SEARCH

SEARCH BY CITATION

Keywords:

  • fracture;
  • osteocalcin;
  • carboxylated osteocalcin;
  • aged

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

We examined serum total osteocalcin (TOC), carboxylated osteocalcin (COC), and their ratio (COC/TOC) by one-step two-site immunofluorescent assays in 87% (n = 792) of all home-dwelling persons of 70 years or older living in a defined area in northern Finland. Other baseline subject-related risk factors of fractures were assessed by postal questionnaires, interviews, clinical examinations, and tests. During a 5-year follow-up period, all falls and fractures (n = 106) were recorded by regular phone calls and by examining all the medical records yearly. Serum TOC and COC concentrations increased with advancing age and were higher in women than in men, but corresponding differences were not found in the case of COC/TOC. The adjusted relative risk of fracture was elevated in association with low (≤−1 SD from the mean) COC; hazard ratio (HR, 95% CI) 2.00 (1.20-3.36) and low COC/TOC; HR 5.32 (3.26-8.68), the relative risk being highest in the population older than 80 years; and HR 7.02 (2.42-20.39). The predictive value of low COC/TOC lasted 3 years. The multivariable-adjusted relative risk of hip fracture (n = 26) in regard to low COC/TOC ratio was 3.49 (1.12-10.86), as compared with the persons who did not suffer hip fractures. Our results suggest that serum COC concentrations and, more strongly, COC/TOC, predict the occurrence of fractures in older community-dwelling adults. The risk of fracture associated with low COC/TOC equals the hip fracture risk previously verified for concomitant high serum undercarboxylated OC concentrations and low bone mineral density.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

OSTEOPOROTIC FRACTURES have serious consequences among the elderly. These fractures result in high mortality, frequent hospitalization, high health care costs, functional impairment, pain, and reduced quality of life.(1) Therefore, easily implemented valid methods to assess the risk of fracture are needed.

Low bone density predicts the occurrence of fracture(2,3) and hip fracture.(4) Serum osteocalcin (OC) is associated with bone turnover, which is a determinant of osteoporosis. The serum concentrations of OC increase with advancing age(5,6) in both men and women.(7) Results concerning the relationship between OC and the occurrence of fracture are contradictory, because some authors suggest that there is a positive association,(8) whereas others suggest that there is no association.(9,10)

The degree of the vitamin K-dependent γ-carboxylation of OC(11,12) becomes reduced with aging.(13,14) High proportions of undercarboxylated OC are associated with the occurrence of hip fractures in the institutionalized elderly.(15) This finding was supported by the results of a case-control study within a large prospective study among healthy elderly community-living women.(16)

Assay designs and antibody specificity for different forms of circulating OC in different studies are highly variable, complicating clinical interpretation.(17) We still lack population-based age- and sex-specific data on the distributions of various biochemical markers of bone metabolism and the relationships between these and the occurrence of fractures among older persons.

Using recently developed two-site immunoassays(18) based on well-characterized monoclonal antibodies (MAbs), we describe here the cross-sectional distributions of total OC (TOC) and carboxylated OC (COC) and their ratio (COC/TOC) in home-dwelling persons aged 70 years or older living in a defined geographical area in northern Finland. By recording all fall-related fractures during a follow-up period of 5 years, fracture prediction by means of these assays was studied, adjusting for other potential risk factors of fracture.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Study population

The population consisted of all home-dwelling persons aged 70 years or over who lived in five rural municipalities around the city of Oulu in northern Finland on September 1, 1991 (n = 954). Baseline examinations were carried out during September 1, 1991 to February 29, 1992 by two teams consisting of a physician, two nurses, and a physiotherapist. In addition, an ophthalmologist examined one-half of the population. Seven hundred and ninety-two (87% of those alive at the midpoint of the examinations) participated in these examinations and the recording of falls. The characteristics of the population are described in Table 1. Based on data from the medical records, the characteristics of those who did not participate (53 men and 64 women) did not differ from those of the participants in terms of age, (men 75.5 years [SD, 4.8] and women 76.9 years [SD, 5.3]), living alone (43%), and inability to go out (11%).

