- Email: [email protected]
Vitamin K status and cognitive function in healthy older adults
Neurobiology of Aging 34 (2013) 2777e2783
Contents lists available at SciVerse ScienceDirect
Neurobiology of Aging journal homepage: www.elsevier.com/locate/neuaging
Vitamin K status and cognitive function in healthy older adults Nancy Presse a, b, Sylvie Belleville a, c, Pierrette Gaudreau d, e, Carol E. Greenwood f, Marie-Jeanne Kergoat a, d, Jose A. Morais g, Hélène Payette h, Bryna Shatenstein a, b, Guylaine Ferland a, b, * a
Centre de recherche, Institut Universitaire de Gériatrie de Montréal, Montréal, Quebec, Canada Département de Nutrition, Université de Montréal, Montréal, Quebec, Canada c Département de Psychologie, Université de Montréal, Montréal, Quebec, Canada d Département de Médecine, Université de Montréal, Montréal, Quebec, Canada e Centre de Recherche, Centre Hospitalier de l’Université de Montréal, Montréal, Quebec, Canada f Kunin Lunenfeld Applied and Evaluative Unit, Baycrest and Department of Nutritional Sciences, University of Toronto, Toronto, Ontario, Canada g McGill Nutrition and Food Science Centre and Department of Medicine, Division of Geriatrics, McGill University Health Centre and School of Dietetics and Human Nutrition, McGill University, Montréal, Quebec, Canada h Centre de recherche sur le vieillissement, Institut Universitaire de Gériatrie de Sherbrooke and Faculté de Médecine et des Sciences de la Santé, Université de Sherbrooke, Sherbrooke, Quebec, Canada b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 19 December 2012 Received in revised form 22 April 2013 Accepted 30 May 2013 Available online 11 July 2013
Evidence is accumulating that vitamin K could have a role in cognition, especially in aging. Using data from the Québec Longitudinal Study on Nutrition and Successful Aging (NuAge), a cross-sectional analysis was conducted to examine the associations between vitamin K status, measured as serum phylloquinone concentrations, and performance in verbal and non-verbal episodic memory, executive functions, and speed of processing. The sample included 320 men and women aged 70 to 85 years who were free of cognitive impairment. After adjustment for covariates, higher serum phylloquinone concentration (log-transformed) was associated with better verbal episodic memory performances (F ¼ 2.43, p ¼ 0.048); specifically with the scores (Z-transformed) on the second (b ¼ 0.47; 95% confidence interval [CI] ¼ 0.13e0.82), third (b ¼ 0.41; 95% CI ¼ 0.06e0.75), and 20-minute delayed (b ¼ 0.47; 95% CI ¼ 0.12 e0.82) free recall trials of the RL/RI-16 Free and Cued Recall Task. No associations were found with nonverbal episodic memory, executive functions, and speed of processing. Our study adds evidence to the possible role of vitamin K in cognition during aging, specifically in the consolidation of the memory trace. Ó 2013 Elsevier Inc. All rights reserved.
Keywords: Vitamin K Phylloquinone Cognition Aging
1. Introduction Vitamin K is a fat-soluble compound found predominantly in green leafy vegetables and, to lesser extent, other vegetables, fruits, and some vegetable oils. It was originally discovered for its role in blood coagulation, activating the vitamin Kedependent clotting factors through its function as a cofactor of the g-glutamylcarboxylase. The discovery of additional vitamin K-dependent proteins (e.g. osteocalcin) and functions independent of the gcarboxylation has expanded the role of vitamin K beyond coagulation (Suttie, 2009). A role in the human brain was first suspected after cases of central nervous system abnormalities in newborns exposed in utero to vitamin K antagonists (Hall et al., 1980). In rat and human brain tissues, high vitamin K levels in the form of * Corresponding author at: Centre de recherche, Institut Universitaire de Gériatrie de Montréal, 4565 Queen Mary, Montréal, Quebec H3W 1W5, Canada. Tel.: þ1 514 340 2800 3236; fax: þ1 514 340 2801. E-mail address: [email protected] (G. Ferland). 0197-4580/$ e see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.neurobiolaging.2013.05.031
menaquinone-4 (MK-4) have been reported (Carrie et al., 2004; Thijssen and Drittij-Reijnders, 1994, 1996). In rodents, brain MK-4 was shown to respond, in a dose-dependent manner, to intakes of vitamin K, whether provided in a natural dietary form (phylloquinone) or a synthetic form (Carrie et al., 2004; Okano et al., 2008; Thijssen and Drittij-Reijnders, 1994). Our understanding of the vitamin K functions in brain is growing (Ferland, 2012). The vitamin K-dependent proteins Gas6 (growth arrest-specific gene 6) and protein S are expressed in the central nervous system (He et al., 1995; Prieto et al., 1999). Protein S is notably known to possess neuroprotective effects during hypoxic/ ischemic injury, whereas Gas6 is now recognized as an important regulator of cell survival, cell growth, and myelination processes (Ferland, 2012). Furthermore, vitamin K participates in the metabolism of sphingolipids, a major constituent of the myelin sheath and neuronal membranes, also involved in important molecular events such as cell signaling (Denisova and Booth, 2005; Ferland, 2012). Finally, the K vitamers were also shown to possess neuroprotective effects against oxidative-induced death of primary
2778
N. Presse et al. / Neurobiology of Aging 34 (2013) 2777e2783
cultures of rat oligodendrocytes, rat neurons, and human neuroblastoma cells by preventing the accumulation of reactive oxygen species (Isaev et al., 2004; Li et al., 2003, 2009; Sakaue et al., 2011). Consistent with mechanistic studies, in vivo rodent models have provided evidence that vitamin K status can modulate brain sphingolipids, behavior, and cognitive performances (Carrie et al., 2004, 2011; Cocchetto et al., 1985; Crivello et al., 2010; Sundaram et al., 1996). Interestingly, current research suggests that vitamin K exerts its actions in aging. Notably, higher concentrations of MK-4 in the hippocampus and cortex were shown to be correlated with higher myelin sulfatides in adult and aged rats, but not in younger animals (Crivello et al., 2010). Similarly, 20-month-old rats fed a vitamin Kerich or eadequate diet throughout their lives had better spatial learning performances in the Morris Water Maze task than those fed a lowevitamin K diet, a result not observed in 6-monthold rats (Carrie et al., 2011). A vulnerability of the aging brain to vitamin K status is also strengthened by studies reporting a marked age-dependent loss of Gas6 in rat brain (Tsaioun et al., 2000) and sphingolipid alterations in brain tissues of aged mice and older adults with Alzheimer’s disease (Cutler et al., 2004). The potential role of vitamin K in cognitive function of older adults has not yet been examined. Thus far, two cross-sectional studies reported a lower vitamin K status in subjects in the early and late stages of Alzheimer’s disease in comparison with controls (Presse et al., 2008; Sato et al., 2005). In the present study, we examined the associations between serum phylloquinone concentration and performances in verbal and non-verbal episodic memory, executive functions, and speed of processing in 320 healthy older adults selected from the Québec Longitudinal Study on Nutrition and Successful Aging (NuAge). To our knowledge, this is the first study to investigate the potential role of vitamin K in specific cognitive domains. 2. Methods The NuAge Study is a prospective cohort study of nutritional factors of successful aging. The methodology has been detailed elsewhere (Gaudreau et al., 2007). Briefly, community-dwelling men and women aged 67 to 84 years were randomly selected from the Québec Medicare Database. To be include in the NuAge Study, participants had to be in good general health, free of disabilities in activities of daily living, free of cognitive impairment as established by a Modified Mini-Mental State Examination (3MS) score of >79, and free of the following health conditions: Class II heart failure, chronic obstructive pulmonary disease requiring home oxygen therapy or oral steroids, inflammatory digestive diseases, and cancer treated by radiation therapy, chemotherapy, or surgery in the previous 5 years. After a telephone interview and a clinical examination, 1793 participants were recruited in 2003 to 2005. Follow-up interviews were performed annually up to 3 years after recruitment. The present study is based on an embedded subsample of 464 French-speaking participants who underwent a cognitive evaluation 1.9 to 4.0 years (mean SD, 3.0 0.6 years) after their recruitment in the NuAge cohort (2006e2008). These participants had to have 3MS scores of >85 to ensure that they could complete the detailed cognitive evaluation, and had to be free of health conditions that could impair cognition (self-reported), namely Parkinson’s disease, history of stroke or cerebral hemorrhage, Alzheimer’s disease or other dementia, Guillain-Barré syndrome, epilepsy, head trauma resulting in unconsciousness, cardiac arrest with resuscitation, brain tumor or metastasis, central nervous system infection (e.g., meningitis), multiple sclerosis, or CO2 or methanol intoxication. All participants provided informed consents approved by the ethics committees of both testing sites, the
Instituts Universitaires de Gériatrie de Montréal and Sherbrooke (QC, Canada). Two health professionals, including a geriatrician, thoroughly reviewed each of the 464 participant files to ensure a sample of older adults free of cognitive impairment who were unlikely to have changed their diet because of underlying cognitive problems. As a result, 36 participants were excluded, as they were found to have Parkinson’s disease (n ¼ 2), actively treated epilepsy (n ¼ 3), Alzheimer’s disease or other dementia (n ¼ 7), or history of stroke/ cerebral hemorrhage (n ¼ 24). We also excluded those treated with vitamin K antagonists (warfarin or acenocoumarol; n ¼ 28), and subjects with incomplete dataset (n ¼ 65). Finally, 15 participants were excluded because they had abnormally low cognitive performances for healthy individuals. These 15 participants were identified by studentized (jackknife) residuals less than 3.0 after cognitive scores were regressed on testing research assistant, age, sex, and number of years of education. The final sample included 320 of the 464 participants (see Supplemental Fig. 1 for a detailed flowchart). 2.1. Cognitive assessment The cognitive test battery was administered by trained research assistants and consisted of 6 tests covering 4 cognitive domains, namely verbal and non-verbal episodic memory, executive functions, and speed of processing (see Supplemental Text 1 for a detailed description). From the 6 cognitive tests, 13 scores were calculated and used to assess performances on each cognitive domain. Verbal episodic memory was assessed using scores of the 3 immediate free recall trials and the 20-minute delayed free recall of the RL/RI-16 Free and Cued Recall Task (RL/RI-16 FCRT) (Groupe de réflexion sur l’évaluation de la mémoire and Van der Linden, 2004); non-verbal episodic memory was tested by scores of the 3-minute and 20-minute recalls of the Rey Complex-Figure (Rey, 1959); executive functions were assessed by those of the Rey Complex-Figure copying, the third plate of the Stroop Test (Spreen and Strauss, 1991), the total number of letters recalled in the interference conditions (10-, 20-, and 30-second delays) of the Adapted BrownePeterson procedure (Belleville et al., 2004; Spreen and Strauss, 1998), and the total number of completed targets of the Weschsler Adult Intelligence ScaleeIII (WAIS-III) Digit Symbol-Coding subtest (Wechsler, 1997); finally, speed of processing was evaluated using scores from the 2 first plates of the Stroop Test and mean initiation time at the Choice-Reaction Time (Cohen et al., 1993). 2.2. Serum phylloquinone, triglycerides, and total cholesterol concentrations Blood sampling, processing, and storing in the NuAge Study followed standard operating procedures (Gaudreau et al., 2007). At annual interviews, a 50-mL blood sample was obtained between 7:30 and 8:30 AM after an overnight fast. Specifically, participants were instructed not to eat, drink (except reasonable amount of water), smoke, or chew gum after 19:00 the evening before the blood sampling. Blood specimens were immediately stored at 4 C and protected from light until processing. Serum was extracted, aliquoted in 1.5-mL, opalescent-colored Eppendorf polypropylene tubes. Serum samples were stored at 20 C in opaque cardboard for few hours and then at 80 C for long-term storage. Overall, light exposure was limited to periods of sample processing, for a total light exposure time of about 100 minutes. Selected blood samples were collected from less than 11.5 months before to 2.5 months after the cognitive evaluation (mean SD, 3.2 3.2 months); that is, between June 2006 and May 2008. In Canada, over-the-counter supplements containing vitamin K
N. Presse et al. / Neurobiology of Aging 34 (2013) 2777e2783
were not available during that period, and none of the 320 participants reported taking vitamin K as a prescribed drug. Thus, circulating phylloquinone reflects dietary phylloquinone intakes. The 480-mL serum samples used for phylloquinone analyses had not previously been thawed and refrozen. Samples were shipped on dry ice in opaque cardboard. Phylloquinone analyses were conducted in July to August 2010 in Dr Guylaine Ferland’s laboratory using standardized HPLC procedures (Davidson and Sadowski, 1997; Wang et al., 2004). Time between blood sampling and phylloquinone analysis was 3.13 0.48 years (mean SD). Quality was controlled using an external serum standard. The intra-assay coefficient of variation (CV) was 9.2% and 6.0% for phylloquinone concentrations of 0.2 and 0.8 nmol/L, respectively, and did not vary over time. The detection limit was 0.03 nmol/L, and none of the 320 serum samples was below this limit. Phylloquinone in blood is mostly carried by triglyceride-rich lipoproteins and, to lesser extent, by the low- and high-density lipoprotein fractions, making the blood lipid profile an important determinant of serum/plasma phylloquinone concentration (Azharuddin et al., 2007; Erkkila et al., 2004; Lamon-Fava et al., 1998; McKeown et al., 2002; Presse et al., 2012; Sadowski et al., 1989; Shea et al., 2009). Given that blood lipid profile was also associated with cognitive performance (Morley and Banks, 2010), triglyceridemia and total cholesterol were considered as covariates in the present study. The overnight fasting concentrations of serum triglycerides and total cholesterol were determined at the Centre Hospitalier de l’Université de Montréal clinical biochemistry laboratory, on a Roche/Hitachi Cobas c311 System analyzer using the TRIGL and CHOL2 kits, respectively (Roche Diagnostic Corp, Indianapolis, IN). 2.3. Other covariates A structured interview conducted annually in the NuAge Study provided self-reported information on race (white vs. other), sex, age, number of years of education, household income (in Canadian dollars), medical history, medications, and lifestyle at the time of the cognitive evaluation. History of hypertension was based on selfreported diagnoses. History of diabetes was determined from medication use (e.g., insulin, oral hypoglycemic agents) or fasting glycemia 7.0 mmol/L. Number of depressive symptoms was determined using the Geriatric Depression Scale (GDS) (Yesavage et al., 1983). Weight and height were measured, and body mass index (BMI) was calculated as weight in kilograms divided by the square of height in meters. Ponderal status was described as normal (BMI <25 kg/m2), overweight (BMI 25 to <30 kg/m2), and obese (BMI 30 kg/m2). Lifestyle was characterized by smoking status (never vs. current/former smoker), alcohol consumption, physical activity, use of vitamin/mineral supplements (regular vs. occasional/ never user), and diet quality. Mean daily alcohol intake (g) was determined by a validated semi-quantitative food-frequency questionnaire completed at the NuAge baseline, which assesses usual eating habits in the past year (Shatenstein et al., 2005a). Weekly consumption of beer, wine, and spirits was expressed as number of drinks (0, >0 to <5, and 5 drinks per week) based on a standard drink of 13.6 g of alcohol. Usual physical activity was assessed using the Physical Activity Scale for the Elderly (PASE), with higher scores indicative of higher physical activity level (Washburn et al., 1999). Physical activity level was then categorized as PASE score <75, 75 to <125, and 125. Diet quality was assessed using the CanadianHealthy Eating Index (C-HEI), a 100-point index with higher scores indicating greater diet quality (Shatenstein et al., 2005b). The C-HEI was calculated from food and nutrient intakes estimated by the food-frequency questionnaire to assess adherence to the 2007 Canada Food Guide and the 1990 Nutrition Recommendations for Canadians (Canada, 2007; Canada, 1990).
