) 10.2 (7.2) 4.7 (7.3) 685 (374) 6183 (3165) 47.7 (30.6)
73 78.4 (61.7) 11.9 (7.6) 8.4 (10.1) 723 (385) 8289 (2884) 68.8 (30.0)
Intakes of calcium and vitamin D include use of supplements. Abbreviations: n: number; SD: standard deviation; BMD: bone mineral density; BMI: body mass index; NA: not applicable; IQR: interquartile range. Follow-up time: time to hip fracture in fracture groups, time to death or end of study in groups with no hip fracture.
65.2 12.6 8.3 691 8252 72.6
(46.1) (8.6) (8.5) (409) (2924) (30.1)
E.M. Apalset et al. / Bone 49 (2011) 990–995
993
Table 2 Cox proportional hazards regression of time to hip fracture in 10 years follow-up of 2807 participants in the Hordaland Health Study, Bergen, Norway 1997–1999. Cohort
Vitamin
All All APOE group (n = 2392) BMD group (n = 1946)
Model 1a
Model 2b
Model 3c
HR
95% CI
p-valued
HR
95% CI
p-valued
HR
95% CI
p-valued
K1 per 10 μg K2 per 1 μg K1 per 10 μg
0.97 0.99 0.98
(0.95, 0.99) (0.97, 1.01) (0.95, 1.00)
0.005 0.399 0.026
0.98 1.01 0.98
(0.95, 1.00) (0.98, 1.03) (0.95, 1.00)
0.016 0.745 0.030
0.98 1.00 0.98
(0.95, 1.00) (0.97, 1.03) (0.96, 1.00)
0.030 0.930 0.030
K1 per 10 μg
0.98
(0.95, 1.00)
0.039
0.98
(0.95, 1.00)
0.084
0.98
(0.95, 1.00)
0.075
Abbreviations: HR: hazard ratio; CI: confidence interval; APOE: apolipoprotein E; BMD: bone mineral density. a Unadjusted. b Adjusted for sex and total energy intake. c Adjusted for sex, total energy intake, smoking, body mass index (BMI), vitamin D and calcium intake in addition to presence of the E4-allele in the APOE group and for BMD in the BMD group. d From likelihood-ratio test.
proportional hazards model [33] did not show any moderating (interaction) or confounding effect of E4. Cox regression analyses stratified by sex did not show any statistically significant association of APOE with risk of hip fracture. This did not change with additional adjustment for use of estrogen in women (data not shown). Discussion In this study we found an inverse association between intake of vitamin K1 and risk of hip fracture in a cohort of older men and women followed for 10 years. Dietary vitamin K2 was not associated with risk of hip fracture. Further, the presence of the APOE4 allele did not increase the risk of hip fracture, nor was there any effect modification of APOE4 in the relation between vitamin K1 and hip fracture. There has been a long-lasting discussion regarding the influence of vitamin K intake on BMD and risk of hip fractures. While the evidence of an association with BMD is inconsistent [3], several epidemiologic studies have found an association with risk of fracture [10,11]. This is consistent with findings reported earlier from this cohort [24], where we did not find vitamin K1 or K2 to be major determinants of BMD. An effect of vitamin K on risk of hip fracture has not yet been established in randomized controlled trials (RCT). While the ECKO-trial [34] found a small reduction in all clinical fractures among vitamin K1-treated postmenopausal women, the study was not powered to conclude on fractures as outcome. The effect of high doses of vitamin K2 on BMD has been investigated in several RCTs, but fracture as outcome is reported by only a few. No RCTs have been a priori designed to show a difference in hip fractures as the main outcome. In two different studies comparing treatment with menatetrenone (a vitamin K2) to placebo, the treatment
groups had fewer vertebral fractures during 2 years of follow-up than controls [35,36]. In a study comparing the fracture preventive effect of calcium compared to calcium and menatetrenone supplementation combined, no difference was found between the groups regarding incidence of clinical fractures (humeral, femoral, radial and high-trauma vertebral fractures) [37]. More recently, a RCT comparing treatment with vitamins K1 and K2 in pharmacological doses and placebo over 1 year, found no effect of vitamin K supplements on BMD, bone loss or bone geometry [38]. These studies do not confirm, nor undermine our findings. Perhaps vitamin K1 intake is more important in older persons with low BMD, than in younger postmenopausal women with better bone health. Even though vitamins K1 and K2 both act as a coenzyme for the vitamin K dependent carboxylase, we only found an association between risk of hip fracture and intake of vitamin K1, but not K2, possibly because of the much higher intake of vitamin K1. A high intake of vitamin K1 may be a marker of a healthy life style and optimal nutrition. Intakes of vitamin K2 in the whole HUSK cohort (including subjects born 1950–51) were mainly through consumption of milk, cheese and other dairy products (50%), meat (24%) and eggs (13%), while K1 intakes were through vegetables, fruits and berries (70%). This suggests that a high consumption of vitamin K1 may represent a healthy lifestyle, while a diet high in vitamin K2 may not. There were, however no differences in BMI or the proportion of current smokers between groups according to quartile intakes of vitamin K1 to support this (data not shown). Further, we found a higher proportion of current smokers among those with the highest quartile intake of vitamin K2. Thus, the proposed positive effect of vitamin K2 on fracture risk may be diminished by an otherwise unhealthy lifestyle, although we cannot conclude on this. There are several forms of menaquinones, each of
Table 3 Cox proportional hazards regression analyses of time to first hip fracture according to intake of vitamin K1 and K2 in quartiles (Q1–Q4) for n = 2807 subjects in the Hordaland Health Study, Bergen, Norway followed from inclusion 1998–99 to 31 December 2009. Vitamin
K1
K2
Quartiles
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
Intake, women (μg/d)
Intake, men (μg/d)
n
b 42.2 42.2–66.7 66.8–108.6 N 108.7 b 7.2 7.2–10.7 10.7–14.5 N 14.5
b 52.9 52.9–77.4 77.4–113.9 N 113.9 b 8.5 8.5–11.9 11.9–16.2 N 16.2
701 703 702 701 701 703 702 701
Model 1a
Model 2b
HR 95% CI 1.57 1.19 0.95 1.00 1.12 1.11 1.15 1.00
(1.09, (0.81, (0.63, (ref) (0.77, (0.76, (0.79, (ref)
p for trend 2.26) 1.75) 1.42)
0.006
1.63) 1.61) 1.67)
0.629
d
Model 3c
HR 95% CI 1.67 1.22 0.96 1.00 0.96 1.02 1.09 1.00
(1.09, (0.82, (0.64, (ref) (0.61, (0.68, (0.75, (ref)
p for trend 2.55) 1.82) 1.43)
0.012
1.52) 1.53) 1.61)
0.786
d
Abbreviations: HR: hazard ratio; CI: confidence interval; ref: reference group. Vitamin K1 and K2 quartile intakes are calculated separately for women and men, and the quartiles combined in the multivariate analyses. a Unadjusted. b Adjusted for sex and total energy intake. c Adjusted for sex, total energy intake, smoking, body mass index (BMI), vitamin D- and calcium intake. d From likelihood-ratio test.
