ORIGINAL E n d o c r i n e

ARTICLE C a r e

Reference Intervals for Bone Turnover Markers and Their Association With Incident Hip Fractures in Older Men: The Health In Men Study S. A. Paul Chubb, Elizabeth Byrnes, Laurens Manning, John P. Beilby, Peter R. Ebeling, Samuel D. Vasikaran, Jonathan Golledge, Leon Flicker, and Bu B. Yeap School of Pathology and Laboratory Medicine (S.A.P.C., J.P.B., S.D.V.) and School of Medicine and Pharmacology (S.A.P.C., L.M., L.F., B.B.Y.), University of Western Australia, Crawley, Western Australia 6009, Australia; Biochemistry Department (S.A.P.C., E.B., J.P.B., S.D.V.), PathWest Laboratory Medicine, Nedlands, 6009 Western Australia, Australia; Department of Medicine (P.R.E.), Monash University, Clayton, Victoria 3146, Australia; Queensland Research Centre for Peripheral Vascular Disease (J.G.), School of Medicine and Dentistry, James Cook University, Townsville 4811, Queensland; Department of Vascular and Endovascular Surgery (J.G.), The Townsville Hospital, Townsville, Queensland 4810, Australia; and Department of Endocrinology and Diabetes (B.B.Y.), Fremantle Hospital, Fremantle 6160, Western Australia, Australia Context: Reference intervals for bone turnover markers (BTMs) and relationships between BTM and fracture risk in older men are not well characterized. Objective: The purpose of this article was to determine the reference intervals for serum total osteocalcin (tOC), undercarboxylated osteocalcin (ucOC), N-terminal propeptide of type I collagen (PINP), and collagen type I C-terminal cross-linked telopeptide (CTX-I) in healthy older men and to explore factors associated with BTMs, including hip fracture risk. Participants and Setting: We studied a population-based cohort of 4248 men aged 70 to 89 years, 4008 of whom had serum samples available for analysis. Interventions: Morning blood samples were collected at the study visit. Comorbid conditions were assessed by questionnaire. The reference sample comprised fasting men (n ⫽ 298, median age 75.3 years [interquartile range 73.9 –78.1 years) reporting excellent or very good health, without a history of diabetes, cardiovascular disease, cancer, depression, or dementia. Main Outcome Measures: Serum tOC, PINP, and CTX-I were estimated by automated electrochemiluminescence immunoassays, ucOC was estimated using hydroxyapatite binding, and incident hip fractures were captured from hospital admission data. Results: Reference intervals for tOC, ucOC, PINP, and CTX-I were 10.2 to 41.0, 5.2 to 21.9, 18 to 129 ␮g/L, and 117 to 740 ng/L, respectively. tOC, ucOC and CTX-I were associated with hip fracture incidence, but after adjustment for other risk factors only tOC remained significantly associated. Conclusions: Reference intervals for BTMs in older men have been defined. tOC may be more informative for hip fracture risk in older men than CTX-I and PINP. Further studies are needed to clarify the utility of BTM reference intervals in the management of aging men at risk of osteoporosis. (J Clin Endocrinol Metab 100: 90 –99, 2015)

ISSN Print 0021-972X ISSN Online 1945-7197 Printed in U.S.A. Copyright © 2015 by the Endocrine Society Received June 13, 2014. Accepted October 9, 2014. First Published Online October 16, 2014

90

jcem.endojournals.org

Abbreviations: BMD, bone mineral density; BTM, bone turnover marker; CI, confidence interval; CTX-I, collagen type I C-terminal cross-linked telopeptide; CV, coefficient of variation; eGFR, estimated glomerular filtration rate; HIMS, Health In Men Study; OC, osteocalcin; 25OHD, 25-hydroxyvitamin D; OR, odds ratio; PINP, N-terminal propeptide of type I collagen; tOC, total osteocalcin; ucOC, undercarboxylated osteocalcin; WADLS, Western Australian Data Linkage System; W1, wave 1; W2, wave 2.

J Clin Endocrinol Metab, January 2015, 100(1):90 –99

doi: 10.1210/jc.2014-2646

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 February 2015. at 04:49 For personal use only. No other uses without permission. . All rights reserved.