2779
2.4. Statistical analyses Cognitive test scores measured as “time” (e.g., seconds in the Rey-Complex Figure) were log-transformed because of departures from normality, and were multiplied by 1 so that an increasing score indicated better performance for all tests. All scores were Ztransformed to allow comparison across cognitive tests. To examine the associations between serum phylloquinone concentration and cognitive performance, multivariate linear regressions were used to simultaneously analyze scores relating to each of the 4 cognitive domains assessed. In the base model, serum phylloquinone (log-transformed) was regressed on cognitive performance adjusting for age, sex, education, serum triglycerides, serum total cholesterol, and testing research assistant. The “testing research assistant” was modeled as a dummy variable, using the research assistant who tested the largest number of participants as reference. In the full model, other covariates chosen a priori were added to the base model: annual household income, hypertension (binary), diabetes (binary), BMI (categorical), number of depressive symptoms, smoking status (binary), alcohol consumption (categorical), physical activity level (categorical), regular use of vitamin/mineral supplements (binary), and diet quality index. Because the sample was basically white (>99%), race was not considered as a covariate. The test assumptions were verified including multivariate normality (Mahalanobis distance), linearity, multicollinearity, and homogeneity of varianceecovariance matrices, with no violations noted. Wilks’ lambda statistics were reported for both the base and full models. Significant results (p < 0.05) on multivariate analyses (full model only) were analyzed further using univariate linear regressions for each of the cognitive scores. Bonferroni adjustment was used to control for type I error. Exploratory analysis was performed for effect modification by sex, age, education, and triglycerides. For these analyses, we evaluated the significance of the cross-product term that resulted from multiplying serum phylloquinone concentration by each variable, 1 variable at a time. Adequacy of the full models was evaluated by the following: (1) examining added variable plots showing adjusted regression lines; (2) comparing these lines with loess regression lines (Epanechnikov kernel function, 50% fit); (3) assessing normality of residuals; and (4) plotting residuals against predicted values. To evaluate the magnitude of effect for significant associations (two-tailed p < .05), test scoreeadjusted means were calculated at the 2.5th, 50th, and 97.5th percentiles of serum phylloquinone concentration. The ManneWhitney U and c2 tests were used to compare the sample to participants not selected from the NuAge cohort. Characteristics of the participants are presented as mean ( SD) or as percentages, and their correlation with serum phylloquinone was determined using regression analysis. All statistical analyses were performed using IBM SPSS Statistics 21 software (SPSS Inc, Chicago, IL). 3. Results The 320 selected participants from the NuAge cohort were younger, more educated, had greater household income, had higher global cognitive scores, had fewer depressive symptoms, were more likely to consume alcohol, and tended to be more physically active than those not selected (Supplemental Table 1). However, they did not differ by sex, race, health and smoking status, use of vitamin/ mineral supplements, or diet quality. At the time of the cognitive evaluation, participants were aged 70.0 to 85.5 years and had 3MS scores indicative of normal cognitive functioning (mean SD, 94.56 4.23; range, 82 to 100), with only 12.5% of participants having a score <90. Mean BMI,
2780
N. Presse et al. / Neurobiology of Aging 34 (2013) 2777e2783
Table 1 Characteristics of the subjects (n ¼ 320) and their correlation with the serum phylloquinone concentration at the time of the cognitive evaluation in the NuAge Study Characteristic Sociodemographic Sex (% female) Age (y) Race (% white) Education (y) Household income (10,000 CAD) Health BMI (kg/m2) Ponderal status (%) Overweight (BMI 25 to < 30 kg/m2) Obese (BMI 30 kg/m2) Diabetes (%) Hypertension (%) GDS (number of depressive symptoms) Serum triglycerides (mmol/L) Serum total cholesterol (mmol/L) Lifestyle Former or current smoker (%) Alcohol consumption (%)c 0 drink/week >0 to <5 drinks/week 5 drinks/week PASE score Physical activity level (%) PASE score <75 PASE score 75 to <125 PASE score 125 Regular use of vitamin/mineral supplements (%) C-HEIc Serum phylloquinone concentration (nmol/L)
Value
Correlation coefficienta
54.1 76.4 3.8b 99.4 12.85 4.56 4.33 2.31
0.15** 0.02 0.02 0.03 0.03
27.8 4.8
0.02
49.7 23.8 10.3 53.4 4.25 3.82 1.65 0.82 5.33 1.07
0.03 0.03 0.11 0.08 0.02 0.40*** 0.30***
43.4
0.06
23.8 41.9 34.9 95.5 45.2
— 0.03 0.06 0.03
35.9 41.9 22.2 59.7
— 0.03 0.02 0.06
79.5 8.6 1.45 1.80
0.04 —
** p < 0.01; *** p < 0.001. Key: BMI, body mass index; CAD, Canadian dollars; C-HEI, Canadian-Healthy Eating Index; GDS, Geriatric Depression Scale; PASE, Physical Activity Scale for the Elderly. a Correlation coefficients were determined by regression analysis with serum phylloquinone concentration (log-transformed) as dependent variable. Partial correlation coefficients for BMI categories, alcohol consumption, and physical activity levels are presented. b Mean SD (all such values). c Computed using data collected by a food frequency questionnaire administered at baseline in the NuAge Study.
triglycerides, and total cholesterol were within the acceptable range for older adults (Cook et al., 2005; NCEP ATP III, 2002) (Table 1). Most participants had healthy lifestyle habits as shown by the presence of only 13 current smokers, low to moderate alcohol consumption, regular use of vitamin/mineral supplements, and the fact that 49.4% of participants had C-HEI scores >80, indicative of “good” diet quality (Shatenstein et al., 2005b). Serum phylloquinone concentrations (median, 1.06 nmol/L; range, 0.08e23.57) were comparable with those found in healthy middle-aged and older adults in the United States (Booth et al., 1995; Booth et al., 1997; McKeown et al., 2002; Neogi et al., 2006; Sadowski et al., 1989; Shea et al., 2009). Serum phylloquinone concentrations were higher in women (median, 1.12 nmol/L; range, 0.23e23.57) than men (median, 0.97 nmol/L; range, 0.08e6.85) and were significantly correlated with triglycerides and total cholesterol. Mean ( SD) cognitive test scores are presented in Supplemental Table 2.
Table 2 Results from the multivariate linear regressions of test scores assessing four cognitive domains on serum phylloquinone concentrations in older adults from the NuAge Study (n ¼ 320) Cognitive domainsa
Verbal episodic memoryd Non-verbal episodic memorye Executive functionsf Speed of processingg
Base modelb
Full modelc
Wilks’ lambda
F (df1, df2)
Wilks’ lambda
F (df1, df2)
0.965 0.992
2.76* (4, 303) 1.30 (2, 305)
0.968 0.994
2.43* (4, 290) 0.90 (2, 292)
0.998 0.989
0.19 (4, 304) 1.14 (3, 305)
0.997 0.993
0.21 (4, 291) 0.67 (3, 292)
* p < 0.05. a All cognitive scores were Z-transformed and standardized for direction. b The base model was adjusted for age (continuous), sex (binary), testing research assistant (categorical), education (continuous), serum triglycerides (log-transformed continuous), and total cholesterol (continuous). c The full model was adjusted for household income (continuous), body mass index (categorical), diabetes (binary), hypertension (binary), number of depressive symptoms, smoking status (binary), alcohol consumption (categorical), physical activity (categorical), vitamin/mineral supplement use (binary), and diet quality index (continuous), in addition to the variables included in the base model. d Based on 4 scores: the 4 free recall trials of the RL/RI-16 Free and Cued Recall Task. e Based on 2 scores: the 3-minute and 20-minute recalls of the Rey ComplexFigure. f Based on 4 scores: the Rey Complex-Figure copying, the 3rd plate of the Stroop Test, the Adapted Brown-Peterson procedure, and the Weschsler Adult Intelligence ScaleeIII Digit Symbol-Coding subtest. g Based on 3 scores: the 2 first plates of the Stroop Test and the Choice-Reaction Time.
although a decrease in its magnitude was observed in comparison with the base model. No association was found with non-verbal episodic memory, executive functions, or speed of processing. When the verbal episodic memory scores were considered separately (Table 3), significant associations were found with the second, the third (marginally considering the Bonferroni adjustement), and the 20-minute delayed free recall trials of the RL/RI-16 FCRT, whereas no associations was observed with the first trial of the RL/RI-16 FCRT. Exploratory analysis did not reveal any consistent evidence of effect modification by sex, age, education, and triglycerides. Analysis of the magnitude of effect in the full models showed that an increase in serum phylloquinone concentration from 0.27 to 1.06 nmol/L (from the 2.5th to 50th percentile) was associated with an increase of 8.2%, 6.1%, and 6.4% in performances on the second, the third, and the 20-minute delayed free recall trials of the RL/RI-16 FCRT, respectively (Fig. 1). However, because the relationship is a logarithmic function, serum phylloquinone had to increase from 1.06 to 5.33 nmol/L (from the 50th to 97.5th percentile) to observe a similar improvement ([6%e8%) in free recall performances. For 203 of the 320 participants, serum specimens drawn 2 years before the cognitive evaluation were available. Therefore, we determined phylloquinone concentrations on these samples and conducted supplemental regression analyses to examine the temporality of the cross-sectional associations observed with the 320 participants. Although we could use only a limited number of confounding factors due to the sample size, results of these supplemental analyses were similar, that is, serum phylloquinone concentration was significantly associated with the second and third free recall trials but not with the first one (see Supplemental Table 3).