Table Table 1.. Characteristics of the Study Population (n = 792)
Thumbnail image of

Study protocol

The research teams collected the data by means of postal questionnaires, interviews, clinical examinations, and clinical tests.(19,20) Functional abilities were recorded by questions concerning the ability to go outdoors, to negotiate stairs, to walk 400 m, and to carry a 5-kg load 100 m without personal assistance. Lifetime history of smoking, current frequency of alcohol use, and the habit of doing heavy outdoor work, such as gardening (no, fewer than once a week, about once a week, and daily), were asked in the questionnaire. Depressive symptoms (short version of the Zung Self-Rating Scale),(21) and feelings of fear of falling (no, sometimes, frequently, always) also were recorded.

Weight (0.1 kg accuracy) in light clothing without shoes and height (cm) were measured by one of the nurses. Body mass index (kg/m2) was calculated out of these values. Cognitive status was assessed by the nurse using the Mini-Mental State Examination test(22) (variation of sum index 0-30). The occurrence of diabetes mellitus (on diet, tablets, or insulin) and the history of transient ischemic attacks and stroke were asked in the questionnaire. Other chronic diseases were recorded by the examining physician from the medical records of the health centers. The examining physician recorded current medication by interviewing the participants. Additional information was derived from the prescriptions and drug packages that the subjects had been asked to take with them to the examination. Choice reaction time was measured in the better upper extremity with a modified Digitest 1000 system (Digitest Oy, Muurame, Finland). Peak expiratory flow (PEF) was measured by a Mini Wright meter (Airmed, Harlow, U.K.). Knee extension strength was assessed by the physiotherapist as suggested by Pact et al.(23) Clinical muscle strength was graded as normal (sufficient for normal instrumental activities and gait) or abnormal. Walking speed (time in 10-m walk) was determined. Corrected distance visual acuity was measured by the ophthalmologist in half of the population(24) and by one of the nurses in the other half of the population under similar lighting conditions.

The nurses collected serum samples, which were stored after centrifugation at −72°C. Concentrations of TOC and COC were measured using one-step, two-site immunofluorescent assays, employing carefully selected MAbs described previously.(18) The MAb combination 2H9/6H9 was used for assay of TOC and the combination 6F9/3H8 for assay of COC. The assays detected both the intact and the large N-terminal midfragment of OC, thus measuring the more stable form of OC in circulation. The combination 6F9/3H8 had only 8% cross-reactivity with decarboxylated OC and was considered relatively specific for γ-carboxylated OC. Human OC purified from bone was used as a standard. The characteristics of the assays have been reported previously.(18) The analysis of serum samples was blind as regards the recording of falls.

Recording of falls lasted from the day of serum collection until December 31, 1996. The participants used diaries for fall recording. In addition, every participant was interviewed by telephone at 3-month intervals. Medical records concerning the entire study population were examined annually to check for falls causing any clinical fracture.(19,25)

Statistical analyses

The age- and sex-specific distributions of the two OC forms and their ratio were compared by analysis of variance (ANOVA). The relationship between serum OC values and the occurrence of fractures was studied by comparing means (±SD) in fracture cases and controls by t-test. OC values were dichotomized toward a greater risk at 1 SD from the mean of each 5-year age group (70-74 years, 75-79 years, 80-84 years, and 85 years or more) separately in men and women. Incidence rates of fractures were calculated by dividing the rate of first fractures by the person-time in stratified classes of OC values. The endpoint of person-time was the day of a move to another district, death, or the end of the follow-up period. Using Cox regression analysis, relative risks of fracture, expressed as hazard ratios (HRs) and supported by their 95% CIs, were determined. These calculations also were carried out separately in the age groups of 70-79 years and 80 years or older and in women. Linearity of the change in fracture risk in regard to increasing or decreasing TOC and COC values was assessed by stratifying the study population into quartiles according to these values.

In the second phase, all the baseline variables according to Cox regression analysis associated (p < 0.05) with the occurrence of fractures were entered in the final Cox regression analyses.

Using Cox regression, relative risk of fracture was determined separately during each of the follow-up years. Persons having suffered fractures during prior years were not excluded from the analyses.

Relative risks of hip fracture were determined, and a comparison was made to the population that did not suffer hip fractures.

Computation was carried out using commercially available software (BMDP Statistical Software, Inc., Los Angeles, CA, U.S.A.) on a SUN/UNIX mainframe computer (Sun Microsystems, Inc., Palo Alto, CA, U.S.A.).