p for trendd
HR 95% CI 1.63 1.21 0.96 1.00 1.03 1.09 1.19 1.00
(1.06, (0.82, (0.64, (ref) (0.64, (0.71, (0.80, (ref)
2.49) 1.80) 1.44)
0.015
1.67) 1.66) 1.75)
0.983
994
E.M. Apalset et al. / Bone 49 (2011) 990–995
Fig. 1. Kaplan–Meier unadjusted survival function for hip fractures in 2807 men and women, 70–72 years old according to quartiles (quartile 1 lowest, quartile 4 highest) of vitamin K1 intake in the Hordaland Health Study.
later sustaining a hip fracture are quite similar to those not sustaining a fracture, with regard to age at inclusion and lifestyle factors. FFQ gives information about the diet during the previous year, in most cases reflecting long time dietary habits which may also be regarded as a strength. The food frequency questionnaire which we used has been validated previously for several nutrients [27–29], but not specifically for vitamin K, which may be regarded as a weakness. The validity of our outcome measure, hip fractures, is expected to be good, relying on diagnoses from high-quality computerized hospital records and confirmation by concurrent surgical procedure codes. The study is community based, and the size of the cohort is substantial with a follow-up time of more than 10 years. The findings from our study suggest that elderly persons around 70 years with the lowest intakes of vitamin K1 have an increased risk of hip fracture. More research is warranted to establish whether a higher vitamin K1 intake after the age of 70 will lower this risk. Subjects with intake of vitamin K1 above average (in our population 92 μg/d for women and 99 μg/d for men) will probably not benefit from increased intake. Our findings do not support APOE polymorphisms as contributing to the high incidence of hip fracture in Norway, nor any effect modification of APOE status in the vitamin K1-hip fracture relation. Financial support
which may have different characteristics regarding absorption, transportation and mechanism of action. Vitamin K2 produced by bacteria in the gut, and a possible conversion of vitamin K1 to a menaquinone (MK4) in vivo [39], further complicates any attempts to separate and measure the effects of different forms of vitamin K in the diet. As shown in Fig. 1, there was no clear difference in the risk of fracture between the groups with the highest and second highest intakes of vitamin K1. This suggests that there may not be any additional effect of increased vitamin K1 intake for people who already have an intake of vitamin K above average. This is in agreement with findings from the Framingham osteoporosis study, where there was a similar shift between the two highest and two lowest quartiles of vitamin K1 intake and risk of fracture [10]. In the Nurses' Health Study cohort an increased risk of fracture was found among women only in the lowest quintile of vitamin K1 intake compared to quintiles 2–5 combined [11]. This study population was younger than in both the present study and the Framingham study, suggesting that a low vitamin K1 intake may be more deleterious in older age. The same trend with higher fracture risk in the two lowest quartiles of vitamin K1 intake was observed after adjusting for BMD, suggesting an independent role of vitamin K in maintaining bone strength. Apolipoprotein E may affect bone metabolism indirectly through transport of the fat soluble vitamin K to bone tissue [15,40]. Theoretically, there may be a connection between the presence of the APOE4-allele, low intake of vitamin K and increased risk of hip fracture. This possible gene– nutrient interaction has been studied earlier in relation to bone loss, BMD and fracture risk, but no interaction was found [10,31]. Furthermore, a meta-analysis did not find any association between APOE gene polymorphisms and bone mineral density of fracture risk [20]. We did not find any effect modification of APOE-status in the vitamin K1-hip fracture relation, further undermining the theory that apolipoprotein E may influence bone health through vitamin K transport and uptake. One limitation of this study is that information about vitamin K intake is available only at baseline. Information on potential dietary changes during the follow-up period might have given a more precise estimate of the connection between vitamin K intake and hip fractures. People at this age have, however, usually adopted dietary habits which are not likely to change as long as they continue their community-dwelling. Another potential limitation of this study is the use of FFQ, which is dependent on the study subjects' ability to remember and estimate his or her food intake. In this study we have analyzed the intake of vitamin K and followed the participants prospectively. As shown in Table 1, the subjects
The data collection of the Hordaland Health Study was partly funded by the Research Council of Norway. Acknowledgments Christian A. Drevon and Anne-Marthe W. Johansen, Department of Nutrition, University of Oslo. References [1] Abrahamsen B, van Staa T, Ariely R, Olson M, Cooper C. Excess mortality following hip fracture: a systematic epidemiological review. Osteoporos Int 2009;20:1633–50. [2] Pasco JA, Sanders KM, Hoekstra FM, Henry MJ, Nicholson GC, Kotowicz MA. The human cost of fracture. Osteoporos Int 2005;16:2046–52. [3] Booth SL. Roles for vitamin K beyond coagulation. Annu Rev Nutr 2009;29:89–110. [4] Ducy P, Desbois C, Boyce B, Pinero G, Story B, Dunstan C, et al. Increased bone formation in osteocalcin-deficient mice. Nature 1996;382:448–52. [5] Boskey AL, Gadaleta S, Gundberg C, Doty SB, Ducy P, Karsenty G. Fourier transform infrared microspectroscopic analysis of bones of osteocalcin-deficient mice provides insight into the function of osteocalcin. Bone 1998;23:187–96. [6] Binkley NC, Krueger DC, Kawahara TN, Engelke JA, Chappell RJ, Suttie JW. A high phylloquinone intake is required to achieve maximal osteocalcin gammacarboxylation. Am J Clin Nutr 2002;76:1055–60. [7] Tabb MM, Sun A, Zhou C, Grun F, Errandi J, Romero K, et al. Vitamin K2 regulation of bone homeostasis is mediated by the steroid and xenobiotic receptor SXR. J Biol Chem 2003;278:43919–27. [8] Atkins GJ, Welldon KJ, Wijenayaka AR, Bonewald LF, Findlay DM. Vitamin K promotes mineralization, osteoblast-to-osteocyte transition, and an anticatabolic phenotype by {gamma}-carboxylation-dependent and -independent mechanisms. Am J Physiol Cell Physiol 2009;297:C1358–67. [9] Suttie JW. The importance of menaquinones in human nutrition. Annu Rev Nutr 1995;15:399–417. [10] Booth SL, Tucker KL, Chen H, Hannan MT, Gagnon DR, Cupples LA, et al. Dietary vitamin K intakes are associated with hip fracture but not with bone mineral density in elderly men and women. Am J Clin Nutr 2000;71:1201–8. [11] Feskanich D, Weber P, Willett WC, Rockett H, Booth SL, Colditz GA. Vitamin K intake and hip fractures in women: a prospective study. Am J Clin Nutr 1999;69:74–9. [12] Rejnmark L, Vestergaard P, Charles P, Hermann AP, Brot C, Eiken P, et al. No effect of vitamin K1 intake on bone mineral density and fracture risk in perimenopausal women. Osteoporos Int 2006;17:1122–32. [13] Iwamoto J, Sato Y, Takeda T, Matsumoto H. High-dose vitamin K supplementation reduces fracture incidence in postmenopausal women: a review of the literature. Nutr Res 2009;29:221–8. [14] Schurgers LJ, Vermeer C. Determination of phylloquinone and menaquinones in food. Effect of food matrix on circulating vitamin K concentrations. Haemostasis 2000;30:298–307. [15] Newman P, Bonello F, Wierzbicki AS, Lumb P, Savidge GF, Shearer MJ. The uptake of lipoprotein-borne phylloquinone (vitamin K1) by osteoblasts and osteoblastlike cells: role of heparan sulfate proteoglycans and apolipoprotein E. J Bone Miner Res 2002;17:426–33.