3.1. Serum phylloquinone concentration and cognitive performances
4. Discussion
In the base model, higher serum phylloquinone was significantly associated with better performance in verbal episodic memory (Table 2). In the full model, the association remained significant,
The present study is important, because it provides support for an emerging role of vitamin K in the aging brain. Specifically, we showed that serum phylloquinone was positively associated with
N. Presse et al. / Neurobiology of Aging 34 (2013) 2777e2783 Table 3 Results from the univariate linear regressions of verbal episodic memory test scores on serum phylloquinone concentrations in older adults from the NuAge Study (n ¼ 320) Cognitive Test
RL/RI-16 Free and Cued Recall Taskc Free recall, Trial 1 Free recall, Trial 2 Free recall, Trial 3 20-Minute delayed free recall
Base modela
Full modelb
b-coefficients for serum phylloquinone (95% CI)
b-coefficients for serum phylloquinone (95% CI)
0.24 0.49 0.43 0.51
0.21 0.47 0.41 0.47
(0.12 to 0.59) (0.15 to 0.83)** (0.09 to 0.77)* (0.16 to 0.85)**
(0.14 to 0.57) (0.13 to 0.82)** (0.06 to 0.75)* (0.12 to 0.82)**
* p < 0.05; ** p < 0.01. a The base model was adjusted for the RL/RI-16 word list version (binary), age (continuous), sex (binary), testing research assistant (categorical), education (continuous), serum triglycerides (log-transformed continuous), and total cholesterol (continuous). b The full model was adjusted for household income (continuous), body mass index (categorical), diabetes (binary), hypertension (binary), number of depressive symptoms, smoking status (binary), alcohol consumption (categorical), physical activity (categorical), vitamin/mineral supplement use (binary), diet quality index (continuous), in addition to the variables included in the base model. c All cognitive scores were Z-transformed and standardized for direction, so a positive coefficient means that test performance enhances with increasing serum phylloquinone concentration (log-transformed).
performance in verbal episodic memory, while being unrelated to non-verbal episodic memory, executive functions, and speed of processing. Episodic memory refers to the memory of events within their spatio-temporal context. For example, remembering where one’s keys were last left relies on episodic memory. It is thus important to everyday life. Episodic memory processes include encoding (when the information to memorize is first encountered) and memory consolidation (when the memory trace is stabilized for long-term storage). Interestingly, our findings advocate for a specific role of vitamin K in memory consolidation. Indeed, a higher serum phylloquinone concentration was associated with better performances on the second, the third, and the 20-minute delayed free recall trials of the verbal episodic memory test while not being associated with the first one. This pattern indicates that the higher the vitamin K status, the more the participants benefited from the successive learning trials. The fact that no association was observed with the
16
Number of words recalled
14 12 10 8 6
5.33 nmol/L 1.06 nmol/L
4
0.27 nmol/L 2 0 Trial 1
Trial 2
Trial 3
20-min delayed trial
Free recall trials of the RL/RI-16 Free and Cued Recall Task
Fig. 1. Performance adjusted-means ( SEM) across free recall trials of the RL/RI-16 Free and Cued Recall Task for serum phylloquinone concentrations at the 2.5th, 50th, and 97.5th percentiles. Means were adjusted for testing research assistant, RL/RI-16 word-list version, age, sex, education, household income, diabetes, hypertension, number of depressive symptoms, body mass index, alcohol consumption, smoking status, physical activity, use of vitamin/mineral supplements, diet quality, serum triglycerides, and total cholesterol.
2781
performances in the non-verbal episodic memory test (Rey Complex-Figure recalls), where the abstract figure is presented only once, also advocates for a specific link between vitamin K and consolidation of the memory trace. A specific role of vitamin K in memory consolidation is in line with a recent study in rodents (Carrie et al., 2011). When rats were subjected to a 5-day spatial learning task (Morris Water Maze), those fed a low vitamin K diet throughout their lives required more exposure time to the visual cues to achieve performances similar to those fed a diet adequate or rich in vitamin K. Moreover, rats fed the lowevitamin K diet presented an altered sphingolipid profile in their hippocampus (Carrie et al., 2011), the cerebral region responsible for memory consolidation. The possible action of vitamin K in the hippocampus is also strengthened by studies on the protein Gas6. Specifically, the receptor Tyro3, one of the Gas6 ligands, is highly expressed in the hippocampal region (Prieto et al., 2007), and its activation by Gas6 results in molecular events known to underlie memory consolidation (Davis and Laroche, 2006; Horwood et al., 2006; Prieto et al., 2007). A role for vitamin K in the aging brain surely raises questions for those treated with vitamin K-antagonists. Although limited, available data do suggest that vitamin K functions in the brain could be altered by these drugs. Notably, warfarin-treated rodents were shown to present decreased concentrations of brain MK-4 (Thijssen et al., 1996) and alterations in brain sphingolipids (Sundaram and Lev, 1988). Whether treatment with vitamin K-antagonists in humans could increase the risk of cognitive problems has, to our knowledge, not yet been examined. Nonetheless, given that warfarin-treated individuals are often told to limit their consumption of vitamin K-rich foods, which could further deplete their brain’s vitamin K status, the implications of a role of vitamin K in cognition could be significant. In the present study, the association between episodic memory performance and serum phylloquinone concentration was found to be a logarithmic function such that the rate of improvement of the scores (number of words recalled) slowed down considerably once the phylloquinone concentration reached approximately 1.0 nmol/ L. Interestingly, 2 studies reported similar “threshold” effects at approximately 1.0 nmol/L when examining associations between vitamin K status and bone health (Kohlmeier et al., 1997; Neogi et al., 2006). Reasons explaining this threshold effect remain to be determined, but this observation should be taken into consideration when planning studies on the role of vitamin K in cognition. Phylloquinone is the main K vitamer in diet and available data indicate that other K vitamers found in the North American diet as well as menaquinones synthesized by gut bacteria would not be significant contributors to vitamin K status (Suttie, 2009). In rodents, 3 studies showed a dose-dependent relationship between phylloquinone intakes and brain MK-4 concentrations (Carrie et al., 2004; Okano et al., 2008; Thijssen and Drittij-Reijnders, 1994). However, in the present study, assessment of phylloquinone exposure through measures of phylloquinone intakes was not retained for 3 reasons: (1) phylloquinone intakes are challenging to assess because of the high intra-individual variability (Booth et al., 1995; Presse et al., 2011); (2) dietary intakes do not account for phylloquinone bioavailability, which varies among food sources (Presse et al., 2012); and (3) dietary intake assessment uses tools that rely on participants’ cognitive abilities (e.g., memory, executive functions), which could lead to equivocal conclusions when examining the relationship between diet and cognition. In this context, circulating phylloquinone has been considered as the most appropriate marker of phylloquinone exposure for the purpose of the present study. Indeed, phylloquinone intake is the primary determinant of serum phylloquinone concentration in humans (McKeown et al., 2002; Presse et al., 2012) and in rodents
2782
N. Presse et al. / Neurobiology of Aging 34 (2013) 2777e2783
(Ronden et al., 1998). Also, circulating phylloquinone has been correlated with functional biomarkers of vitamin K status such as the percentage of under-g-carboxylated osteocalcin (McKeown et al., 2002; Shea et al., 2009; Sokoll and Sadowski, 1996). Finally, we previously showed that a single measurement of serum phylloquinone in older adults is reproducible over 2 years and is strongly associated with phylloquinone intakes assessed over 12 months (Presse et al., 2012). The strengths of the present study include the use of a comprehensive cognitive battery assessing 4 cognitive domains, the relatively large sample size, and the robust analytical design, including the exploratory analysis of effect modifiers. Although the cross-sectional nature of our study constitutes a limitation, the temporality of the associations remains probable. First, we previously showed that a single measurement of serum phylloquinone concentration is an adequate indicator of long-term phylloquinone exposure in healthy older adults (Presse et al., 2012). Furthermore, using a smaller subsample of our participants, we found similar associations between verbal episodic memory scores and serum phylloquinone concentration determined 2 years before the cognitive evaluation. Although analyses included a number of potential confounders, we cannot rule out the possibility of residual confounding such as the ApoE genotype. In fact, carrying the ApoEε4 allele has been shown to exert small adverse effects on cognitive function in healthy individuals (Wisdom et al., 2011). However, the association with circulating phylloquinone is controversial, with studies reporting ApoEε4 carriers having lower (Kohlmeier et al., 1995; Saupe et al., 1993), higher (Yan et al., 2005), or similar (Beavan et al., 2005) phylloquinone concentrations as compared with non-carriers. In addition, although we controlled for overall quality of the diet (the C-HEI score), confounding by specific nutrients associated with cognition (e.g., vitamin E) should be ruled out in future studies. Limitations also include the fact that diet quality and alcohol consumption were assessed using a food frequency questionnaire administered at the NuAge Study baseline, whereas serum phylloquinone concentration was determined at the time of the cognitive evaluation. Finally, we acknowledge that our sample consisted of overselected individuals who had higher cognitive abilities than those not selected from the NuAge Study. Although this may limit generalization, the sample still included the oldest old (>80 years). To our knowledge, this is the first published study to examine the association between vitamin K and cognitive outcomes in healthy older adults. Notably, we found a positive association between serum phylloquinone and performances in verbal episodic memory, with a pattern suggesting a specific role in memory consolidation. In line with rodent models, our results are biologically plausible and add evidence for an expanding role of vitamin K from coagulation to cognition. The present report emphasizes the need to consider vitamin K as a nutritional factor in cognitive health in the aging population. Future studies will be needed to further our understanding of the role of vitamin K in brain and to examine whether treatment with vitamin K antagonists could affect cognition. Disclosure statement None of the authors has a conflict of interest. Acknowledgements We thank Doriane Kwong Hoi Fung for assistance with the serum phylloquinone analysis and Mélanie Drisdelle for coordinating the cognitive evaluation. The NuAge Study was supported by
the research grant MOP-62842 from the Canadian Institutes of Health Research (CIHR) and by the Quebec Network for Research on Aging funded by the Fonds de Recherche du Québec- Santé (FRQS; formerly Fonds de la Recherche en Santé du Québec). Access to NuAge databases and measurement of serum phylloquinone concentration were made possible through the NutCog study funded by the research grant MOP-82825 from the CIHR. Nancy Presse was supported by doctoral scholarships from the FRQS and from the CIHR. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.neurobiolaging. 2013.05.031. References Azharuddin, M.K., O’Reilly, D.S., Gray, A., Talwar, D., 2007. HPLC method for plasma vitamin K1: effect of plasma triglyceride and acute-phase response on circulating concentrations. Clin. Chem. 53, 1706e1713. Beavan, S.R., Prentice, A., Stirling, D.M., Dibba, B., Yan, L., Harrington, D.J., Shearer, M.J., 2005. Ethnic differences in osteocalcin gamma-carboxylation, plasma phylloquinone (vitamin K1) and apolipoprotein E genotype. Eur. J. Clin. Nutr. 59, 72e81. Belleville, S., Chatelois, J., Fontaine, F., Peretz, I., 2004. Mémoria: Batterie informatisée d’évaluation de la mémoire pour Mac et PC. Institut Universitaire de Gériatrie de Montréal, Montréal. Booth, S.L., Sokoll, L.J., O’Brien, M.E., Tucker, K., Dawson-Hughes, B., Sadowski, J.A., 1995. Assessment of dietary phylloquinone intake and vitamin K status in postmenopausal women. Eur. J. Clin. Nutr. 49, 832e841. Booth, S.L., Tucker, K.L., McKeown, N.M., Davidson, K.W., Dallal, G.E., Sadowski, J.A., 1997. Relationships between dietary intakes and fasting plasma concentrations of fat-soluble vitamins in humans. J. Nutr. 127, 587e592. Canada, H. 2007. Eating well with Canada’s Food Guide. Ó Her Majesty the Queen in Right of Canada, represented by the Minister of Health Canada, HC Publication no. 4651. Catalog no. H164e38/1-2007E. Canada, H.W., 1990. Nutrition recommendations: The report of the Scientific Review Committee. Minister of Supply and Services Canada. Catalog no. H-49e2/1990E. Carrie, I., Belanger, E., Portoukalian, J., Rochford, J., Ferland, G., 2011. Lifelong lowphylloquinone intake is associated with cognitive impairments in old rats. J. Nutr. 141, 1495e1501. Carrie, I., Portoukalian, J., Vicaretti, R., Rochford, J., Potvin, S., Ferland, G., 2004. Menaquinone-4 concentration is correlated with sphingolipid concentrations in rat brain. J. Nutr. 134, 167e172. Cocchetto, D.M., Miller, D.B., Miller, L.L., Bjornsson, T.D., 1985. Behavioral perturbations in the vitamin K-deficient rat. Physiol. Behav. 34, 727e734. Cohen, J.D., MacWhinney, B., Flatt, M., Provost, J., 1993. PsyScope: a new graphic interactive environment for designing psychology experiments. Behav. Res. Methods Instrum. Comput. 25, 257e271. Cook, Z., Kirk, S., Lawrenson, S., Sandford, S., 2005. Use of BMI in the assessment of undernutrition in older subjects: reflecting on practice. Proc. Nutr. Soc. 64, 313e317. Crivello, N.A., Casseus, S.L., Peterson, J.W., Smith, D.E., Booth, S.L., 2010. Age- and brain region-specific effects of dietary vitamin K on myelin sulfatides. J. Nutr. Biochem. 21, 1083e1088. Cutler, R.G., Kelly, J., Storie, K., Pedersen, W.A., Tammara, A., Hatanpaa, K., Troncoso, J.C., Mattson, M.P., 2004. Involvement of oxidative stress-induced abnormalities in ceramide and cholesterol metabolism in brain aging and Alzheimer’s disease. Proc. Natl. Acad. Sci. U.S.A 101, 2070e2075. Davidson, K.W., Sadowski, J.A., 1997. Determination of vitamin K compounds in plasma or serum by high-performance liquid chromatography using postcolumn chemical reduction and fluorimetric detection. Methods Enzymol. 282, 408e421. Davis, S., Laroche, S., 2006. Mitogen-activated protein kinase/extracellular regulated kinase signalling and memory stabilization: a review. Genes Brain Behav. 5 (Suppl 2), 61e72. Denisova, N.A., Booth, S.L., 2005. Vitamin K and sphingolipid metabolism: evidence to date. Nutr. Rev. 63, 111e121. Erkkila, A.T., Lichtenstein, A.H., Dolnikowski, G.G., Grusak, M.A., Jalbert, S.M., Aquino, K.A., Peterson, J.W., Booth, S.L., 2004. Plasma transport of vitamin K in men using deuterium-labeled collard greens. Metabolism 53, 215e221. Ferland, G., 2012. Vitamin K and the nervous system: an overview of its actions. Adv. Nutr. 3, 204e212. Gaudreau, P., Morais, J.A., Shatenstein, B., Gray-Donald, K., Khalil, A., Dionne, I., Ferland, G., Fulop, T., Jacques, D., Kergoat, M.J., Tessier, D., Wagner, R., Payette, H., 2007. Nutrition as a determinant of successful aging: description of the Quebec longitudinal study Nuage and results from cross-sectional pilot studies. Rejuvenation Res. 10, 377e386.