doi: 10.1210/jc.2014-2646

he risk of minimal trauma fracture is as important in aging men as it is for women. Men older than 50 years lose bone at 0.5 to 1.0% per year (1), contributing to an increase in the 10-year risk of hip or femur fracture from 0.2% to 2.9% between 50 and 80 years of age (2). About one third of hip fractures occur in men (1), and mortality after hip fracture is significantly higher in men than in women (3). Serum concentrations of bone turnover markers (BTMs) are widely used to monitor responses to antiosteoporosis treatment in clinical trials and in routine clinical practice. In addition, some studies have found that BTM measurements contribute independently to fracture risk prediction (4, 5). The International Osteoporosis Federation and International Federation of Clinical Chemistry Working Group nominated serum collagen type I Cterminal cross-linked telopeptide (CTX-I) and serum N-terminal propeptide of type I collagen (PINP) as reference BTMs for use in all studies (4). Reference intervals for these BTMs in premenopausal women have been extensively characterized for use as treatment targets in older women (6 –9). However, until recently, few high-quality reference data were available for men, particularly those in older age groups (10, 11). Thus, the ability of these BTMs to identify individuals with abnormally high bone turnover is limited in these patient groups. Another BTM, serum osteocalcin (OC), has also received increased interest in recent years and may have utility in fracture prediction. In particular, some authors have found higher concentrations of the undercarboxylated form of osteocalcin (ucOC) to be associated with fracture risk, particularly hip fracture (12–14). Only one previous study has examined this in elderly men (14). Whether this result can be generalized to elderly men in other populations has not been investigated. We measured the BTMs total osteocalcin (tOC), ucOC, PINP, and CTX-I in a large well-characterized cohort of community-dwelling older Australian men. We aimed to (1) define the reference intervals for BTMs in a subset of older men in optimal health, (2) identify the factors associated with the BTMs across the whole cohort, and (3) explore whether BTMs in this age group were associated with incident fracture during follow-up.

T

jcem.endojournals.org

91

2001–2004), 4248 of these men, then aged 70 to 89 years, completed a second questionnaire and attended for physical examination and venesection. Of these, 4008 men had specimens available for measurement of BTMs. The University of Western Australia Human Research Ethics Committee approved HIMS, and all men gave voluntary informed consent to participate (15).

Assessment of medical comorbidities Demographic, physical, medical, and medications data were collected by questionnaire at W2. Height, weight, and blood pressure were measured at this time. Body mass index was calculated as weight/(height)2, with weight measured in kilograms and height in meters; overweight was defined as body mass index of ⬎25 kg/m2. Men with diabetes and those reporting the use of glucose-lowering medication or with fasting or nonfasting glucose at W2 of ⱖ7.0 or ⱖ11.1 mmol/L, respectively, were considered to have diabetes. Morbidity was assessed using the Western Australian Data Linkage System (WADLS), which provides electronic linkage to records from death and hospital and cancer registries and captures admissions to all public and private hospitals in Western Australia (16). Prevalent cardiovascular disease was defined as a self-reported history of angina, acute myocardial infarction, stroke, or abdominal aortic aneurysm by questionnaire responses in W1 and W2 or hospital diagnoses of these conditions before W2. Cancer diagnoses were identified from the cancer registry between 1990 and W2. Dementia was assessed by medical history and hospital admissions data up to W2. Depression was classified according to mental health and hospital morbidity data up to W2 or by a score of ⱖ7 on the Geriatric Depression Scale administered at W2 (17). Any history of osteoporosis, fracture, or Paget disease or the use of bisphosphonates, oral glucocorticoids, or warfarin was obtained from questionnaire data at W2. The list of medications included all then available oral and parenteral bisphosphonates (alendronate, risedronate, and zoledronic acid) and the range of oral glucocorticoid preparations (cortisone, hydrocortisone, dexamethasone, and prednisolone). The occurrence and date of incident hip fractures were determined via WADLS from the time of assessment at W2 until the end of 2012.

Defining optimal health in older men

Materials and Methods

We studied a sample of men of the W2 cohort who were considered to be in optimal health. First, global perception of health was assessed at W2 with the question, “In general, would you say your health is . . . ?” The 5 responses available were “excellent,” “very good,” “good,” “fair,” and “poor.” This measure has been validated as a predictor of mortality in older men (18). The subset of older men reporting “excellent” or “very good” health without a history of smoking, diabetes, cardiovascular disease, cancer, depression, or dementia were considered to be in optimal health for defining reference intervals. Healthy men with nonfasting samples were excluded from the analysis of reference intervals.

Study population

Laboratory assays

The Health In Men Study (HIMS) is a population-based cohort study of community-dwelling older men in Perth, Western Australia. A total of 12 203 men of predominantly Caucasian origin completed a questionnaire and attended a screening medical examination in wave 1 (W1, 1996 –1999). In wave 2 (W2,

Blood samples were collected between 8:00 and 10:30 AM at W2. Serum was prepared immediately after phlebotomy and stored at ⫺80°C until assayed. The seasonality of the blood sample collection was defined as summer (December to February), autumn (March to May), winter (June to August), and spring

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 February 2015. at 04:49 For personal use only. No other uses without permission. . All rights reserved.