E.M. Apalset et al. / Bone 49 (2011) 990–995 [16] Pluijm SM, Dik MG, Jonker C, Deeg DJ, van Kamp GJ, Lips P. Effects of gender and age on the association of apolipoprotein E epsilon4 with bone mineral density, bone turnover and the risk of fractures in older people. Osteoporos Int 2002;13:701–9. [17] Cauley JA, Zmuda JM, Yaffe K, Kuller LH, Ferrell RE, Wisniewski SR, et al. Apolipoprotein E polymorphism: a new genetic marker of hip fracture risk—the study of osteoporotic fractures. J Bone Miner Res 1999;14:1175–81. [18] Kohlmeier M, Saupe J, Schaefer K, Asmus G. Bone fracture history and prospective bone fracture risk of hemodialysis patients are related to apolipoprotein E genotype. Calcif Tissue Int 1998;62:278–81. [19] von Muhlen DG, Barrett-Connor E, Schneider DL, Morin PA, Parry P. Osteoporosis and apolipoprotein E genotype in older adults: the Rancho Bernardo study. Osteoporos Int 2001;12:332–5. [20] Peter I, Crosier MD, Yoshida M, Booth SL, Cupples LA, Dawson-Hughes B, et al. Associations of APOE gene polymorphisms with bone mineral density and fracture risk: a meta-analysis. Osteoporos Int 2011;22:1199–209. [21] Gerdes LU, Klausen IC, Sihm I, Faergeman O. Apolipoprotein E polymorphism in a Danish population compared to findings in 45 other study populations around the world. Genet Epidemiol 1992;9:155–67. [22] Bagger YZ, Rasmussen HB, Alexandersen P, Werge T, Christiansen C, Tanko LB. Links between cardiovascular disease and osteoporosis in postmenopausal women: serum lipids or atherosclerosis per se? Osteoporos Int 2007;18:505–12. [23] Sennels HP, Sand JC, Madsen B, Lauritzen JB, Fenger M, Jorgensen HL. Association between polymorphisms of apolipoprotein E, bone mineral density of the lower forearm, quantitative ultrasound of the calcaneus and osteoporotic fractures in postmenopausal women with hip or lower forearm fracture. Scand J Clin Lab Invest 2003;63:247–58. [24] Apalset EM, Gjesdal CG, Eide GE, Johansen A-MW, Drevon CA, Tell GS. Dietary vitamins K1, K2 and bone mineral density: the Hordaland Health Study. Arch Osteoporos 2010, doi:10.1007/s11657-010-0036-6. [25] Lehmann DJ, Refsum H, Nurk E, Warden DR, Tell GS, Vollset SE, et al. Apolipoprotein E epsilon4 and impaired episodic memory in communitydwelling elderly people: a marked sex difference. The Hordaland Health Study. J Neurol Neurosurg Psychiatry 2006;77:902–8. [26] Gjesdal CG, Vollset SE, Ueland PM, Refsum H, Meyer HE, Tell GS. Plasma homocysteine, folate, and vitamin B 12 and the risk of hip fracture: the Hordaland Homocysteine Study. J Bone Miner Res 2007;22:747–56. [27] Nes M, Frost Andersen L, Solvoll K, Sandstad B, Hustvedt BE, Lovo A, et al. Accuracy of a quantitative food frequency questionnaire applied in elderly Norwegian women. Eur J Clin Nutr 1992;46:809–21.
995
[28] Andersen LF, Solvoll K, Johansson LR, Salminen I, Aro A, Drevon CA. Evaluation of a food frequency questionnaire with weighed records, fatty acids, and alphatocopherol in adipose tissue and serum. Am J Epidemiol 1999;150:75–87. [29] Andersen LF, Veierod MB, Johansson L, Sakhi A, Solvoll K, Drevon CA. Evaluation of three dietary assessment methods and serum biomarkers as measures of fruit and vegetable intake, using the method of triads. Br J Nutr 2005;93:519–27. [30] Nurk E, Refsum H, Tell GS, Engedal K, Vollset SE, Ueland PM, et al. Plasma total homocysteine and memory in the elderly: the Hordaland Homocysteine Study. Ann Neurol 2005;58:847–57. [31] Macdonald HM, McGuigan FE, Lanham-New SA, Fraser WD, Ralston SH, Reid DM. Vitamin K1 intake is associated with higher bone mineral density and reduced bone resorption in early postmenopausal Scottish women: no evidence of gene–nutrient interaction with apolipoprotein E polymorphisms. Am J Clin Nutr 2008;87:1513–20. [32] Cox DR. Regression models and life-tables. J R Stat Soc B 1972;34:187–220. [33] Eide GE, Omenaas E, Gulsvik A. The semi-proportional hazards model revisited: practical reparametrizations. Stat Med 1996;15:1771–7. [34] Cheung AM, Tile L, Lee Y, Tomlinson G, Hawker G, Scher J, et al. Vitamin K supplementation in postmenopausal women with osteopenia (ECKO trial): a randomized controlled trial. PLoS Med 2008;5:e196. [35] Iwamoto J, Takeda T, Ichimura S. Effect of menatetrenone on bone mineral density and incidence of vertebral fractures in postmenopausal women with osteoporosis: a comparison with the effect of etidronate. J Orthop Sci 2001;6:487–92. [36] Shiraki M, Shiraki Y, Aoki C, Miura M. Vitamin K2 (menatetrenone) effectively prevents fractures and sustains lumbar bone mineral density in osteoporosis. J Bone Miner Res 2000;15:515–21. [37] Inoue T, Fujita T, Kishimoto H, Makino T, Nakamura T, Sato T, et al. Randomized controlled study on the prevention of osteoporotic fractures (OF study): a phase IV clinical study of 15-mg menatetrenone capsules. J Bone Miner Metab 2009;27: 66–75. [38] Binkley N, Harke J, Krueger D, Engelke J, Vallarta-Ast N, Gemar D, et al. Vitamin K treatment reduces undercarboxylated osteocalcin but does not alter bone turnover, density, or geometry in healthy postmenopausal North American women. J Bone Miner Res 2009;24:983–91. [39] Ronden JE, Drittij-Reijnders MJ, Vermeer C, Thijssen HH. Intestinal flora is not an intermediate in the phylloquinone-menaquinone-4 conversion in the rat. Biochim Biophys Acta 1998;1379:69–75. [40] Kohlmeier M, Salomon A, Saupe J, Shearer MJ. Transport of vitamin K to bone in humans. J Nutr 1996;126:1192S–6S.