N. Presse et al. / Neurobiology of Aging 34 (2013) 2777e2783 Groupe de réflexion sur l’évaluation de la mémoire, Van der Linden, M., 2004. L’épreuve de rappel libre/rappel indicé à 16 items (RL/RI-16). L’évaluation des troubles de la mémoire: Présentation de quatre tests de mémoire épisodique (avec leur étalonnage), Solal, Marseille, pp. 25e47. Hall, J.G., Pauli, R.M., Wilson, K.M., 1980. Maternal and fetal sequelae of anticoagulation during pregnancy. Am. J. Med. 68, 122e140. He, X., Shen, L., Bjartell, A., Dahlback, B., 1995. The gene encoding vitamin Kdependent anticoagulant protein S is expressed in multiple rabbit organs as demonstrated by northern blotting, in situ hybridization, and immunohistochemistry. J. Histochem. Cytochem. 43, 85e96. Horwood, J.M., Dufour, F., Laroche, S., Davis, S., 2006. Signalling mechanisms mediated by the phosphoinositide 3-kinase/Akt cascade in synaptic plasticity and memory in the rat. Eur. J. Neurosci. 23, 3375e3384. Isaev, N.K., Stelmashook, E.V., Ruscher, K., Andreeva, N.A., Zorov, D.B., 2004. Menadione reduces rotenone-induced cell death in cerebellar granule neurons. Neuroreport 15, 2227e2231. Kohlmeier, M., Saupe, J., Drossel, H.J., Shearer, M.J., 1995. Variation of phylloquinone (vitamin K1) concentrations in hemodialysis patients. Thromb. Haemost. 74, 1252e1254. Kohlmeier, M., Saupe, J., Shearer, M.J., Schaefer, K., Asmus, G., 1997. Bone health of adult hemodialysis patients is related to vitamin K status. Kidney Int. 51, 1218e1221. Lamon-Fava, S., Sadowski, J.A., Davidson, K.W., O’Brien, M.E., McNamara, J.R., Schaefer, E.J., 1998. Plasma lipoproteins as carriers of phylloquinone (vitamin K1) in humans. Am. J. Clin. Nutr. 67, 1226e1231. Li, J., Lin, J.C., Wang, H., Peterson, J.W., Furie, B.C., Furie, B., Booth, S.L., Volpe, J.J., Rosenberg, P.A., 2003. Novel role of vitamin K in preventing oxidative injury to developing oligodendrocytes and neurons. J. Neurosci. 23, 5816e5826. Li, J., Wang, H., Rosenberg, P.A., 2009. Vitamin K prevents oxidative cell death by inhibiting activation of 12-lipoxygenase in developing oligodendrocytes. J. Neurosci. Res. 87, 1997e2005. McKeown, N.M., Jacques, P.F., Gundberg, C.M., Peterson, J.W., Tucker, K.L., Kiel, D.P., Wilson, P.W., Booth, S.L., 2002. Dietary and nondietary determinants of vitamin K biochemical measures in men and women. J. Nutr. 132, 1329e1334. Morley, J.E., Banks, W.A., 2010. Lipids and cognition. J. Alzheimers. Dis. 20, 737e747. NCEP ATP III, 2002. Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation. 106, 3143e3421. Neogi, T., Booth, S.L., Zhang, Y.Q., Jacques, P.F., Terkeltaub, R., Aliabadi, P., Felson, D.T., 2006. Low vitamin K status is associated with osteoarthritis in the hand and knee. Arthritis Rheum. 54, 1255e1261. Okano, T., Shimomura, Y., Yamane, M., Suhara, Y., Kamao, M., Sugiura, M., Nakagawa, K., 2008. Conversion of phylloquinone (vitamin K1) into menaquinone-4 (vitamin K2) in mice: two possible routes for menaquinone-4 accumulation in cerebra of mice. J. Biol. Chem. 283, 11270e11279. Presse, N., Gaudreau, P., Greenwood, C.E., Kergoat, M.J., Morais, J.A., Payette, H., Shatenstein, B., Ferland, G., 2012. A single measurement of serum phylloquinone is an adequate indicator of long-term phylloquinone exposure in healthy older adults. J. Nutr. 142, 1910e1916. Presse, N., Payette, H., Shatenstein, B., Greenwood, C.E., Kergoat, M.J., Ferland, G., 2011. A minimum of six days of diet recording is needed to assess usual vitamin K intake among older adults. J. Nutr. 141, 341e346. Presse, N., Shatenstein, B., Kergoat, M.J., Ferland, G., 2008. Low vitamin K intakes in community-dwelling elders at an early stage of Alzheimer’s disease. J. Am. Diet. Assoc. 108, 2095e2099. Prieto, A.L., O’Dell, S., Varnum, B., Lai, C., 2007. Localization and signaling of the receptor protein tyrosine kinase Tyro3 in cortical and hippocampal neurons. Neuroscience 150, 319e334. Prieto, A.L., Weber, J.L., Tracy, S., Heeb, M.J., Lai, C., 1999. Gas6, a ligand for the receptor protein-tyrosine kinase Tyro-3, is widely expressed in the central nervous system. Brain Res. 816, 646e661. Rey, A., 1959. Test de copie d’une figure complexe. Centre de psychologie appliquée, Manuel, Paris. Ronden, J.E., Drittij-Reijnders, M.J., Vermeer, C., Thijssen, H.H., 1998. Intestinal flora is not an intermediate in the phylloquinone-menaquinone-4 conversion in the rat. Biochim. Biophys. Acta 1379, 69e75.
2783
Sadowski, J.A., Hood, S.J., Dallal, G.E., Garry, P.J., 1989. Phylloquinone in plasma from elderly and young adults: factors influencing its concentration. Am. J. Clin. Nutr. 50, 100e108. Sakaue, M., Mori, N., Okazaki, M., Kadowaki, E., Kaneko, T., Hemmi, N., Sekiguchi, H., Maki, T., Ozawa, A., Hara, S., Arishima, K., Yamamoto, M., 2011. Vitamin K has the potential to protect neurons from methylmercury-induced cell death in vitro. J. Neurosci. Res. 89, 1052e1058. Sato, Y., Honda, Y., Hayashida, N., Iwamoto, J., Kanoko, T., Satoh, K., 2005. Vitamin K deficiency and osteopenia in elderly women with Alzheimer’s disease. Arch. Phys. Med. Rehabil. 86, 576e581. Saupe, J., Shearer, M.J., Kohlmeier, M., 1993. Phylloquinone transport and its influence on gamma-carboxyglutamate residues of osteocalcin in patients on maintenance hemodialysis. Am. J. Clin. Nutr. 58, 204e208. Shatenstein, B., Nadon, S., Godin, C., Ferland, G., 2005a. Development and validation of a food frequency questionnaire. Can. J. Diet. Pract. Res. 66, 67e75. Shatenstein, B., Nadon, S., Godin, C., Ferland, G., 2005b. Diet quality of Montrealarea adults needs improvement: estimates from a self-administered food frequency questionnaire furnishing a dietary indicator score. J. Am. Diet. Assoc. 105, 1251e1260. Shea, M.K., Benjamin, E.J., Dupuis, J., Massaro, J.M., Jacques, P.F., D’Agostino Sr., R.B., Ordovas, J.M., O’Donnell, C.J., Dawson-Hughes, B., Vasan, R.S., Booth, S.L., 2009. Genetic and non-genetic correlates of vitamins K and D. Eur. J. Clin. Nutr. 63, 458e464. Sokoll, L.J., Sadowski, J.A., 1996. Comparison of biochemical indexes for assessing vitamin K nutritional status in a healthy adult population. Am. J. Clin. Nutr. 63, 566e573. Spreen, O., Strauss, E., 1991. Profile of test results. A compendium of neuropsychological tests: Administration, norms, and commentary. Oxford University Press, New York, pp. 10e76. Spreen, O., Strauss, E., 1998. A compendium of neuropsychological tests: Administration, norms, and commentary, second ed. Oxford University Press, New York. Sundaram, K.S., Fan, J.H., Engelke, J.A., Foley, A.L., Suttie, J.W., Lev, M., 1996. Vitamin K status influences brain sulfatide metabolism in young mice and rats. J. Nutr. 126, 2746e2751. Sundaram, K.S., Lev, M., 1988. Warfarin administration reduces synthesis of sulfatides and other sphingolipids in mouse brain. J. Lipid Res. 29, 1475e1479. Suttie, J.W., 2009. Vitamin K in Health and Disease. CRC Press Taylor & Francis Group, New York. Thijssen, H.H., Drittij-Reijnders, M.J., 1994. Vitamin K distribution in rat tissues: dietary phylloquinone is a source of tissue menaquinone-4. Br. J. Nutr. 72, 415e425. Thijssen, H.H., Drittij-Reijnders, M.J., 1996. Vitamin K status in human tissues: tissue-specific accumulation of phylloquinone and menaquinone-4. Br. J. Nutr. 75, 121e127. Thijssen, H.H., Drittij-Reijnders, M.J., Fischer, M.A., 1996. Phylloquinone and menaquinone-4 distribution in rats: synthesis rather than uptake determines menaquinone-4 organ concentrations. J. Nutr. 126, 537e543. Tsaioun, K.I., Denisova, N.A., Obin, M., Joseph, J., 2000. Novel growth factor Gas6, phosphatidylserine and their age-related changes in the rat brain. Neurosci. Res. Commun. 26, 113e122. Wang, L.Y., Bates, C.J., Yan, L., Harrington, D.J., Shearer, M.J., Prentice, A., 2004. Determination of phylloquinone (vitamin K1) in plasma and serum by HPLC with fluorescence detection. Clin. Chim. Acta 347, 199e207. Washburn, R.A., McAuley, E., Katula, J., Mihalko, S.L., Boileau, R.A., 1999. The Physical Activity Scale for the Elderly (PASE): evidence for validity. J. Clin. Epidemiol. 52, 643e651. Wechsler, D., 1997. Weschsler Adult Intelligence Scale, Third (WAIS-III) ed. Psychological Corporation, San Antonio, TX. Wisdom, N.M., Callahan, J.L., Hawkins, K.A., 2011. The effects of apolipoprotein E on non-impaired cognitive functioning: a meta-analysis. Neurobiol. Aging 32, 63e74. Yan, L., Zhou, B., Nigdikar, S., Wang, X., Bennett, J., Prentice, A., 2005. Effect of apolipoprotein E genotype on vitamin K status in healthy older adults from China and the UK. Br. J. Nutr. 94, 956e961. Yesavage, J.A., Brink, T.L., Rose, F.L., Lum, O., Huang, V., Adey, M., Leirer, V.O., 1983. Development and validation of a geriatric depression screening scale: a preliminary report. J. Psychiatr. Res. 17, 37e49.