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Age- and sex-specific distributions of different OC forms

The distributions of TOC, COC, and COC/TOC by age and sex are described in Table 2. Concentrations of TOC increased with advancing age in men (ANOVA, p = 0.009) and women (p < 0.001) and were higher in women than men (t-test, p < 0.001). Concentrations of COC also increased with advancing age in both men (ANOVA, p = 0.012) and women (p < 0.001) and were higher in women than men (t-test, p < 0.001). There was no difference in the COC/TOC values between the age groups, in either men (ANOVA, p = 0.587) or women (p = 0.250), but the COC/TOC values were higher in men than women (t-test, p < 0.001).

Table Table 2.. Distributions (Mean SD) of TOC and COC Concentrations (Units) and the COC/TOC Ratio in Older Persons by Sex and Age
Thumbnail image of

Occurrence of fractures relative to different OC forms

Hip fractures totaled 26, wrist fractures totaled 21, vertebral fractures totaled 5, and other fractures totaled 61.

In men, COC and COC/TOC values were lower in fracture cases than controls, but no difference was found with regard to TOC. Concentrations of TOC were higher, and COC/TOC values were lower in the female population with fractures than in their controls, and the corresponding COC values tended to be lower in the fracture cases (Table 3).

Table Table 3.. Distributions (Mean, SD) of TOC, COC (Units), and the COC/TOC Ratio in Fracture Cases and Controls Among Home-Dwelling Older Men and Women
Thumbnail image of

The incidence rates and relative risks of fracture according to higher (≥+1 SD from the mean) TOC, lower (≤−1 SD from the mean) COC, and lower (≤−1 SD from the mean) COC/TOC are described in Table 4. A higher TOC value was not associated with the occurrence of fracture, but lower values of COC and COC/TOC were positively associated.

Table Table 4.. Incidence Rate/1000 Person-Years and Unadjusted and Adjusted Risk Ratio (95% CI) of Fracture by TOC, COC, and COC/TOC in Older Persons (n = 792)
Thumbnail image of

Dichotomized at −1 SD from the mean of the total study population, the crude relative risk of fracture as regards a low COC/TOC ratio was 2.97 (1.91-4.63).

Compared with the lowest quartile, relative risks of fracture as regards higher quartiles of TOC concentrations were 1.58 (0.83-2.98), 1.83 (0.99-3.38), and 3.07 (1.71-5.52), respectively. Compared with the highest quartile, the corresponding figures as regards lower COC values were 0.81 (0.44-1.49), 1.18 (0.68-2.09), and 1.56 (0.91-2.67). For lower COC/TOC ratios, the relative risks were 1.36 (0.63-2.92), 2.40 (1.17-4.90), and 5.51 (2.88-10.5).

Other risk factors of fractures

The following baseline variables were associated (p < 0.05) with the occurrence of fracture during the 5-year follow-up period: higher (for the 5-year increase) age (HR, 1.35), female sex (2.44), doing heavy outdoor work (0.50), inability to go out unassisted (2.28), inability to carry a 5-kg load 100 m (2.28), poor distant visual acuity (2.65), reduced knee extension strength (2.64), frequent fear of falling (2.86), low (≤−1 SD from the mean) PEF (1.81), low cognitive status (Mini-Mental State Examination test result <23 points; 1.58), past stroke (4.35), use of anti-Parkinson medication (3.78), and use of psychotropic medication (2.08).

Results of multivariate analyses

After entering these variables in Cox regression analyses, low COC and COC/TOC predicted the occurrence of fracture, but TOC did not (Table 4). Female sex, frequent fear of falling, past stroke, and poor visual acuity were the other independent risk factors of fracture in each of these multivariate models. When analyzing the risk of fracture concerning COC/TOC ratio the following relative risks (95% CI) of fracture were found: female sex, 2.47 (1.46-4.20); frequent fear of falling, 2.14 (1.37-3.35); past stroke, 3.38 (1.27-8.99); and poor visual acuity, 2.16 (1.30-3.60).

Dichotomized at −1 SD from the mean of the total study population, the adjusted relative risk of fracture as regards a low COC/TOC ratio was 3.26 (2.01-5.25). Female sex, 2.00 (1.19-3.37); frequent fear of falling, 2.12 (1.36-3.31); past stroke, 3.33 (1.25-8.86); and poor visual acuity, 2.16 (1.30-3.60), were the other independent risk factors for fracture according to this multivariate model.