Contents lists available at ScienceDirect
Bone j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / b o n e
Original Full Length Article
Intake of vitamin K1 and K2 and risk of hip fractures: The Hordaland Health Study Ellen M. Apalset a, b,⁎, Clara G. Gjesdal b, c, Geir E. Eide a, d, Grethe S. Tell a a
Department of Public Health and Primary Health Care, University of Bergen, Kalfarveien 31, N-5018 Bergen, Norway Department of Rheumatology, Haukeland University Hospital, N-5021 Bergen, Norway c Section of Rheumatology, Institute of Medicine, University of Bergen, Bergen, Norway d Centre for Clinical Research, Haukeland University Hospital, N-5021 Bergen, Norway b
a r t i c l e
i n f o
Article history: Received 26 April 2011 Revised 5 July 2011 Accepted 23 July 2011 Available online 2 August 2011 Edited by: Robert Recker Keywords: Hip fracture Osteoporosis Apolipoprotein E Diet Vitamin K1 Vitamin K2
a b s t r a c t Background: Evidence of the effect of vitamin K on bone health is conflicting. The aim was to investigate the association between intake of vitamins K1 and K2 and subsequent risk of hip fracture in a general population sample, as well as potential effect modification by apolipoprotein E gene (APOE) status by presence of the E4 allele. Methods: 1238 men and 1569 women 71–75 years of age were included in the community-based Hordaland Health Study 1997–1999 in Western Norway. Information on hip fracture was obtained from hospitalizations in the region from enrolment until 31 December 2009. Information on intake of vitamins K1 and K2 collected at baseline was used as potential predictors of hip fracture in Cox proportional hazards regression analyses. Results: Participants in the lowest compared to the highest quartile of vitamin K1 intake had increased risk of suffering a hip fracture (hazard ratio (HR)= 1.57 [95% CI 1.09, 2.26]). Vitamin K2 intake was not associated with hip fracture. Presence of APOE4-allele did not increase the risk of hip fracture, nor was there any effect modification with vitamin K1 in relation to risk of hip fracture. Conclusions: A low intake of vitamin K1, but not K2, was associated with an increased risk of hip fractures. © 2011 Elsevier Inc. All rights reserved.
Introduction Hip fractures have serious consequences, leading to impaired health, disability and increased mortality [1,2]. Possible preventive strategies include change in health behaviors and optimization of diet. Vitamin K is a nutrient which possibly influences bone mineral density (BMD) and the risk of fracture [3]. One method of action may be regulation of osteoblast function [4] and bone mineral maturation by enhancing carboxylation of osteocalcin [5,6], the most abundant non-collagen protein in bone. Another proposed mechanism is through vitamin K2 induced activation of the steroid and xenobiotic receptor, SXR, influencing gene expressions in osteoblasts [7]. Vitamin K2 has also been shown to increase bone mineralization by a carboxylation independent mechanism [8], possibly through the SXR. Naturally occurring vitamin K-variants are phylloquinone (vitamin K1) found mainly in green vegetables and some plant oils, and multiple forms of menaquinones (vitamin K2), found in e.g. fermented food, meat and milk products. Vitamin K2 is also produced by bacteria in the gut, and can partly satisfy the amount required although not a major source of vitamin K [9].
⁎ Corresponding author at: Department of Public Health and Primary Health Care, University of Bergen, Kalfarveien 31, N-5018 Bergen, Norway. Fax: + 47 55586130. E-mail addresses: [email protected] (E.M. Apalset), [email protected] (C.G. Gjesdal), [email protected] (G.E. Eide), [email protected] (G.S. Tell). 8756-3282/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.bone.2011.07.035
There is conflicting evidence on whether dietary or supplementary vitamin K has any preventive effect on osteoporosis and fracture risk. Some studies have suggested an association between low dietary vitamin K1 and fracture risk [10,11], whereas others have not [12]. Due to better absorption and longer half-life, dietary vitamin K2 has been suggested to contribute more to the total vitamin K effect than its 10–20% of the total vitamin K intake [13,14]. Most studies in favor of vitamin K as treatment for osteoporosis and prevention of fractures have investigated vitamin K in pharmacological doses [13], and published studies on the association of dietary (non-supplemental) vitamin K2 and fracture risk are, as far as we know, non-existing. Apolipoprotein E is a protein crucial for gastrointestinal uptake and transport of fat molecules and may also play a role in uptake and transportation of the fat soluble vitamin K to bone tissue [15]. Genetic polymorphisms for the apolipoprotein E gene (APOE) are most commonly expressed as the isoforms E2, E3 and E4. Some studies have found an association between presence of the E4 allele of APOE and low bone mineral density of the hip [16,17] or increased fracture risk [18], possibly through rapid clearance of fat and phylloquinone from the circulation. Other studies have not found any influence of APOE status on BMD of the hip or risk of fracture [10,19,20]. Thus, the importance of the E4 allele regarding BMD and fracture risk is ambiguous. The reported frequency of the E4 allele is higher in the Scandinavian countries than in the south of Europe [21]. The high incidence of osteoporosis and osteoporotic fractures in Norway compared to southern European
E.M. Apalset et al. / Bone 49 (2011) 990–995
countries is thought to be caused by genetic differences. In two Danish studies, no significant association between E4 status and fractures of hip and forearm or with vertebral fractures was found [22,23]. The role of APOE status as a risk factor for osteoporotic fractures in the Scandinavian countries has, however, not been fully investigated. As previously reported [24], dietary vitamins K1 and K2 are not major determinants of BMD in the Hordaland Health Study. The aim of the present study was to expand this previous investigation to the effect of these nutrients on the risk of hip fracture. We also wanted to examine the possible influence of APOE status and the potential effect modification of the E4 allele on vitamin K in relation to hip fractures. Material and methods Study population The community-based Hordaland Health Study (HUSK) was conducted in 1997–1999 in collaboration between the National Health Screening Service (now the Norwegian Institute of Public Health), the University of Bergen, and local health services. The study cohort included here consisted of people born from 1925–27 who had previously participated in the Hordaland Homocysteine study in 1992–93. Inclusion was from April 1998 to June 1999. Of 4338 persons invited, 3341 (77.0%) participated and of these, 3019 completed a food frequency questionnaire (FFQ). A total of 212 were excluded (65 because of treatment with warfarin, a vitamin K antagonist, and 147 for energy intake extremes (less than 2.5 or above 97.5 percentiles) leaving 2807 (65.0%) subjects for analysis. APOE was previously analyzed for participants residing in the city of Bergen (2392 subjects) [25], this constituted the APOE sub-study cohort. Some 1946 took part in the osteoporosis sub-study with BMD measurements. Both APOE status and BMD was known for 1684 subjects. The Regional Committee for Medical Research Ethics approved the study protocol, as did the Norwegian Data Inspectorate. All participants signed an informed consent. Hip fractures The computerized discharge diagnoses records of the hospitals serving Hordaland County were searched for hip fractures occurring between baseline and 31 December 2009. Only hip fractures confirmed by a concurrent code of an adequate surgical procedure were included, as previously described [26]. A hip fracture was defined as the first fracture of the proximal femur occurring during the observation period. The discharge diagnoses used to classify a hip fracture were according to the International Classification of Diseases, Ninth Revision (ICD-9): 820–820.9 and Tenth Revision (ICD-10): S720-S722. Surgical treatments for each hospital stay are coded according to the Norwegian Classification of Surgical Procedures (version 2, 1989 and 3, 1995) and later the NOMESCO (Nordic Medico-Statistical Committee) Classification of Surgical Procedures (version 1, 1999 and version 3, 2004). Information on time of death was obtained from the Norwegian Population Register. Food frequency questionnaire A FFQ was used to measure usual intake of 169 food items and vitamin and mineral supplements over the past year. The FFQ was developed at the Department of Nutrition at the University of Oslo and previously validated [27–29]. The FFQ offered alternatives for frequency, number of units eaten and portion sizes. Intakes of vitamin K and other nutrients were computed from a food database developed at the Department of Nutrition, University of Oslo, as described previously [24]. No distinctions were made between the different menaquinones, thus vitamin K2 has been studied as one entity. At the time of the study, the most commonly used supplements in Norway did not contain
991