Contents lists available at SciVerse ScienceDirect
Neurobiology of Aging journal homepage: www.elsevier.com/locate/neuaging
Vitamin K status and cognitive function in healthy older adults Nancy Presse a, b, Sylvie Belleville a, c, Pierrette Gaudreau d, e, Carol E. Greenwood f, Marie-Jeanne Kergoat a, d, Jose A. Morais g, Hélène Payette h, Bryna Shatenstein a, b, Guylaine Ferland a, b, * a
Centre de recherche, Institut Universitaire de Gériatrie de Montréal, Montréal, Quebec, Canada Département de Nutrition, Université de Montréal, Montréal, Quebec, Canada c Département de Psychologie, Université de Montréal, Montréal, Quebec, Canada d Département de Médecine, Université de Montréal, Montréal, Quebec, Canada e Centre de Recherche, Centre Hospitalier de l’Université de Montréal, Montréal, Quebec, Canada f Kunin Lunenfeld Applied and Evaluative Unit, Baycrest and Department of Nutritional Sciences, University of Toronto, Toronto, Ontario, Canada g McGill Nutrition and Food Science Centre and Department of Medicine, Division of Geriatrics, McGill University Health Centre and School of Dietetics and Human Nutrition, McGill University, Montréal, Quebec, Canada h Centre de recherche sur le vieillissement, Institut Universitaire de Gériatrie de Sherbrooke and Faculté de Médecine et des Sciences de la Santé, Université de Sherbrooke, Sherbrooke, Quebec, Canada b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 19 December 2012 Received in revised form 22 April 2013 Accepted 30 May 2013 Available online 11 July 2013
Evidence is accumulating that vitamin K could have a role in cognition, especially in aging. Using data from the Québec Longitudinal Study on Nutrition and Successful Aging (NuAge), a cross-sectional analysis was conducted to examine the associations between vitamin K status, measured as serum phylloquinone concentrations, and performance in verbal and non-verbal episodic memory, executive functions, and speed of processing. The sample included 320 men and women aged 70 to 85 years who were free of cognitive impairment. After adjustment for covariates, higher serum phylloquinone concentration (log-transformed) was associated with better verbal episodic memory performances (F ¼ 2.43, p ¼ 0.048); specifically with the scores (Z-transformed) on the second (b ¼ 0.47; 95% confidence interval [CI] ¼ 0.13e0.82), third (b ¼ 0.41; 95% CI ¼ 0.06e0.75), and 20-minute delayed (b ¼ 0.47; 95% CI ¼ 0.12 e0.82) free recall trials of the RL/RI-16 Free and Cued Recall Task. No associations were found with nonverbal episodic memory, executive functions, and speed of processing. Our study adds evidence to the possible role of vitamin K in cognition during aging, specifically in the consolidation of the memory trace. Ó 2013 Elsevier Inc. All rights reserved.
Keywords: Vitamin K Phylloquinone Cognition Aging
1. Introduction Vitamin K is a fat-soluble compound found predominantly in green leafy vegetables and, to lesser extent, other vegetables, fruits, and some vegetable oils. It was originally discovered for its role in blood coagulation, activating the vitamin Kedependent clotting factors through its function as a cofactor of the g-glutamylcarboxylase. The discovery of additional vitamin K-dependent proteins (e.g. osteocalcin) and functions independent of the gcarboxylation has expanded the role of vitamin K beyond coagulation (Suttie, 2009). A role in the human brain was first suspected after cases of central nervous system abnormalities in newborns exposed in utero to vitamin K antagonists (Hall et al., 1980). In rat and human brain tissues, high vitamin K levels in the form of * Corresponding author at: Centre de recherche, Institut Universitaire de Gériatrie de Montréal, 4565 Queen Mary, Montréal, Quebec H3W 1W5, Canada. Tel.: þ1 514 340 2800 3236; fax: þ1 514 340 2801. E-mail address: [email protected] (G. Ferland). 0197-4580/$ e see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.neurobiolaging.2013.05.031
menaquinone-4 (MK-4) have been reported (Carrie et al., 2004; Thijssen and Drittij-Reijnders, 1994, 1996). In rodents, brain MK-4 was shown to respond, in a dose-dependent manner, to intakes of vitamin K, whether provided in a natural dietary form (phylloquinone) or a synthetic form (Carrie et al., 2004; Okano et al., 2008; Thijssen and Drittij-Reijnders, 1994). Our understanding of the vitamin K functions in brain is growing (Ferland, 2012). The vitamin K-dependent proteins Gas6 (growth arrest-specific gene 6) and protein S are expressed in the central nervous system (He et al., 1995; Prieto et al., 1999). Protein S is notably known to possess neuroprotective effects during hypoxic/ ischemic injury, whereas Gas6 is now recognized as an important regulator of cell survival, cell growth, and myelination processes (Ferland, 2012). Furthermore, vitamin K participates in the metabolism of sphingolipids, a major constituent of the myelin sheath and neuronal membranes, also involved in important molecular events such as cell signaling (Denisova and Booth, 2005; Ferland, 2012). Finally, the K vitamers were also shown to possess neuroprotective effects against oxidative-induced death of primary
2778
N. Presse et al. / Neurobiology of Aging 34 (2013) 2777e2783
cultures of rat oligodendrocytes, rat neurons, and human neuroblastoma cells by preventing the accumulation of reactive oxygen species (Isaev et al., 2004; Li et al., 2003, 2009; Sakaue et al., 2011). Consistent with mechanistic studies, in vivo rodent models have provided evidence that vitamin K status can modulate brain sphingolipids, behavior, and cognitive performances (Carrie et al., 2004, 2011; Cocchetto et al., 1985; Crivello et al., 2010; Sundaram et al., 1996). Interestingly, current research suggests that vitamin K exerts its actions in aging. Notably, higher concentrations of MK-4 in the hippocampus and cortex were shown to be correlated with higher myelin sulfatides in adult and aged rats, but not in younger animals (Crivello et al., 2010). Similarly, 20-month-old rats fed a vitamin Kerich or eadequate diet throughout their lives had better spatial learning performances in the Morris Water Maze task than those fed a lowevitamin K diet, a result not observed in 6-monthold rats (Carrie et al., 2011). A vulnerability of the aging brain to vitamin K status is also strengthened by studies reporting a marked age-dependent loss of Gas6 in rat brain (Tsaioun et al., 2000) and sphingolipid alterations in brain tissues of aged mice and older adults with Alzheimer’s disease (Cutler et al., 2004). The potential role of vitamin K in cognitive function of older adults has not yet been examined. Thus far, two cross-sectional studies reported a lower vitamin K status in subjects in the early and late stages of Alzheimer’s disease in comparison with controls (Presse et al., 2008; Sato et al., 2005). In the present study, we examined the associations between serum phylloquinone concentration and performances in verbal and non-verbal episodic memory, executive functions, and speed of processing in 320 healthy older adults selected from the Québec Longitudinal Study on Nutrition and Successful Aging (NuAge). To our knowledge, this is the first study to investigate the potential role of vitamin K in specific cognitive domains. 2. Methods The NuAge Study is a prospective cohort study of nutritional factors of successful aging. The methodology has been detailed elsewhere (Gaudreau et al., 2007). Briefly, community-dwelling men and women aged 67 to 84 years were randomly selected from the Québec Medicare Database. To be include in the NuAge Study, participants had to be in good general health, free of disabilities in activities of daily living, free of cognitive impairment as established by a Modified Mini-Mental State Examination (3MS) score of >79, and free of the following health conditions: Class II heart failure, chronic obstructive pulmonary disease requiring home oxygen therapy or oral steroids, inflammatory digestive diseases, and cancer treated by radiation therapy, chemotherapy, or surgery in the previous 5 years. After a telephone interview and a clinical examination, 1793 participants were recruited in 2003 to 2005. Follow-up interviews were performed annually up to 3 years after recruitment. The present study is based on an embedded subsample of 464 French-speaking participants who underwent a cognitive evaluation 1.9 to 4.0 years (mean SD, 3.0 0.6 years) after their recruitment in the NuAge cohort (2006e2008). These participants had to have 3MS scores of >85 to ensure that they could complete the detailed cognitive evaluation, and had to be free of health conditions that could impair cognition (self-reported), namely Parkinson’s disease, history of stroke or cerebral hemorrhage, Alzheimer’s disease or other dementia, Guillain-Barré syndrome, epilepsy, head trauma resulting in unconsciousness, cardiac arrest with resuscitation, brain tumor or metastasis, central nervous system infection (e.g., meningitis), multiple sclerosis, or CO2 or methanol intoxication. All participants provided informed consents approved by the ethics committees of both testing sites, the
Instituts Universitaires de Gériatrie de Montréal and Sherbrooke (QC, Canada). Two health professionals, including a geriatrician, thoroughly reviewed each of the 464 participant files to ensure a sample of older adults free of cognitive impairment who were unlikely to have changed their diet because of underlying cognitive problems. As a result, 36 participants were excluded, as they were found to have Parkinson’s disease (n ¼ 2), actively treated epilepsy (n ¼ 3), Alzheimer’s disease or other dementia (n ¼ 7), or history of stroke/ cerebral hemorrhage (n ¼ 24). We also excluded those treated with vitamin K antagonists (warfarin or acenocoumarol; n ¼ 28), and subjects with incomplete dataset (n ¼ 65). Finally, 15 participants were excluded because they had abnormally low cognitive performances for healthy individuals. These 15 participants were identified by studentized (jackknife) residuals less than 3.0 after cognitive scores were regressed on testing research assistant, age, sex, and number of years of education. The final sample included 320 of the 464 participants (see Supplemental Fig. 1 for a detailed flowchart). 2.1. Cognitive assessment The cognitive test battery was administered by trained research assistants and consisted of 6 tests covering 4 cognitive domains, namely verbal and non-verbal episodic memory, executive functions, and speed of processing (see Supplemental Text 1 for a detailed description). From the 6 cognitive tests, 13 scores were calculated and used to assess performances on each cognitive domain. Verbal episodic memory was assessed using scores of the 3 immediate free recall trials and the 20-minute delayed free recall of the RL/RI-16 Free and Cued Recall Task (RL/RI-16 FCRT) (Groupe de réflexion sur l’évaluation de la mémoire and Van der Linden, 2004); non-verbal episodic memory was tested by scores of the 3-minute and 20-minute recalls of the Rey Complex-Figure (Rey, 1959); executive functions were assessed by those of the Rey Complex-Figure copying, the third plate of the Stroop Test (Spreen and Strauss, 1991), the total number of letters recalled in the interference conditions (10-, 20-, and 30-second delays) of the Adapted BrownePeterson procedure (Belleville et al., 2004; Spreen and Strauss, 1998), and the total number of completed targets of the Weschsler Adult Intelligence ScaleeIII (WAIS-III) Digit Symbol-Coding subtest (Wechsler, 1997); finally, speed of processing was evaluated using scores from the 2 first plates of the Stroop Test and mean initiation time at the Choice-Reaction Time (Cohen et al., 1993). 2.2. Serum phylloquinone, triglycerides, and total cholesterol concentrations Blood sampling, processing, and storing in the NuAge Study followed standard operating procedures (Gaudreau et al., 2007). At annual interviews, a 50-mL blood sample was obtained between 7:30 and 8:30 AM after an overnight fast. Specifically, participants were instructed not to eat, drink (except reasonable amount of water), smoke, or chew gum after 19:00 the evening before the blood sampling. Blood specimens were immediately stored at 4 C and protected from light until processing. Serum was extracted, aliquoted in 1.5-mL, opalescent-colored Eppendorf polypropylene tubes. Serum samples were stored at 20 C in opaque cardboard for few hours and then at 80 C for long-term storage. Overall, light exposure was limited to periods of sample processing, for a total light exposure time of about 100 minutes. Selected blood samples were collected from less than 11.5 months before to 2.5 months after the cognitive evaluation (mean SD, 3.2 3.2 months); that is, between June 2006 and May 2008. In Canada, over-the-counter supplements containing vitamin K
N. Presse et al. / Neurobiology of Aging 34 (2013) 2777e2783
were not available during that period, and none of the 320 participants reported taking vitamin K as a prescribed drug. Thus, circulating phylloquinone reflects dietary phylloquinone intakes. The 480-mL serum samples used for phylloquinone analyses had not previously been thawed and refrozen. Samples were shipped on dry ice in opaque cardboard. Phylloquinone analyses were conducted in July to August 2010 in Dr Guylaine Ferland’s laboratory using standardized HPLC procedures (Davidson and Sadowski, 1997; Wang et al., 2004). Time between blood sampling and phylloquinone analysis was 3.13 0.48 years (mean SD). Quality was controlled using an external serum standard. The intra-assay coefficient of variation (CV) was 9.2% and 6.0% for phylloquinone concentrations of 0.2 and 0.8 nmol/L, respectively, and did not vary over time. The detection limit was 0.03 nmol/L, and none of the 320 serum samples was below this limit. Phylloquinone in blood is mostly carried by triglyceride-rich lipoproteins and, to lesser extent, by the low- and high-density lipoprotein fractions, making the blood lipid profile an important determinant of serum/plasma phylloquinone concentration (Azharuddin et al., 2007; Erkkila et al., 2004; Lamon-Fava et al., 1998; McKeown et al., 2002; Presse et al., 2012; Sadowski et al., 1989; Shea et al., 2009). Given that blood lipid profile was also associated with cognitive performance (Morley and Banks, 2010), triglyceridemia and total cholesterol were considered as covariates in the present study. The overnight fasting concentrations of serum triglycerides and total cholesterol were determined at the Centre Hospitalier de l’Université de Montréal clinical biochemistry laboratory, on a Roche/Hitachi Cobas c311 System analyzer using the TRIGL and CHOL2 kits, respectively (Roche Diagnostic Corp, Indianapolis, IN). 2.3. Other covariates A structured interview conducted annually in the NuAge Study provided self-reported information on race (white vs. other), sex, age, number of years of education, household income (in Canadian dollars), medical history, medications, and lifestyle at the time of the cognitive evaluation. History of hypertension was based on selfreported diagnoses. History of diabetes was determined from medication use (e.g., insulin, oral hypoglycemic agents) or fasting glycemia 7.0 mmol/L. Number of depressive symptoms was determined using the Geriatric Depression Scale (GDS) (Yesavage et al., 1983). Weight and height were measured, and body mass index (BMI) was calculated as weight in kilograms divided by the square of height in meters. Ponderal status was described as normal (BMI <25 kg/m2), overweight (BMI 25 to <30 kg/m2), and obese (BMI 30 kg/m2). Lifestyle was characterized by smoking status (never vs. current/former smoker), alcohol consumption, physical activity, use of vitamin/mineral supplements (regular vs. occasional/ never user), and diet quality. Mean daily alcohol intake (g) was determined by a validated semi-quantitative food-frequency questionnaire completed at the NuAge baseline, which assesses usual eating habits in the past year (Shatenstein et al., 2005a). Weekly consumption of beer, wine, and spirits was expressed as number of drinks (0, >0 to <5, and 5 drinks per week) based on a standard drink of 13.6 g of alcohol. Usual physical activity was assessed using the Physical Activity Scale for the Elderly (PASE), with higher scores indicative of higher physical activity level (Washburn et al., 1999). Physical activity level was then categorized as PASE score <75, 75 to <125, and 125. Diet quality was assessed using the CanadianHealthy Eating Index (C-HEI), a 100-point index with higher scores indicating greater diet quality (Shatenstein et al., 2005b). The C-HEI was calculated from food and nutrient intakes estimated by the food-frequency questionnaire to assess adherence to the 2007 Canada Food Guide and the 1990 Nutrition Recommendations for Canadians (Canada, 2007; Canada, 1990).