In the female population the multivariable-adjusted relative risks of fracture as regards high TOC, low COC, and low COC/TOC were 1.23 (0.60-2.51), 1.53 (0.82-2.88), and 4.64 (2.57-8.40), respectively.

In the younger population (70-79 years) the multivariable-adjusted relative risks of fracture as regards high TOC, low COC, and low COC/TOC were 1.34 (0.60-3.00), 2.04 (1.12-3.73), and 4.36 (2.46-7.75), respectively. The corresponding figures in the older (≥80 years) population were 1.44 (0.51-4.11), 1.51 (0.50-4.59), and 7.02 (2.42-20.39).

The risk of fracture in successive years

According to age- and sex-adjusted Cox regression analyses, relative risks of fracture during 1992, 1993, 1994, 1995, and 1996 associated with high TOC values (≤+1 SD from the mean) were 0.59 (0.14-2.51), 1.86 (0.70-4.93), 0.46 (0.06-3.48), 2.41 (0.80-7.25), and 0.79 (0.18-3.41), respectively. The corresponding figures as regards low COC values (≤−1 SD from the mean) were 3.45 (1.50-7.95), 2.37 (0.999-5.60), 1.54 (0.44-5.37), 0.77 (0.18-3.30), and 0.98 (0.29-3.30), respectively. For low COC/TOC ratios (≤−1 SD from the mean) the relative risks of fracture were 6.87 (3.12-15.1), 3.98 (1.74-9.13), 3.14 (1.02-9.73), 1.83 (0.53-6.29), and 0.78 (0.18-3.36).

The risk of hip fracture according to low COC/TOC ratio

The multivariable-adjusted relative risk of hip fracture (n = 26) in regard to low COC/TOC ratio was 3.49 (1.12-10.86), as compared with the persons who did not suffer hip fractures. The other variables used as covariates in this analysis were not statistically significantly (95% CI for HR crossing 1.00) associated with the occurrence of hip fracture.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

We present the first reliable age- and sex-specific reference values for different OC forms based on a representative sample of home-dwelling persons. In agreement with earlier findings, serum TOC concentrations increased with advancing age(6,10) and were higher in women than men.(7) Human vitamin K status has not been found to deteriorate clearly with age,(11,26) and we found vitamin K-dependent COC values to increase with age, similarly to the concentrations of TOC. The slight but nonsignificant decrease in the COC/TOC ratio with advancing age found in women but not in men was primarily because of a stronger age-associated increase of TOC concentrations in women.

Although serum TOC concentrations were higher in the women with fractures than in those without, we did not find an increased risk associated with high TOC, in agreement with the findings of Garnero et al.(10) Therefore, high bone turnover alone was not critical as regards the risk of fracture in the present study population.

The mechanisms concerning how low COC and COC/TOC increased fracture risk remain obscure. The linear increase in fracture risk seen with increasing TOC values may indicate contribution of high bone turnover.(5,6) The fact that low COC predicted the occurrence of fractures and the relationship between decreasing serum COC concentrations and the occurrence of fractures was J-shaped rather than linear presupposes additional explanations for the increased fracture risk. The high fracture risk associated with low serum COC concentrations and low COC/TOC ratios increased after adjusting for other risk factors of fracture, indicating that the contributions were independent of general risk factors of fractures. Serum undercarboxylated OC has been found to be associated more strongly with ultrasonic transmitted velocity at os calcis than bone mineral density,(27) suggesting that poor bone quality may explain the fracture risk associated with undercarboxylated OC, possibly as a result of inadequate vitamin K status.(27) To evaluate the possible contribution of vitamin K to the different OC forms and the occurrence of fractures in the present study, we examined the relationships between use of warfarin and the different OC forms and the occurrence of fractures. In line with the findings of Obrant et al.,(12) we found a strong positive association between the use of warfarin (n = 52) and low (≤−1 SD) COC (χ2, 88.9; df = 1; p = 0.001) and low COC/TOC (χ2, 63.3; df = 1; p = 0.001) but not high (≥1 SD) TOC (χ2, 0.167; df = 1; p = 0.682). However, use of warfarin was not associated with the occurrence of fractures (HR, 0.96 [0.42-2.18]) and does not support the results of a previous study, which found decreased levels of circulating vitamin K in elderly women with hip fractures.(28) In addition to impaired vitamin K status,(28) poor vitamin D status also may affect γ-carboxylation of OC,(29) and elevated proportions of undercarboxylated OC can be reversed by taking a supplement containing vitamin K(30) or by calcium and vitamin D administration.(15) Inadequate vitamin D status could be one explanation for the low serum COC concentrations that we found in older people living in northern Finland. In France, major determinants of low vitamin D in older women are personal outdoor score and to a lesser extent, the amount of daily sunlight.(31) Poor bone quality together with a simultaneous high bone turnover might explain why low COC/TOC strongly predicted the occurrence of fractures and hip fracture in the present study.