vitamin K. Thus, vitamin K intake in the present report is from diet only. Vitamin D supplements were, however, common, as part of multivitamins or calcium supplements, or through intake of cod liver oil. Vitamin D and calcium intakes are calculated from dietary intake and supplements combined. Additional data Participants filled out a self-administered questionnaire providing information on health status, use of medications, sociodemographic data and lifestyle factors. Smokers were categorized as never smokers or ex/current smokers. For women estrogen use was categorized as current or no use. Height and weight were measured in light clothing and body mass index (BMI; in kg/m 2) calculated. Apolipoprotein E A subgroup consisting of residents in Bergen City was genotyped for APOE during the 1992–93 survey. Details of the collection of blood samples have been described previously [30]. Genotyping for APOE was carried out by a standard polymerase chain reaction method, and the APOE genotypes were in Hardy–Weinberg equilibrium [25]. The APOE4 allele was regarded as the risk allele and subjects with one or two E4 alleles were considered E4-carriers. In some studies, APOE2E4 carriers have been excluded [20,31], as opposing effects of the E2 and E4 allele have been reported for other conditions. We did the analyses both including and excluding APOE2E4 carriers, and results including these carriers are reported. Bone mineral density A stationary, dual X-ray densitometer (EXPERT-XL; Lunar Company Inc, Madison, Wis., software version 1.72 and 1.90) was used for BMD measurements during 1998–2000. The left hip was scanned except in those with a left hip prosthesis or former hip fracture. A right hip scan was used in 3.8%. Calibration of the scanner was performed every day through measurements against a standard calibration block. The coefficient of variation for total hip BMD measurements based on duplicate measurement in 27 individuals was 1.2%, and daily phantom measurements showed no drift. Several areas of the hip were measured. Here we report the findings for BMD in total hip. Statistics Follow-up time was from enrolment until first hip fracture, death or 31 December 2009. Cox proportional hazards regression model [32] was used to study the associations between intake of vitamin K and hip fracture. Intakes of vitamins K1 and K2 were evaluated as continuous variables and adjusted for additional covariates in different model specifications. Further, vitamins K1 and K2 intakes were categorized into quartiles separately for men and women and the results reported as hazard ratios (HR) and 95% confidence intervals (CI). The analyses were performed for women and men combined. A Kaplan–Meier plot was made for cumulative survival of men and women combined, stratified on quartiles of vitamin K1 intake. A possible trend between quartiles was tested by the log-rank test. The results of analyses with total hip BMD are reported, but all analyses were also performed with BMD of the femoral neck. Risk of hip fracture according to E4-allele status (E4 carriers vs. noncarriers) was investigated by Cox proportional hazards regression models for women and men combined. Analyses were done both including and excluding subjects with E2E4 gene status. Possible effect modification of APOE-status on the relation between vitamin K1 intake and risk of hip fracture was investigated in a semi-proportional hazards model [33].
992
E.M. Apalset et al. / Bone 49 (2011) 990–995
models, possibly because of lack of statistical power. Additional adjustment for BMD in addition to sex, total energy intake, BMI, smoking, calcium- and vitamin D intake, did not change the results materially. The results were similar whether BMD of the total hip or femoral neck was used. There was no significant relation between vitamin K2 intake and subsequent hip fractures (Table 2). Table 3 shows that the hazard for hip fractures was higher in the lowest compared to the highest quartile of vitamin K1 intake, also after adjusting for sex, total energy intake, BMI, smoking and calcium and vitamin D intakes. No significant association was found between vitamin K2 intake and risk of hip fracture. Cumulative survival according to baseline quartile intake of vitamin K1 is shown in Fig. 1, with highest occurrence of fractures in the lowest quartile, and a significant trend across quartiles.
SPSS version 15.0 (SPSS inc., Chicago, Illinois) was used for all statistical analyses. Results Subject characteristics Baseline characteristics of the study population are shown in Table 1. The fracture group had lower BMI and lower BMD compared to the non-fracture group of the same sex. The proportion of smokers was larger in the fracture group in both sexes. A higher proportion of women who later sustained a hip fracture had regular physical activity than those not sustaining a fracture. The level of physical activity was quite similar between fracture and non-fracture groups among men. The total number of hip fractures was 225. The proportions of subjects who were heterozygote and homozygote for APOE4 were similar among women and men. During the study period 306 women and 413 men died. Personyears of follow-up were 27,108 for the whole cohort. 61.6% of the study population had a total daily intake of vitamin K equal to or above the recommended intake in Norway of 1 μg/kg body weight.
APOE genotypes APOE genotypes were classified into two categories: APOE4 noncarriers (E2E2, E2E3 and E3E3, n = 1652) and APOE4 carriers (E2E4, E3E4, E4E4, n = 740). E4 allele frequency (the proportion with presence of one or two E4 alleles) was 16.9%. There were no major differences between the groups regarding BMD, vitamin K1 or K2 intake. Adjusting for presence of E4 did not change the estimates for vitamin K1 in Table 2 materially, although the statistical significance of the associations was weakened, possibly due to loss of statistical power. Of the 1652 participants with no E4, 119 (7.2%) sustained a hip fracture. Of the 740 who were hetero- or homozygotes for E4, 67 (9.1%) had a hip fracture during follow-up. The hazard ratio for hip fractures among E4 carriers vs. non-carriers was 1.30 (95% CI: 0.96 to 1.75). Excluding subjects with E2E4-allele in analyses including the APOE gene status did not change the results materially. Analyzing the association of dietary vitamin K1 with risk of hip fracture by adjusting for the presence of the E4-allele either in a proportional hazards model [32] or by stratifying on E4 in a semi-
Vitamin K intake and hip fracture As shown in Table 2, there was a negative association between baseline vitamin K1 intake and subsequent risk of hip fracture. HR was 0.97 (95% CI: 0.95 to 0.99) for each 10 μg/d increase in vitamin K1 intake with no major change after additional adjustment for sex, total energy intake, BMI, smoking and intakes of vitamin D and calcium. Analyses of the association between total vitamin K intake and hip fractures showed similar results as for vitamin K1 with HR 0.97 (95% CI: 0.95 to 0.99). The association between vitamin K1 intake and risk of hip fracture was similar for the BMD sub-cohort as for the whole cohort (Table 2), but the results were not significant in the adjusted
Table 1 Baseline characteristics of the Hordaland Health Study cohort (n = 2807). Characteristics
Age at inclusion in years, mean (SD) Follow-up time in years, median (range) BMD total hip in g/cm2, mean (SD) BMI in kg/m2, mean (SD) APOE, E4-allele, n (%) No E4 Heterozygote Homozygote Smoking status, n (%) Never Current/former Alcohol consumption in g/week, median (IQR) Physical activity Regular No/irregular Estrogen treatment, n (%) Current user Non user Dietary intake Vitamin K1 (μg/d), median (IQR) Vitamin K2 (μg/d), median (IQR) Vitamin D (μg/d), median (IQR) Calcium (mg/d), median (IQR) Total energy (kJ/d), median (IQR) Fat (g/d), median (IQR)
Women (n = 1569)
Men (n = 1238)
No hip fracture
Hip fracture
No hip fracture
Hip fracture
n
n
n
n
1417 1417 947 1414 1186
72.4 (0.85) 10.9 (11.5) 0.81 (0.12) 26.4 (4.4)
152 152 93 126
854 (72.0) 299 (25.2) 33 (2.8) 1417
1117 1302
72.5 (0.89) 7.0 (11.3) 0.74 (0.11) 24.8 (4.0) 84 (66.7) 37 (29.4) 5 (4.0)
152 872 (61.5) 545 (38.5) 0.0 (1.4)
152 140
572 (43.9) 730 (56.1) 1152
1165 1165 855 1163 967
1165 86 (56.6) 66 (43.3) 0.0 (1.1)
1165 1130
72.5 6.4 0.87 25.5
(0.90) (10.8) (0.12) (3.3)
35 (61.4) 19 (33.3) 3 (5.3) 73
279 (23.9) 886 (76.1) 2.0 (7.6)
73 69
14 (19.2) 59 (80.8) 1.4 (5.2) 21 (30.4) 48 (69.6) NA
20 (16.4) 102 (83.6) 152
67.0 (66.6) 10.8 (7.4) 5.5 (7.1) 663 (360) 6361 (2548) 51.5 (29.1)
73 73 51 73 57
336 (29.7) 794 (70.3) NA
122
1417
(0.86) (11.6) (0.15) (3.2)
679 (70.2) 260 (22.3) 28 (2.9)
69 (49.3) 71 (50.7)
165 (11.6) 987 (85.7)
72.4 10.8 0.97 26.0
1165 57.9 (64.3) 10.2 (7.2) 4.7 (7.3) 685 (374) 6183 (3165) 47.7 (30.6)
73 78.4 (61.7) 11.9 (7.6) 8.4 (10.1) 723 (385) 8289 (2884) 68.8 (30.0)
Intakes of calcium and vitamin D include use of supplements. Abbreviations: n: number; SD: standard deviation; BMD: bone mineral density; BMI: body mass index; NA: not applicable; IQR: interquartile range. Follow-up time: time to hip fracture in fracture groups, time to death or end of study in groups with no hip fracture.