2779
2.4. Statistical analyses Cognitive test scores measured as “time” (e.g., seconds in the Rey-Complex Figure) were log-transformed because of departures from normality, and were multiplied by 1 so that an increasing score indicated better performance for all tests. All scores were Ztransformed to allow comparison across cognitive tests. To examine the associations between serum phylloquinone concentration and cognitive performance, multivariate linear regressions were used to simultaneously analyze scores relating to each of the 4 cognitive domains assessed. In the base model, serum phylloquinone (log-transformed) was regressed on cognitive performance adjusting for age, sex, education, serum triglycerides, serum total cholesterol, and testing research assistant. The “testing research assistant” was modeled as a dummy variable, using the research assistant who tested the largest number of participants as reference. In the full model, other covariates chosen a priori were added to the base model: annual household income, hypertension (binary), diabetes (binary), BMI (categorical), number of depressive symptoms, smoking status (binary), alcohol consumption (categorical), physical activity level (categorical), regular use of vitamin/mineral supplements (binary), and diet quality index. Because the sample was basically white (>99%), race was not considered as a covariate. The test assumptions were verified including multivariate normality (Mahalanobis distance), linearity, multicollinearity, and homogeneity of varianceecovariance matrices, with no violations noted. Wilks’ lambda statistics were reported for both the base and full models. Significant results (p < 0.05) on multivariate analyses (full model only) were analyzed further using univariate linear regressions for each of the cognitive scores. Bonferroni adjustment was used to control for type I error. Exploratory analysis was performed for effect modification by sex, age, education, and triglycerides. For these analyses, we evaluated the significance of the cross-product term that resulted from multiplying serum phylloquinone concentration by each variable, 1 variable at a time. Adequacy of the full models was evaluated by the following: (1) examining added variable plots showing adjusted regression lines; (2) comparing these lines with loess regression lines (Epanechnikov kernel function, 50% fit); (3) assessing normality of residuals; and (4) plotting residuals against predicted values. To evaluate the magnitude of effect for significant associations (two-tailed p < .05), test scoreeadjusted means were calculated at the 2.5th, 50th, and 97.5th percentiles of serum phylloquinone concentration. The ManneWhitney U and c2 tests were used to compare the sample to participants not selected from the NuAge cohort. Characteristics of the participants are presented as mean ( SD) or as percentages, and their correlation with serum phylloquinone was determined using regression analysis. All statistical analyses were performed using IBM SPSS Statistics 21 software (SPSS Inc, Chicago, IL). 3. Results The 320 selected participants from the NuAge cohort were younger, more educated, had greater household income, had higher global cognitive scores, had fewer depressive symptoms, were more likely to consume alcohol, and tended to be more physically active than those not selected (Supplemental Table 1). However, they did not differ by sex, race, health and smoking status, use of vitamin/ mineral supplements, or diet quality. At the time of the cognitive evaluation, participants were aged 70.0 to 85.5 years and had 3MS scores indicative of normal cognitive functioning (mean SD, 94.56 4.23; range, 82 to 100), with only 12.5% of participants having a score <90. Mean BMI,
2780
N. Presse et al. / Neurobiology of Aging 34 (2013) 2777e2783
Table 1 Characteristics of the subjects (n ¼ 320) and their correlation with the serum phylloquinone concentration at the time of the cognitive evaluation in the NuAge Study Characteristic Sociodemographic Sex (% female) Age (y) Race (% white) Education (y) Household income (10,000 CAD) Health BMI (kg/m2) Ponderal status (%) Overweight (BMI 25 to < 30 kg/m2) Obese (BMI 30 kg/m2) Diabetes (%) Hypertension (%) GDS (number of depressive symptoms) Serum triglycerides (mmol/L) Serum total cholesterol (mmol/L) Lifestyle Former or current smoker (%) Alcohol consumption (%)c 0 drink/week >0 to <5 drinks/week 5 drinks/week PASE score Physical activity level (%) PASE score <75 PASE score 75 to <125 PASE score 125 Regular use of vitamin/mineral supplements (%) C-HEIc Serum phylloquinone concentration (nmol/L)
Value
Correlation coefficienta
54.1 76.4 3.8b 99.4 12.85 4.56 4.33 2.31
0.15** 0.02 0.02 0.03 0.03
27.8 4.8
0.02
49.7 23.8 10.3 53.4 4.25 3.82 1.65 0.82 5.33 1.07
0.03 0.03 0.11 0.08 0.02 0.40*** 0.30***
43.4
0.06
23.8 41.9 34.9 95.5 45.2
— 0.03 0.06 0.03
35.9 41.9 22.2 59.7
— 0.03 0.02 0.06
79.5 8.6 1.45 1.80
0.04 —
** p < 0.01; *** p < 0.001. Key: BMI, body mass index; CAD, Canadian dollars; C-HEI, Canadian-Healthy Eating Index; GDS, Geriatric Depression Scale; PASE, Physical Activity Scale for the Elderly. a Correlation coefficients were determined by regression analysis with serum phylloquinone concentration (log-transformed) as dependent variable. Partial correlation coefficients for BMI categories, alcohol consumption, and physical activity levels are presented. b Mean SD (all such values). c Computed using data collected by a food frequency questionnaire administered at baseline in the NuAge Study.
triglycerides, and total cholesterol were within the acceptable range for older adults (Cook et al., 2005; NCEP ATP III, 2002) (Table 1). Most participants had healthy lifestyle habits as shown by the presence of only 13 current smokers, low to moderate alcohol consumption, regular use of vitamin/mineral supplements, and the fact that 49.4% of participants had C-HEI scores >80, indicative of “good” diet quality (Shatenstein et al., 2005b). Serum phylloquinone concentrations (median, 1.06 nmol/L; range, 0.08e23.57) were comparable with those found in healthy middle-aged and older adults in the United States (Booth et al., 1995; Booth et al., 1997; McKeown et al., 2002; Neogi et al., 2006; Sadowski et al., 1989; Shea et al., 2009). Serum phylloquinone concentrations were higher in women (median, 1.12 nmol/L; range, 0.23e23.57) than men (median, 0.97 nmol/L; range, 0.08e6.85) and were significantly correlated with triglycerides and total cholesterol. Mean ( SD) cognitive test scores are presented in Supplemental Table 2.
Table 2 Results from the multivariate linear regressions of test scores assessing four cognitive domains on serum phylloquinone concentrations in older adults from the NuAge Study (n ¼ 320) Cognitive domainsa
Verbal episodic memoryd Non-verbal episodic memorye Executive functionsf Speed of processingg
Base modelb
Full modelc
Wilks’ lambda
F (df1, df2)
Wilks’ lambda
F (df1, df2)
0.965 0.992
2.76* (4, 303) 1.30 (2, 305)
0.968 0.994
2.43* (4, 290) 0.90 (2, 292)
0.998 0.989
0.19 (4, 304) 1.14 (3, 305)
0.997 0.993
0.21 (4, 291) 0.67 (3, 292)
* p < 0.05. a All cognitive scores were Z-transformed and standardized for direction. b The base model was adjusted for age (continuous), sex (binary), testing research assistant (categorical), education (continuous), serum triglycerides (log-transformed continuous), and total cholesterol (continuous). c The full model was adjusted for household income (continuous), body mass index (categorical), diabetes (binary), hypertension (binary), number of depressive symptoms, smoking status (binary), alcohol consumption (categorical), physical activity (categorical), vitamin/mineral supplement use (binary), and diet quality index (continuous), in addition to the variables included in the base model. d Based on 4 scores: the 4 free recall trials of the RL/RI-16 Free and Cued Recall Task. e Based on 2 scores: the 3-minute and 20-minute recalls of the Rey ComplexFigure. f Based on 4 scores: the Rey Complex-Figure copying, the 3rd plate of the Stroop Test, the Adapted Brown-Peterson procedure, and the Weschsler Adult Intelligence ScaleeIII Digit Symbol-Coding subtest. g Based on 3 scores: the 2 first plates of the Stroop Test and the Choice-Reaction Time.
although a decrease in its magnitude was observed in comparison with the base model. No association was found with non-verbal episodic memory, executive functions, or speed of processing. When the verbal episodic memory scores were considered separately (Table 3), significant associations were found with the second, the third (marginally considering the Bonferroni adjustement), and the 20-minute delayed free recall trials of the RL/RI-16 FCRT, whereas no associations was observed with the first trial of the RL/RI-16 FCRT. Exploratory analysis did not reveal any consistent evidence of effect modification by sex, age, education, and triglycerides. Analysis of the magnitude of effect in the full models showed that an increase in serum phylloquinone concentration from 0.27 to 1.06 nmol/L (from the 2.5th to 50th percentile) was associated with an increase of 8.2%, 6.1%, and 6.4% in performances on the second, the third, and the 20-minute delayed free recall trials of the RL/RI-16 FCRT, respectively (Fig. 1). However, because the relationship is a logarithmic function, serum phylloquinone had to increase from 1.06 to 5.33 nmol/L (from the 50th to 97.5th percentile) to observe a similar improvement ([6%e8%) in free recall performances. For 203 of the 320 participants, serum specimens drawn 2 years before the cognitive evaluation were available. Therefore, we determined phylloquinone concentrations on these samples and conducted supplemental regression analyses to examine the temporality of the cross-sectional associations observed with the 320 participants. Although we could use only a limited number of confounding factors due to the sample size, results of these supplemental analyses were similar, that is, serum phylloquinone concentration was significantly associated with the second and third free recall trials but not with the first one (see Supplemental Table 3).