We found the predictive value of COC/TOC as regards fractures to last only 3 years, which calls for studies with repeated measurements of the COC/TOC ratio. The increased risk of fracture associated with low COC/TOC ratio found in our study resembles in magnitude the relative risk of hip fracture presented previously as regards a combination of low bone mineral density and OC with decreased carboxylation.(16) We found the increased risk of fracture with low COC/TOC ratio to be especially high among the oldest persons, which is in line with the finding that high undercarboxylated OC predicts hip fractures.(16) According to our results, simple immunoassays for total and carboxylated OC are likely to prove to be efficient tools in the short-term assessment of fracture risk among the elderly.

ACKNOWLEDGMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES

The expert technical assistance of Mrs. Pirjo Laaksonen is gratefully acknowledged. This study was supported by the Finnish Academy, Ministry of Social Welfare and Health, and the State Technology Development Center of Finland.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. REFERENCES
  • 1
    Ross PD 1996 Osteoporosis. Frequency, consequences, and risk factors. Arch Intern Med 156:13991411.
  • 2
    Hui SL, Slemenda CW, Johnston CC Jr 1989 Baseline measurement of bone mass predicts fracture in white women. Ann Intern Med 111:355361.
  • 3
    Melton LJ III, Atkinson EJ, O'Fallon WM, Wahner HW, Riggs BL 1993 Long-term fracture prediction by bone mineral assessed at different skeletal sites. J Bone Miner Res 8:12271233.
  • 4
    Cummings SR, Black DM, Nevitt MC, Browner W, Cauley J, Ensrud K, Genant HK, Palermo L, Scott J, Vogt TM 1993 Bone density at various sites for prediction of hip fractures. Lancet 341:7275.
  • 5
    Garnero P, Sornay-Rendu E, Chapuy M-C, Delmas PD 1996 Increased bone turnover in late postmenopausal women is a major determinant of osteoporosis. J Bone Miner Res 11:337349.
  • 6
    Melton LJ III, Khosla S, Atkinson EJ, O'Fallon WM, Riggs BL 1997 Relationship of bone turnover to bone density and fractures. J Bone Miner Res 12:10831091.
  • 7
    Epstein S, McClintock R, Bryce G, Poser J, Johnston CC Jr, Hui S 1984 Differences in serum bone Gla protein with age and sex. Lancet 1:307310.
  • 8
    Riis BJ, Hansen AM, Jensen AM, Overgaard K, Christiansen C 1996 Low bone mass and fast rate of bone loss at menopause: Equal risk factors for future fracture: A 15-year follow-up study. Bone 19:912.
  • 9
    Åkesson K, Ljunghall S, Jonsson B, Sernbo I, Johnell O, Gärdsell P, Obrant KJ 1995 Assessment of biochemical markers of bone metabolism in relation to the occurrence of fracture: A retrospective and prospective population-based study of women. J Bone Miner Res 10:18231829.
  • 10
    Garnero P, Hausherr E, Chapuy M-C, Marcelli C, Grandjean H, Muller C, Cormier C, Breart G, Meunier PJ Delmas PD 1996 Markers of bone resorption predict hip fracture in elderly women: The EPIDOS Prospective Study. J Bone Miner Res 11:15311538.
  • 11
    Plantalech L, Guillamont M, Vergnaud P, Leclercq M, Delmas PD 1991 Impairment of gamma carboxylation of circulating osteocalcin (bone Gla protein) in elderly women. J Bone Miner Res 6:12111216.
  • 12
    Obrant KJ, Käkönen S-M, Astermark J, Lilja H, Lövgren T, Åkesson K, Pettersson K 1999 The proportion of carboxylated to total or intact osteocalcin in serum discriminates warfarin treated patients from control subjects. J Bone Miner Res 14:555560.
  • 13