65.2 12.6 8.3 691 8252 72.6
(46.1) (8.6) (8.5) (409) (2924) (30.1)
E.M. Apalset et al. / Bone 49 (2011) 990–995
993
Table 2 Cox proportional hazards regression of time to hip fracture in 10 years follow-up of 2807 participants in the Hordaland Health Study, Bergen, Norway 1997–1999. Cohort
Vitamin
All All APOE group (n = 2392) BMD group (n = 1946)
Model 1a
Model 2b
Model 3c
HR
95% CI
p-valued
HR
95% CI
p-valued
HR
95% CI
p-valued
K1 per 10 μg K2 per 1 μg K1 per 10 μg
0.97 0.99 0.98
(0.95, 0.99) (0.97, 1.01) (0.95, 1.00)
0.005 0.399 0.026
0.98 1.01 0.98
(0.95, 1.00) (0.98, 1.03) (0.95, 1.00)
0.016 0.745 0.030
0.98 1.00 0.98
(0.95, 1.00) (0.97, 1.03) (0.96, 1.00)
0.030 0.930 0.030
K1 per 10 μg
0.98
(0.95, 1.00)
0.039
0.98
(0.95, 1.00)
0.084
0.98
(0.95, 1.00)
0.075
Abbreviations: HR: hazard ratio; CI: confidence interval; APOE: apolipoprotein E; BMD: bone mineral density. a Unadjusted. b Adjusted for sex and total energy intake. c Adjusted for sex, total energy intake, smoking, body mass index (BMI), vitamin D and calcium intake in addition to presence of the E4-allele in the APOE group and for BMD in the BMD group. d From likelihood-ratio test.
proportional hazards model [33] did not show any moderating (interaction) or confounding effect of E4. Cox regression analyses stratified by sex did not show any statistically significant association of APOE with risk of hip fracture. This did not change with additional adjustment for use of estrogen in women (data not shown). Discussion In this study we found an inverse association between intake of vitamin K1 and risk of hip fracture in a cohort of older men and women followed for 10 years. Dietary vitamin K2 was not associated with risk of hip fracture. Further, the presence of the APOE4 allele did not increase the risk of hip fracture, nor was there any effect modification of APOE4 in the relation between vitamin K1 and hip fracture. There has been a long-lasting discussion regarding the influence of vitamin K intake on BMD and risk of hip fractures. While the evidence of an association with BMD is inconsistent [3], several epidemiologic studies have found an association with risk of fracture [10,11]. This is consistent with findings reported earlier from this cohort [24], where we did not find vitamin K1 or K2 to be major determinants of BMD. An effect of vitamin K on risk of hip fracture has not yet been established in randomized controlled trials (RCT). While the ECKO-trial [34] found a small reduction in all clinical fractures among vitamin K1-treated postmenopausal women, the study was not powered to conclude on fractures as outcome. The effect of high doses of vitamin K2 on BMD has been investigated in several RCTs, but fracture as outcome is reported by only a few. No RCTs have been a priori designed to show a difference in hip fractures as the main outcome. In two different studies comparing treatment with menatetrenone (a vitamin K2) to placebo, the treatment
groups had fewer vertebral fractures during 2 years of follow-up than controls [35,36]. In a study comparing the fracture preventive effect of calcium compared to calcium and menatetrenone supplementation combined, no difference was found between the groups regarding incidence of clinical fractures (humeral, femoral, radial and high-trauma vertebral fractures) [37]. More recently, a RCT comparing treatment with vitamins K1 and K2 in pharmacological doses and placebo over 1 year, found no effect of vitamin K supplements on BMD, bone loss or bone geometry [38]. These studies do not confirm, nor undermine our findings. Perhaps vitamin K1 intake is more important in older persons with low BMD, than in younger postmenopausal women with better bone health. Even though vitamins K1 and K2 both act as a coenzyme for the vitamin K dependent carboxylase, we only found an association between risk of hip fracture and intake of vitamin K1, but not K2, possibly because of the much higher intake of vitamin K1. A high intake of vitamin K1 may be a marker of a healthy life style and optimal nutrition. Intakes of vitamin K2 in the whole HUSK cohort (including subjects born 1950–51) were mainly through consumption of milk, cheese and other dairy products (50%), meat (24%) and eggs (13%), while K1 intakes were through vegetables, fruits and berries (70%). This suggests that a high consumption of vitamin K1 may represent a healthy lifestyle, while a diet high in vitamin K2 may not. There were, however no differences in BMI or the proportion of current smokers between groups according to quartile intakes of vitamin K1 to support this (data not shown). Further, we found a higher proportion of current smokers among those with the highest quartile intake of vitamin K2. Thus, the proposed positive effect of vitamin K2 on fracture risk may be diminished by an otherwise unhealthy lifestyle, although we cannot conclude on this. There are several forms of menaquinones, each of
Table 3 Cox proportional hazards regression analyses of time to first hip fracture according to intake of vitamin K1 and K2 in quartiles (Q1–Q4) for n = 2807 subjects in the Hordaland Health Study, Bergen, Norway followed from inclusion 1998–99 to 31 December 2009. Vitamin
K1
K2
Quartiles
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
Intake, women (μg/d)
Intake, men (μg/d)
n
b 42.2 42.2–66.7 66.8–108.6 N 108.7 b 7.2 7.2–10.7 10.7–14.5 N 14.5
b 52.9 52.9–77.4 77.4–113.9 N 113.9 b 8.5 8.5–11.9 11.9–16.2 N 16.2
701 703 702 701 701 703 702 701
Model 1a
Model 2b
HR 95% CI 1.57 1.19 0.95 1.00 1.12 1.11 1.15 1.00
(1.09, (0.81, (0.63, (ref) (0.77, (0.76, (0.79, (ref)
p for trend 2.26) 1.75) 1.42)
0.006
1.63) 1.61) 1.67)
0.629
d
Model 3c
HR 95% CI 1.67 1.22 0.96 1.00 0.96 1.02 1.09 1.00
(1.09, (0.82, (0.64, (ref) (0.61, (0.68, (0.75, (ref)
p for trend 2.55) 1.82) 1.43)
0.012
1.52) 1.53) 1.61)
0.786
d
Abbreviations: HR: hazard ratio; CI: confidence interval; ref: reference group. Vitamin K1 and K2 quartile intakes are calculated separately for women and men, and the quartiles combined in the multivariate analyses. a Unadjusted. b Adjusted for sex and total energy intake. c Adjusted for sex, total energy intake, smoking, body mass index (BMI), vitamin D- and calcium intake. d From likelihood-ratio test.