3.1. Serum phylloquinone concentration and cognitive performances
4. Discussion
In the base model, higher serum phylloquinone was significantly associated with better performance in verbal episodic memory (Table 2). In the full model, the association remained significant,
The present study is important, because it provides support for an emerging role of vitamin K in the aging brain. Specifically, we showed that serum phylloquinone was positively associated with
N. Presse et al. / Neurobiology of Aging 34 (2013) 2777e2783 Table 3 Results from the univariate linear regressions of verbal episodic memory test scores on serum phylloquinone concentrations in older adults from the NuAge Study (n ¼ 320) Cognitive Test
RL/RI-16 Free and Cued Recall Taskc Free recall, Trial 1 Free recall, Trial 2 Free recall, Trial 3 20-Minute delayed free recall
Base modela
Full modelb
b-coefficients for serum phylloquinone (95% CI)
b-coefficients for serum phylloquinone (95% CI)
0.24 0.49 0.43 0.51
0.21 0.47 0.41 0.47
(0.12 to 0.59) (0.15 to 0.83)** (0.09 to 0.77)* (0.16 to 0.85)**
(0.14 to 0.57) (0.13 to 0.82)** (0.06 to 0.75)* (0.12 to 0.82)**
* p < 0.05; ** p < 0.01. a The base model was adjusted for the RL/RI-16 word list version (binary), age (continuous), sex (binary), testing research assistant (categorical), education (continuous), serum triglycerides (log-transformed continuous), and total cholesterol (continuous). b The full model was adjusted for household income (continuous), body mass index (categorical), diabetes (binary), hypertension (binary), number of depressive symptoms, smoking status (binary), alcohol consumption (categorical), physical activity (categorical), vitamin/mineral supplement use (binary), diet quality index (continuous), in addition to the variables included in the base model. c All cognitive scores were Z-transformed and standardized for direction, so a positive coefficient means that test performance enhances with increasing serum phylloquinone concentration (log-transformed).
performance in verbal episodic memory, while being unrelated to non-verbal episodic memory, executive functions, and speed of processing. Episodic memory refers to the memory of events within their spatio-temporal context. For example, remembering where one’s keys were last left relies on episodic memory. It is thus important to everyday life. Episodic memory processes include encoding (when the information to memorize is first encountered) and memory consolidation (when the memory trace is stabilized for long-term storage). Interestingly, our findings advocate for a specific role of vitamin K in memory consolidation. Indeed, a higher serum phylloquinone concentration was associated with better performances on the second, the third, and the 20-minute delayed free recall trials of the verbal episodic memory test while not being associated with the first one. This pattern indicates that the higher the vitamin K status, the more the participants benefited from the successive learning trials. The fact that no association was observed with the
16
Number of words recalled
14 12 10 8 6
5.33 nmol/L 1.06 nmol/L
4
0.27 nmol/L 2 0 Trial 1
Trial 2
Trial 3
20-min delayed trial
Free recall trials of the RL/RI-16 Free and Cued Recall Task
Fig. 1. Performance adjusted-means ( SEM) across free recall trials of the RL/RI-16 Free and Cued Recall Task for serum phylloquinone concentrations at the 2.5th, 50th, and 97.5th percentiles. Means were adjusted for testing research assistant, RL/RI-16 word-list version, age, sex, education, household income, diabetes, hypertension, number of depressive symptoms, body mass index, alcohol consumption, smoking status, physical activity, use of vitamin/mineral supplements, diet quality, serum triglycerides, and total cholesterol.
2781
performances in the non-verbal episodic memory test (Rey Complex-Figure recalls), where the abstract figure is presented only once, also advocates for a specific link between vitamin K and consolidation of the memory trace. A specific role of vitamin K in memory consolidation is in line with a recent study in rodents (Carrie et al., 2011). When rats were subjected to a 5-day spatial learning task (Morris Water Maze), those fed a low vitamin K diet throughout their lives required more exposure time to the visual cues to achieve performances similar to those fed a diet adequate or rich in vitamin K. Moreover, rats fed the lowevitamin K diet presented an altered sphingolipid profile in their hippocampus (Carrie et al., 2011), the cerebral region responsible for memory consolidation. The possible action of vitamin K in the hippocampus is also strengthened by studies on the protein Gas6. Specifically, the receptor Tyro3, one of the Gas6 ligands, is highly expressed in the hippocampal region (Prieto et al., 2007), and its activation by Gas6 results in molecular events known to underlie memory consolidation (Davis and Laroche, 2006; Horwood et al., 2006; Prieto et al., 2007). A role for vitamin K in the aging brain surely raises questions for those treated with vitamin K-antagonists. Although limited, available data do suggest that vitamin K functions in the brain could be altered by these drugs. Notably, warfarin-treated rodents were shown to present decreased concentrations of brain MK-4 (Thijssen et al., 1996) and alterations in brain sphingolipids (Sundaram and Lev, 1988). Whether treatment with vitamin K-antagonists in humans could increase the risk of cognitive problems has, to our knowledge, not yet been examined. Nonetheless, given that warfarin-treated individuals are often told to limit their consumption of vitamin K-rich foods, which could further deplete their brain’s vitamin K status, the implications of a role of vitamin K in cognition could be significant. In the present study, the association between episodic memory performance and serum phylloquinone concentration was found to be a logarithmic function such that the rate of improvement of the scores (number of words recalled) slowed down considerably once the phylloquinone concentration reached approximately 1.0 nmol/ L. Interestingly, 2 studies reported similar “threshold” effects at approximately 1.0 nmol/L when examining associations between vitamin K status and bone health (Kohlmeier et al., 1997; Neogi et al., 2006). Reasons explaining this threshold effect remain to be determined, but this observation should be taken into consideration when planning studies on the role of vitamin K in cognition. Phylloquinone is the main K vitamer in diet and available data indicate that other K vitamers found in the North American diet as well as menaquinones synthesized by gut bacteria would not be significant contributors to vitamin K status (Suttie, 2009). In rodents, 3 studies showed a dose-dependent relationship between phylloquinone intakes and brain MK-4 concentrations (Carrie et al., 2004; Okano et al., 2008; Thijssen and Drittij-Reijnders, 1994). However, in the present study, assessment of phylloquinone exposure through measures of phylloquinone intakes was not retained for 3 reasons: (1) phylloquinone intakes are challenging to assess because of the high intra-individual variability (Booth et al., 1995; Presse et al., 2011); (2) dietary intakes do not account for phylloquinone bioavailability, which varies among food sources (Presse et al., 2012); and (3) dietary intake assessment uses tools that rely on participants’ cognitive abilities (e.g., memory, executive functions), which could lead to equivocal conclusions when examining the relationship between diet and cognition. In this context, circulating phylloquinone has been considered as the most appropriate marker of phylloquinone exposure for the purpose of the present study. Indeed, phylloquinone intake is the primary determinant of serum phylloquinone concentration in humans (McKeown et al., 2002; Presse et al., 2012) and in rodents
2782
N. Presse et al. / Neurobiology of Aging 34 (2013) 2777e2783
(Ronden et al., 1998). Also, circulating phylloquinone has been correlated with functional biomarkers of vitamin K status such as the percentage of under-g-carboxylated osteocalcin (McKeown et al., 2002; Shea et al., 2009; Sokoll and Sadowski, 1996). Finally, we previously showed that a single measurement of serum phylloquinone in older adults is reproducible over 2 years and is strongly associated with phylloquinone intakes assessed over 12 months (Presse et al., 2012). The strengths of the present study include the use of a comprehensive cognitive battery assessing 4 cognitive domains, the relatively large sample size, and the robust analytical design, including the exploratory analysis of effect modifiers. Although the cross-sectional nature of our study constitutes a limitation, the temporality of the associations remains probable. First, we previously showed that a single measurement of serum phylloquinone concentration is an adequate indicator of long-term phylloquinone exposure in healthy older adults (Presse et al., 2012). Furthermore, using a smaller subsample of our participants, we found similar associations between verbal episodic memory scores and serum phylloquinone concentration determined 2 years before the cognitive evaluation. Although analyses included a number of potential confounders, we cannot rule out the possibility of residual confounding such as the ApoE genotype. In fact, carrying the ApoEε4 allele has been shown to exert small adverse effects on cognitive function in healthy individuals (Wisdom et al., 2011). However, the association with circulating phylloquinone is controversial, with studies reporting ApoEε4 carriers having lower (Kohlmeier et al., 1995; Saupe et al., 1993), higher (Yan et al., 2005), or similar (Beavan et al., 2005) phylloquinone concentrations as compared with non-carriers. In addition, although we controlled for overall quality of the diet (the C-HEI score), confounding by specific nutrients associated with cognition (e.g., vitamin E) should be ruled out in future studies. Limitations also include the fact that diet quality and alcohol consumption were assessed using a food frequency questionnaire administered at the NuAge Study baseline, whereas serum phylloquinone concentration was determined at the time of the cognitive evaluation. Finally, we acknowledge that our sample consisted of overselected individuals who had higher cognitive abilities than those not selected from the NuAge Study. Although this may limit generalization, the sample still included the oldest old (>80 years). To our knowledge, this is the first published study to examine the association between vitamin K and cognitive outcomes in healthy older adults. Notably, we found a positive association between serum phylloquinone and performances in verbal episodic memory, with a pattern suggesting a specific role in memory consolidation. In line with rodent models, our results are biologically plausible and add evidence for an expanding role of vitamin K from coagulation to cognition. The present report emphasizes the need to consider vitamin K as a nutritional factor in cognitive health in the aging population. Future studies will be needed to further our understanding of the role of vitamin K in brain and to examine whether treatment with vitamin K antagonists could affect cognition. Disclosure statement None of the authors has a conflict of interest. Acknowledgements We thank Doriane Kwong Hoi Fung for assistance with the serum phylloquinone analysis and Mélanie Drisdelle for coordinating the cognitive evaluation. The NuAge Study was supported by
the research grant MOP-62842 from the Canadian Institutes of Health Research (CIHR) and by the Quebec Network for Research on Aging funded by the Fonds de Recherche du Québec- Santé (FRQS; formerly Fonds de la Recherche en Santé du Québec). Access to NuAge databases and measurement of serum phylloquinone concentration were made possible through the NutCog study funded by the research grant MOP-82825 from the CIHR. Nancy Presse was supported by doctoral scholarships from the FRQS and from the CIHR. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.neurobiolaging. 2013.05.031. References Azharuddin, M.K., O’Reilly, D.S., Gray, A., Talwar, D., 2007. HPLC method for plasma vitamin K1: effect of plasma triglyceride and acute-phase response on circulating concentrations. Clin. Chem. 53, 1706e1713. Beavan, S.R., Prentice, A., Stirling, D.M., Dibba, B., Yan, L., Harrington, D.J., Shearer, M.J., 2005. Ethnic differences in osteocalcin gamma-carboxylation, plasma phylloquinone (vitamin K1) and apolipoprotein E genotype. Eur. J. Clin. Nutr. 59, 72e81. Belleville, S., Chatelois, J., Fontaine, F., Peretz, I., 2004. Mémoria: Batterie informatisée d’évaluation de la mémoire pour Mac et PC. Institut Universitaire de Gériatrie de Montréal, Montréal. Booth, S.L., Sokoll, L.J., O’Brien, M.E., Tucker, K., Dawson-Hughes, B., Sadowski, J.A., 1995. Assessment of dietary phylloquinone intake and vitamin K status in postmenopausal women. Eur. J. Clin. Nutr. 49, 832e841. Booth, S.L., Tucker, K.L., McKeown, N.M., Davidson, K.W., Dallal, G.E., Sadowski, J.A., 1997. Relationships between dietary intakes and fasting plasma concentrations of fat-soluble vitamins in humans. J. Nutr. 127, 587e592. Canada, H. 2007. Eating well with Canada’s Food Guide. Ó Her Majesty the Queen in Right of Canada, represented by the Minister of Health Canada, HC Publication no. 4651. Catalog no. H164e38/1-2007E. Canada, H.W., 1990. Nutrition recommendations: The report of the Scientific Review Committee. Minister of Supply and Services Canada. Catalog no. H-49e2/1990E. Carrie, I., Belanger, E., Portoukalian, J., Rochford, J., Ferland, G., 2011. Lifelong lowphylloquinone intake is associated with cognitive impairments in old rats. J. Nutr. 141, 1495e1501. Carrie, I., Portoukalian, J., Vicaretti, R., Rochford, J., Potvin, S., Ferland, G., 2004. Menaquinone-4 concentration is correlated with sphingolipid concentrations in rat brain. J. Nutr. 134, 167e172. Cocchetto, D.M., Miller, D.B., Miller, L.L., Bjornsson, T.D., 1985. Behavioral perturbations in the vitamin K-deficient rat. Physiol. Behav. 34, 727e734. Cohen, J.D., MacWhinney, B., Flatt, M., Provost, J., 1993. PsyScope: a new graphic interactive environment for designing psychology experiments. Behav. Res. Methods Instrum. Comput. 25, 257e271. Cook, Z., Kirk, S., Lawrenson, S., Sandford, S., 2005. Use of BMI in the assessment of undernutrition in older subjects: reflecting on practice. Proc. Nutr. Soc. 64, 313e317. Crivello, N.A., Casseus, S.L., Peterson, J.W., Smith, D.E., Booth, S.L., 2010. Age- and brain region-specific effects of dietary vitamin K on myelin sulfatides. J. Nutr. Biochem. 21, 1083e1088. Cutler, R.G., Kelly, J., Storie, K., Pedersen, W.A., Tammara, A., Hatanpaa, K., Troncoso, J.C., Mattson, M.P., 2004. Involvement of oxidative stress-induced abnormalities in ceramide and cholesterol metabolism in brain aging and Alzheimer’s disease. Proc. Natl. Acad. Sci. U.S.A 101, 2070e2075. Davidson, K.W., Sadowski, J.A., 1997. Determination of vitamin K compounds in plasma or serum by high-performance liquid chromatography using postcolumn chemical reduction and fluorimetric detection. Methods Enzymol. 282, 408e421. Davis, S., Laroche, S., 2006. Mitogen-activated protein kinase/extracellular regulated kinase signalling and memory stabilization: a review. Genes Brain Behav. 5 (Suppl 2), 61e72. Denisova, N.A., Booth, S.L., 2005. Vitamin K and sphingolipid metabolism: evidence to date. Nutr. Rev. 63, 111e121. Erkkila, A.T., Lichtenstein, A.H., Dolnikowski, G.G., Grusak, M.A., Jalbert, S.M., Aquino, K.A., Peterson, J.W., Booth, S.L., 2004. Plasma transport of vitamin K in men using deuterium-labeled collard greens. Metabolism 53, 215e221. Ferland, G., 2012. Vitamin K and the nervous system: an overview of its actions. Adv. Nutr. 3, 204e212. Gaudreau, P., Morais, J.A., Shatenstein, B., Gray-Donald, K., Khalil, A., Dionne, I., Ferland, G., Fulop, T., Jacques, D., Kergoat, M.J., Tessier, D., Wagner, R., Payette, H., 2007. Nutrition as a determinant of successful aging: description of the Quebec longitudinal study Nuage and results from cross-sectional pilot studies. Rejuvenation Res. 10, 377e386.