    Knapen MHJ, Hamulyak K, Vermeer C 1989 The effect of vitamin K supplementation on circulating osteocalcin (bone Gla protein) and urinary excretion. Ann Intern Med 111:10011005.
  • 14
    Merle B, Delmas PD 1990 Normal carboxylation of circulating osteocalcin (bone Gla protein) in Paget's disease of bone. Bone Miner 11:237245.
  • 15
    Szulc P, Chapuy M-C, Meunier PJ, Delmas PD 1993 Serum undercarboxylated osteocalcin is a marker of the risk of hip fracture in elderly women. J Clin Invest 91:17691774.
  • 16
    Vergnaud P, Garnero P, Meunier PJ, Breart G, Kamihagi K, Delmas PD 1997 Undercarboxylated osteocalcin measured with a specific immunoassay predict hip fracture in elderly women: The EPIDOS Study. J Clin Endocrinol Metab 82:719724.
  • 17
    Masters PW, Jones RG, Purves DA, Cooper EH, Cooney JM 1994 Commercial assays for serum osteocalcin give clinically discordant results. Clin Chem 40:358363.
  • 18
    Hellman J, Käkönen S-M, Matikainen M-T, Karp M, Lövgren T, Väänäen HK, Pettersson K 1996 Epitope mapping of nine monoclonal antibodies against osteocalcin: Combinations into two-site assays affect both assay specificity and sample stability. J Bone Miner Res 11:11651175.
  • 19
    Luukinen H, Koski K, Laippala P, Kivelä S-L 1997 Factors predicting fractures during falling impacts among home-dwelling older adults. J Am Geriatr Soc 45:13021309.
  • 20
    Koski K, Luukinen H, Laippala P, Kivelä S-L 1996 Physiological factors and medications as predictors of injurious falls by elderly: A prospective population-based study. Age Ageing 25:2938.
  • 21
    Tucker M, Ogle S, Davison J, Eilenberg MD 1987 Validation of a brief screening test for depression in the elderly. Age Ageing 16:139144.
  • 22
    Folstein MF, Folstein SE, McHugh PR 1975 Mini-mental state: A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12:189198.
  • 23
    Pact V, Sirotkin-Roses M, Beatus J 1984 Testing in the sitting position. In: PactV, Sirotkin-RosesM, BeatusJ (eds.) The Muscle Testing Handbook. Little, Brown and Company, Boston, MA, U.S.A./Toronto, Canada, p. 62.
  • 24
    Hirvelä H, Luukinen H, Laatikainen L 1995 Prevalence and risk factors of lens opacities in the elderly in Finland. A population-based study. Ophthalmology 102:108117.
  • 25
    Luukinen H, Koski K, Honkanen R, Kivelä S-L 1995 Incidence of injury-causing falls among older adults by place of residence: A population-based study. J Am Geriatr Soc 43:871876.
  • 26
    Hodges SJ, Pilkington MJ, Shearer MJ, Bitensky L, Chayen J 1990 Age-related changes in the circulating levels of congeners of vitamin K2, menaquinone-7, and menaquinone-8. Clin Sci 78:6366.
  • 27
    Liu G, Peakcock M 1998 Age-related changes in serum undercarboxylated osteocalcin and its relationships with bone density, bone quality, and hip fracture. Calcif Tissue Int 62:286289.
  • 28
    Hodges SJ, Åkesson K, Vergnaud P, Obrant K, Delmas PD 1993 Circulating levels of vitamins K1 and K2 decreased in elderly women with hip fracture. J Bone Miner Res 8:12411245.
  • 29
    Deyl Z, Adam M 1983 Evidence for vitamin D dependent γ-carboxylation in osteocalcin related proteins. Biochem Biophys Res Commun 113:294300.
  • 30
    Knapen MK, Hamulyak K, Vermeer C 1989 The effect of vitamin K supplementation on circulating osteocalcin (bone Gla protein) and urinary calcium excretion. Ann Intern Med 111:10011005.
  • 31
    Chapuy MC, Schott AM, Garnero P, Hans D, Delmas PD, Meunier PJ 1996 Healthy elderly French women living at home have secondary hyperparathyroidism and high bone turnover in winter. J Clin Endocrinol Metab 81:11291133.