p for trendd
HR 95% CI 1.63 1.21 0.96 1.00 1.03 1.09 1.19 1.00
(1.06, (0.82, (0.64, (ref) (0.64, (0.71, (0.80, (ref)
2.49) 1.80) 1.44)
0.015
1.67) 1.66) 1.75)
0.983
994
E.M. Apalset et al. / Bone 49 (2011) 990–995
Fig. 1. Kaplan–Meier unadjusted survival function for hip fractures in 2807 men and women, 70–72 years old according to quartiles (quartile 1 lowest, quartile 4 highest) of vitamin K1 intake in the Hordaland Health Study.
later sustaining a hip fracture are quite similar to those not sustaining a fracture, with regard to age at inclusion and lifestyle factors. FFQ gives information about the diet during the previous year, in most cases reflecting long time dietary habits which may also be regarded as a strength. The food frequency questionnaire which we used has been validated previously for several nutrients [27–29], but not specifically for vitamin K, which may be regarded as a weakness. The validity of our outcome measure, hip fractures, is expected to be good, relying on diagnoses from high-quality computerized hospital records and confirmation by concurrent surgical procedure codes. The study is community based, and the size of the cohort is substantial with a follow-up time of more than 10 years. The findings from our study suggest that elderly persons around 70 years with the lowest intakes of vitamin K1 have an increased risk of hip fracture. More research is warranted to establish whether a higher vitamin K1 intake after the age of 70 will lower this risk. Subjects with intake of vitamin K1 above average (in our population 92 μg/d for women and 99 μg/d for men) will probably not benefit from increased intake. Our findings do not support APOE polymorphisms as contributing to the high incidence of hip fracture in Norway, nor any effect modification of APOE status in the vitamin K1-hip fracture relation. Financial support
which may have different characteristics regarding absorption, transportation and mechanism of action. Vitamin K2 produced by bacteria in the gut, and a possible conversion of vitamin K1 to a menaquinone (MK4) in vivo [39], further complicates any attempts to separate and measure the effects of different forms of vitamin K in the diet. As shown in Fig. 1, there was no clear difference in the risk of fracture between the groups with the highest and second highest intakes of vitamin K1. This suggests that there may not be any additional effect of increased vitamin K1 intake for people who already have an intake of vitamin K above average. This is in agreement with findings from the Framingham osteoporosis study, where there was a similar shift between the two highest and two lowest quartiles of vitamin K1 intake and risk of fracture [10]. In the Nurses' Health Study cohort an increased risk of fracture was found among women only in the lowest quintile of vitamin K1 intake compared to quintiles 2–5 combined [11]. This study population was younger than in both the present study and the Framingham study, suggesting that a low vitamin K1 intake may be more deleterious in older age. The same trend with higher fracture risk in the two lowest quartiles of vitamin K1 intake was observed after adjusting for BMD, suggesting an independent role of vitamin K in maintaining bone strength. Apolipoprotein E may affect bone metabolism indirectly through transport of the fat soluble vitamin K to bone tissue [15,40]. Theoretically, there may be a connection between the presence of the APOE4-allele, low intake of vitamin K and increased risk of hip fracture. This possible gene– nutrient interaction has been studied earlier in relation to bone loss, BMD and fracture risk, but no interaction was found [10,31]. Furthermore, a meta-analysis did not find any association between APOE gene polymorphisms and bone mineral density of fracture risk [20]. We did not find any effect modification of APOE-status in the vitamin K1-hip fracture relation, further undermining the theory that apolipoprotein E may influence bone health through vitamin K transport and uptake. One limitation of this study is that information about vitamin K intake is available only at baseline. Information on potential dietary changes during the follow-up period might have given a more precise estimate of the connection between vitamin K intake and hip fractures. People at this age have, however, usually adopted dietary habits which are not likely to change as long as they continue their community-dwelling. Another potential limitation of this study is the use of FFQ, which is dependent on the study subjects' ability to remember and estimate his or her food intake. In this study we have analyzed the intake of vitamin K and followed the participants prospectively. As shown in Table 1, the subjects
The data collection of the Hordaland Health Study was partly funded by the Research Council of Norway. Acknowledgments Christian A. Drevon and Anne-Marthe W. Johansen, Department of Nutrition, University of Oslo. References [1] Abrahamsen B, van Staa T, Ariely R, Olson M, Cooper C. Excess mortality following hip fracture: a systematic epidemiological review. Osteoporos Int 2009;20:1633–50. [2] Pasco JA, Sanders KM, Hoekstra FM, Henry MJ, Nicholson GC, Kotowicz MA. The human cost of fracture. Osteoporos Int 2005;16:2046–52. [3] Booth SL. Roles for vitamin K beyond coagulation. Annu Rev Nutr 2009;29:89–110. [4] Ducy P, Desbois C, Boyce B, Pinero G, Story B, Dunstan C, et al. Increased bone formation in osteocalcin-deficient mice. Nature 1996;382:448–52. [5] Boskey AL, Gadaleta S, Gundberg C, Doty SB, Ducy P, Karsenty G. Fourier transform infrared microspectroscopic analysis of bones of osteocalcin-deficient mice provides insight into the function of osteocalcin. Bone 1998;23:187–96. [6] Binkley NC, Krueger DC, Kawahara TN, Engelke JA, Chappell RJ, Suttie JW. A high phylloquinone intake is required to achieve maximal osteocalcin gammacarboxylation. Am J Clin Nutr 2002;76:1055–60. [7] Tabb MM, Sun A, Zhou C, Grun F, Errandi J, Romero K, et al. Vitamin K2 regulation of bone homeostasis is mediated by the steroid and xenobiotic receptor SXR. J Biol Chem 2003;278:43919–27. [8] Atkins GJ, Welldon KJ, Wijenayaka AR, Bonewald LF, Findlay DM. Vitamin K promotes mineralization, osteoblast-to-osteocyte transition, and an anticatabolic phenotype by {gamma}-carboxylation-dependent and -independent mechanisms. Am J Physiol Cell Physiol 2009;297:C1358–67. [9] Suttie JW. The importance of menaquinones in human nutrition. Annu Rev Nutr 1995;15:399–417. [10] Booth SL, Tucker KL, Chen H, Hannan MT, Gagnon DR, Cupples LA, et al. Dietary vitamin K intakes are associated with hip fracture but not with bone mineral density in elderly men and women. Am J Clin Nutr 2000;71:1201–8. [11] Feskanich D, Weber P, Willett WC, Rockett H, Booth SL, Colditz GA. Vitamin K intake and hip fractures in women: a prospective study. Am J Clin Nutr 1999;69:74–9. [12] Rejnmark L, Vestergaard P, Charles P, Hermann AP, Brot C, Eiken P, et al. No effect of vitamin K1 intake on bone mineral density and fracture risk in perimenopausal women. Osteoporos Int 2006;17:1122–32. [13] Iwamoto J, Sato Y, Takeda T, Matsumoto H. High-dose vitamin K supplementation reduces fracture incidence in postmenopausal women: a review of the literature. Nutr Res 2009;29:221–8. [14] Schurgers LJ, Vermeer C. Determination of phylloquinone and menaquinones in food. Effect of food matrix on circulating vitamin K concentrations. Haemostasis 2000;30:298–307. [15] Newman P, Bonello F, Wierzbicki AS, Lumb P, Savidge GF, Shearer MJ. The uptake of lipoprotein-borne phylloquinone (vitamin K1) by osteoblasts and osteoblastlike cells: role of heparan sulfate proteoglycans and apolipoprotein E. J Bone Miner Res 2002;17:426–33.