N. Presse et al. / Neurobiology of Aging 34 (2013) 2777e2783 Groupe de réflexion sur l’évaluation de la mémoire, Van der Linden, M., 2004. L’épreuve de rappel libre/rappel indicé à 16 items (RL/RI-16). L’évaluation des troubles de la mémoire: Présentation de quatre tests de mémoire épisodique (avec leur étalonnage), Solal, Marseille, pp. 25e47. Hall, J.G., Pauli, R.M., Wilson, K.M., 1980. Maternal and fetal sequelae of anticoagulation during pregnancy. Am. J. Med. 68, 122e140. He, X., Shen, L., Bjartell, A., Dahlback, B., 1995. The gene encoding vitamin Kdependent anticoagulant protein S is expressed in multiple rabbit organs as demonstrated by northern blotting, in situ hybridization, and immunohistochemistry. J. Histochem. Cytochem. 43, 85e96. Horwood, J.M., Dufour, F., Laroche, S., Davis, S., 2006. Signalling mechanisms mediated by the phosphoinositide 3-kinase/Akt cascade in synaptic plasticity and memory in the rat. Eur. J. Neurosci. 23, 3375e3384. Isaev, N.K., Stelmashook, E.V., Ruscher, K., Andreeva, N.A., Zorov, D.B., 2004. Menadione reduces rotenone-induced cell death in cerebellar granule neurons. Neuroreport 15, 2227e2231. Kohlmeier, M., Saupe, J., Drossel, H.J., Shearer, M.J., 1995. Variation of phylloquinone (vitamin K1) concentrations in hemodialysis patients. Thromb. Haemost. 74, 1252e1254. Kohlmeier, M., Saupe, J., Shearer, M.J., Schaefer, K., Asmus, G., 1997. Bone health of adult hemodialysis patients is related to vitamin K status. Kidney Int. 51, 1218e1221. Lamon-Fava, S., Sadowski, J.A., Davidson, K.W., O’Brien, M.E., McNamara, J.R., Schaefer, E.J., 1998. Plasma lipoproteins as carriers of phylloquinone (vitamin K1) in humans. Am. J. Clin. Nutr. 67, 1226e1231. Li, J., Lin, J.C., Wang, H., Peterson, J.W., Furie, B.C., Furie, B., Booth, S.L., Volpe, J.J., Rosenberg, P.A., 2003. Novel role of vitamin K in preventing oxidative injury to developing oligodendrocytes and neurons. J. Neurosci. 23, 5816e5826. Li, J., Wang, H., Rosenberg, P.A., 2009. Vitamin K prevents oxidative cell death by inhibiting activation of 12-lipoxygenase in developing oligodendrocytes. J. Neurosci. Res. 87, 1997e2005. McKeown, N.M., Jacques, P.F., Gundberg, C.M., Peterson, J.W., Tucker, K.L., Kiel, D.P., Wilson, P.W., Booth, S.L., 2002. Dietary and nondietary determinants of vitamin K biochemical measures in men and women. J. Nutr. 132, 1329e1334. Morley, J.E., Banks, W.A., 2010. Lipids and cognition. J. Alzheimers. Dis. 20, 737e747. NCEP ATP III, 2002. Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation. 106, 3143e3421. Neogi, T., Booth, S.L., Zhang, Y.Q., Jacques, P.F., Terkeltaub, R., Aliabadi, P., Felson, D.T., 2006. Low vitamin K status is associated with osteoarthritis in the hand and knee. Arthritis Rheum. 54, 1255e1261. Okano, T., Shimomura, Y., Yamane, M., Suhara, Y., Kamao, M., Sugiura, M., Nakagawa, K., 2008. Conversion of phylloquinone (vitamin K1) into menaquinone-4 (vitamin K2) in mice: two possible routes for menaquinone-4 accumulation in cerebra of mice. J. Biol. Chem. 283, 11270e11279. Presse, N., Gaudreau, P., Greenwood, C.E., Kergoat, M.J., Morais, J.A., Payette, H., Shatenstein, B., Ferland, G., 2012. A single measurement of serum phylloquinone is an adequate indicator of long-term phylloquinone exposure in healthy older adults. J. Nutr. 142, 1910e1916. Presse, N., Payette, H., Shatenstein, B., Greenwood, C.E., Kergoat, M.J., Ferland, G., 2011. A minimum of six days of diet recording is needed to assess usual vitamin K intake among older adults. J. Nutr. 141, 341e346. Presse, N., Shatenstein, B., Kergoat, M.J., Ferland, G., 2008. Low vitamin K intakes in community-dwelling elders at an early stage of Alzheimer’s disease. J. Am. Diet. Assoc. 108, 2095e2099. Prieto, A.L., O’Dell, S., Varnum, B., Lai, C., 2007. Localization and signaling of the receptor protein tyrosine kinase Tyro3 in cortical and hippocampal neurons. Neuroscience 150, 319e334. Prieto, A.L., Weber, J.L., Tracy, S., Heeb, M.J., Lai, C., 1999. Gas6, a ligand for the receptor protein-tyrosine kinase Tyro-3, is widely expressed in the central nervous system. Brain Res. 816, 646e661. Rey, A., 1959. Test de copie d’une figure complexe. Centre de psychologie appliquée, Manuel, Paris. Ronden, J.E., Drittij-Reijnders, M.J., Vermeer, C., Thijssen, H.H., 1998. Intestinal flora is not an intermediate in the phylloquinone-menaquinone-4 conversion in the rat. Biochim. Biophys. Acta 1379, 69e75.
2783
Sadowski, J.A., Hood, S.J., Dallal, G.E., Garry, P.J., 1989. Phylloquinone in plasma from elderly and young adults: factors influencing its concentration. Am. J. Clin. Nutr. 50, 100e108. Sakaue, M., Mori, N., Okazaki, M., Kadowaki, E., Kaneko, T., Hemmi, N., Sekiguchi, H., Maki, T., Ozawa, A., Hara, S., Arishima, K., Yamamoto, M., 2011. Vitamin K has the potential to protect neurons from methylmercury-induced cell death in vitro. J. Neurosci. Res. 89, 1052e1058. Sato, Y., Honda, Y., Hayashida, N., Iwamoto, J., Kanoko, T., Satoh, K., 2005. Vitamin K deficiency and osteopenia in elderly women with Alzheimer’s disease. Arch. Phys. Med. Rehabil. 86, 576e581. Saupe, J., Shearer, M.J., Kohlmeier, M., 1993. Phylloquinone transport and its influence on gamma-carboxyglutamate residues of osteocalcin in patients on maintenance hemodialysis. Am. J. Clin. Nutr. 58, 204e208. Shatenstein, B., Nadon, S., Godin, C., Ferland, G., 2005a. Development and validation of a food frequency questionnaire. Can. J. Diet. Pract. Res. 66, 67e75. Shatenstein, B., Nadon, S., Godin, C., Ferland, G., 2005b. Diet quality of Montrealarea adults needs improvement: estimates from a self-administered food frequency questionnaire furnishing a dietary indicator score. J. Am. Diet. Assoc. 105, 1251e1260. Shea, M.K., Benjamin, E.J., Dupuis, J., Massaro, J.M., Jacques, P.F., D’Agostino Sr., R.B., Ordovas, J.M., O’Donnell, C.J., Dawson-Hughes, B., Vasan, R.S., Booth, S.L., 2009. Genetic and non-genetic correlates of vitamins K and D. Eur. J. Clin. Nutr. 63, 458e464. Sokoll, L.J., Sadowski, J.A., 1996. Comparison of biochemical indexes for assessing vitamin K nutritional status in a healthy adult population. Am. J. Clin. Nutr. 63, 566e573. Spreen, O., Strauss, E., 1991. Profile of test results. A compendium of neuropsychological tests: Administration, norms, and commentary. Oxford University Press, New York, pp. 10e76. Spreen, O., Strauss, E., 1998. A compendium of neuropsychological tests: Administration, norms, and commentary, second ed. Oxford University Press, New York. Sundaram, K.S., Fan, J.H., Engelke, J.A., Foley, A.L., Suttie, J.W., Lev, M., 1996. Vitamin K status influences brain sulfatide metabolism in young mice and rats. J. Nutr. 126, 2746e2751. Sundaram, K.S., Lev, M., 1988. Warfarin administration reduces synthesis of sulfatides and other sphingolipids in mouse brain. J. Lipid Res. 29, 1475e1479. Suttie, J.W., 2009. Vitamin K in Health and Disease. CRC Press Taylor & Francis Group, New York. Thijssen, H.H., Drittij-Reijnders, M.J., 1994. Vitamin K distribution in rat tissues: dietary phylloquinone is a source of tissue menaquinone-4. Br. J. Nutr. 72, 415e425. Thijssen, H.H., Drittij-Reijnders, M.J., 1996. Vitamin K status in human tissues: tissue-specific accumulation of phylloquinone and menaquinone-4. Br. J. Nutr. 75, 121e127. Thijssen, H.H., Drittij-Reijnders, M.J., Fischer, M.A., 1996. Phylloquinone and menaquinone-4 distribution in rats: synthesis rather than uptake determines menaquinone-4 organ concentrations. J. Nutr. 126, 537e543. Tsaioun, K.I., Denisova, N.A., Obin, M., Joseph, J., 2000. Novel growth factor Gas6, phosphatidylserine and their age-related changes in the rat brain. Neurosci. Res. Commun. 26, 113e122. Wang, L.Y., Bates, C.J., Yan, L., Harrington, D.J., Shearer, M.J., Prentice, A., 2004. Determination of phylloquinone (vitamin K1) in plasma and serum by HPLC with fluorescence detection. Clin. Chim. Acta 347, 199e207. Washburn, R.A., McAuley, E., Katula, J., Mihalko, S.L., Boileau, R.A., 1999. The Physical Activity Scale for the Elderly (PASE): evidence for validity. J. Clin. Epidemiol. 52, 643e651. Wechsler, D., 1997. Weschsler Adult Intelligence Scale, Third (WAIS-III) ed. Psychological Corporation, San Antonio, TX. Wisdom, N.M., Callahan, J.L., Hawkins, K.A., 2011. The effects of apolipoprotein E on non-impaired cognitive functioning: a meta-analysis. Neurobiol. Aging 32, 63e74. Yan, L., Zhou, B., Nigdikar, S., Wang, X., Bennett, J., Prentice, A., 2005. Effect of apolipoprotein E genotype on vitamin K status in healthy older adults from China and the UK. Br. J. Nutr. 94, 956e961. Yesavage, J.A., Brink, T.L., Rose, F.L., Lum, O., Huang, V., Adey, M., Leirer, V.O., 1983. Development and validation of a geriatric depression screening scale: a preliminary report. J. Psychiatr. Res. 17, 37e49.