E.M. Apalset et al. / Bone 49 (2011) 990–995 [16] Pluijm SM, Dik MG, Jonker C, Deeg DJ, van Kamp GJ, Lips P. Effects of gender and age on the association of apolipoprotein E epsilon4 with bone mineral density, bone turnover and the risk of fractures in older people. Osteoporos Int 2002;13:701–9. [17] Cauley JA, Zmuda JM, Yaffe K, Kuller LH, Ferrell RE, Wisniewski SR, et al. Apolipoprotein E polymorphism: a new genetic marker of hip fracture risk—the study of osteoporotic fractures. J Bone Miner Res 1999;14:1175–81. [18] Kohlmeier M, Saupe J, Schaefer K, Asmus G. Bone fracture history and prospective bone fracture risk of hemodialysis patients are related to apolipoprotein E genotype. Calcif Tissue Int 1998;62:278–81. [19] von Muhlen DG, Barrett-Connor E, Schneider DL, Morin PA, Parry P. Osteoporosis and apolipoprotein E genotype in older adults: the Rancho Bernardo study. Osteoporos Int 2001;12:332–5. [20] Peter I, Crosier MD, Yoshida M, Booth SL, Cupples LA, Dawson-Hughes B, et al. Associations of APOE gene polymorphisms with bone mineral density and fracture risk: a meta-analysis. Osteoporos Int 2011;22:1199–209. [21] Gerdes LU, Klausen IC, Sihm I, Faergeman O. Apolipoprotein E polymorphism in a Danish population compared to findings in 45 other study populations around the world. Genet Epidemiol 1992;9:155–67. [22] Bagger YZ, Rasmussen HB, Alexandersen P, Werge T, Christiansen C, Tanko LB. Links between cardiovascular disease and osteoporosis in postmenopausal women: serum lipids or atherosclerosis per se? Osteoporos Int 2007;18:505–12. [23] Sennels HP, Sand JC, Madsen B, Lauritzen JB, Fenger M, Jorgensen HL. Association between polymorphisms of apolipoprotein E, bone mineral density of the lower forearm, quantitative ultrasound of the calcaneus and osteoporotic fractures in postmenopausal women with hip or lower forearm fracture. Scand J Clin Lab Invest 2003;63:247–58. [24] Apalset EM, Gjesdal CG, Eide GE, Johansen A-MW, Drevon CA, Tell GS. Dietary vitamins K1, K2 and bone mineral density: the Hordaland Health Study. Arch Osteoporos 2010, doi:10.1007/s11657-010-0036-6. [25] Lehmann DJ, Refsum H, Nurk E, Warden DR, Tell GS, Vollset SE, et al. Apolipoprotein E epsilon4 and impaired episodic memory in communitydwelling elderly people: a marked sex difference. The Hordaland Health Study. J Neurol Neurosurg Psychiatry 2006;77:902–8. [26] Gjesdal CG, Vollset SE, Ueland PM, Refsum H, Meyer HE, Tell GS. Plasma homocysteine, folate, and vitamin B 12 and the risk of hip fracture: the Hordaland Homocysteine Study. J Bone Miner Res 2007;22:747–56. [27] Nes M, Frost Andersen L, Solvoll K, Sandstad B, Hustvedt BE, Lovo A, et al. Accuracy of a quantitative food frequency questionnaire applied in elderly Norwegian women. Eur J Clin Nutr 1992;46:809–21.
995
[28] Andersen LF, Solvoll K, Johansson LR, Salminen I, Aro A, Drevon CA. Evaluation of a food frequency questionnaire with weighed records, fatty acids, and alphatocopherol in adipose tissue and serum. Am J Epidemiol 1999;150:75–87. [29] Andersen LF, Veierod MB, Johansson L, Sakhi A, Solvoll K, Drevon CA. Evaluation of three dietary assessment methods and serum biomarkers as measures of fruit and vegetable intake, using the method of triads. Br J Nutr 2005;93:519–27. [30] Nurk E, Refsum H, Tell GS, Engedal K, Vollset SE, Ueland PM, et al. Plasma total homocysteine and memory in the elderly: the Hordaland Homocysteine Study. Ann Neurol 2005;58:847–57. [31] Macdonald HM, McGuigan FE, Lanham-New SA, Fraser WD, Ralston SH, Reid DM. Vitamin K1 intake is associated with higher bone mineral density and reduced bone resorption in early postmenopausal Scottish women: no evidence of gene–nutrient interaction with apolipoprotein E polymorphisms. Am J Clin Nutr 2008;87:1513–20. [32] Cox DR. Regression models and life-tables. J R Stat Soc B 1972;34:187–220. [33] Eide GE, Omenaas E, Gulsvik A. The semi-proportional hazards model revisited: practical reparametrizations. Stat Med 1996;15:1771–7. [34] Cheung AM, Tile L, Lee Y, Tomlinson G, Hawker G, Scher J, et al. Vitamin K supplementation in postmenopausal women with osteopenia (ECKO trial): a randomized controlled trial. PLoS Med 2008;5:e196. [35] Iwamoto J, Takeda T, Ichimura S. Effect of menatetrenone on bone mineral density and incidence of vertebral fractures in postmenopausal women with osteoporosis: a comparison with the effect of etidronate. J Orthop Sci 2001;6:487–92. [36] Shiraki M, Shiraki Y, Aoki C, Miura M. Vitamin K2 (menatetrenone) effectively prevents fractures and sustains lumbar bone mineral density in osteoporosis. J Bone Miner Res 2000;15:515–21. [37] Inoue T, Fujita T, Kishimoto H, Makino T, Nakamura T, Sato T, et al. Randomized controlled study on the prevention of osteoporotic fractures (OF study): a phase IV clinical study of 15-mg menatetrenone capsules. J Bone Miner Metab 2009;27: 66–75. [38] Binkley N, Harke J, Krueger D, Engelke J, Vallarta-Ast N, Gemar D, et al. Vitamin K treatment reduces undercarboxylated osteocalcin but does not alter bone turnover, density, or geometry in healthy postmenopausal North American women. J Bone Miner Res 2009;24:983–91. [39] Ronden JE, Drittij-Reijnders MJ, Vermeer C, Thijssen HH. Intestinal flora is not an intermediate in the phylloquinone-menaquinone-4 conversion in the rat. Biochim Biophys Acta 1998;1379:69–75. [40] Kohlmeier M, Salomon A, Saupe J, Shearer MJ. Transport of vitamin K to bone in humans. J Nutr 1996;126:1192S–6S.