92

Chubb et al

BTM Reference Intervals in Older Men

J Clin Endocrinol Metab, January 2015, 100(1):90 –99

(September to November). Serum tOC, PINP, and CTX-I were measured by electrochemiluminescence immunoassay using a modular E170 analyzer and reagents supplied by Roche Diagnostics Australia. Between-day imprecision for these assays expressed as coefficient of variation (CV) was 3.7% and 2.9% at 18 and 89 ␮g/L tOC, 4.0% and 5.7% at 28 and 191 ␮g/L PINP, and 4.1% and 3.8% at 311 and 708 ␮g/L CTX-I. Plasma 25hydroxyvitamin D (25OHD) was determined by an automated chemiluminescence immunoassay on a LIAISON analyzer (DiaSorin) with interassay CVs of 13.2% and 11.3% at 38 and 130 nmol/L, respectively. Vitamin D deficiency was defined as a 25OHD concentration of ⬍50 nmol/L. Serum creatinine was measured with a Roche Hitachi 917 analyzer (Roche Diagnostics). The estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) algorithm (19); renal impairment was defined as eGFR of ⬍60 mL/min/1.73 m2. ucOC was estimated after incubation of serum samples with hydroxyapatite (5 mg/mL) (catalog number 391947; Calbiochem) for 1 hour with continuous mixing as described by Gundberg et al (20). The OC in the supernatant was estimated using the same assay as that for tOC and was reported as a concentration and as a fraction of the total. For an OC reference sample with expected fractional hydroxyapatite binding of 0.80 (kindly supplied by Professor Caren Gundberg, Yale School of Medicine, New Haven, Connecticut), the mean fractional hydroxyapatitebound OC was 0.77 and the between-run imprecision (CV) was 6.0%. Because OC has 3 potential carboxylation sites, and binding to hydroxyapatite is only relatively specific for carboxylated OC (20), the proportion of unbound OC is a surrogate for the overall ucOC status.

Statistical analysis All statistical procedures were performed using R (21). Statistical methodologies from published guidelines were applied to

define BTM reference intervals between the 2.5th to 97.5th percentiles (22). Outlying values were removed using the Dixon outlier statistic or the one-third rule. To explore the effect of age within the group of optimally healthy men, we performed quantile regression (R package: quantreg, Roger Koenker; http:// cran.r-project.org/web/packages/quantreg/index.html). Visual inspection of quantile regression plots was performed to determine whether any practical advantage would be gained by stratifying by age within this age-defined group of older men. Data are presented as medians and interquartile ranges. Between-group comparisons were by nonparametric methods. Multivariate analyses were performed using backward stepwise logistic regression. Candidate variables were included based on a value of P ⬍ .10 on bivariate analysis. The most parsimonious model was chosen using the minimum Akaike information criterion after the removal of each variable. For inclusion in the multivariate analyses, nonnormally distributed data were logtransformed to approximate a normal distribution. These data were then normalized to a distribution for which the mean value was 0 and the SD was 1.

Results Demographic and clinical characteristics of the entire cohort and the reference sample of older men in optimal health are shown in Table 1. A total of 298 men were included in the reference sample. None had a history of osteoporosis, previous hip fracture, or Paget disease, and none were taking bisphosphonates or warfarin. Three (1.0%) were taking or had a history of taking oral glucocorticoids. The distributions for each BTM in the reference sample are shown in Figure 1. With the exception of ucOC/tOC,

Table 1. Characteristics of 4008 Community-Dwelling Older Men With BTM Results and of the Subset of 298 Men in Optimal Health Forming the Reference Group

Age, y Body mass index, kg/m2 Creatinine, ␮mol/L eGFR, mL/min/1.73 m2 eGFR ⬍60 mL/min/1.73 m2 25OHD, nmol/L 25OHD ⬍50 nmol/L Osteoporosis Paget disease Bisphosphonate use Prior fracture Incident hip fracturea Glucocorticoid use Warfarin use tOC, ␮g/L ucOC, ␮g/L ucOC/tOC PINP, ␮g/L CTX-I, ng/L

Entire Cohort (n ⴝ 4008)

Healthy Subjects (n ⴝ 298)

76.3 (74.2–79.1) 26.28 (24.17–28.62) 89.0 (78.0 –101.0) 72.4 (61.5– 82.7) 887 (22.1) 67.3 (52.9 – 81.6) 848 (21.2) 157 (3.9) 24 (0.6) 71 (1.8) 193 (4.8) 114 (2.8) 76 (1.9) 235 (9.9) 18.5 (14.4 –24.0) 10.2 (8.1–13.1) 0.55 (0.49 – 0.62) 36.8 (27.8 – 49.7) 282 (199 –394)

75.3 (73.9 –78.1) 25.77 (23.93–28.08) 85.0 (76.0 –94.3) 76.5 (66.6 – 84.4) 35 (11.7) 68.6 (53.2– 82.5) 60 (20.1) 0 0 0 0 1 (0.3) 3 (1.0) 0 19.1 (15.2–24.1) 10.45 (8.1–12.9) 0.54 (0.49 – 0.61) 39.3 (28.1–50.1) 301 (227– 417)

Results are median (interquartile range) or number (%). a

Until death or December 31, 2012.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 February 2015. at 04:49 For personal use only. No other uses without permission. . All rights reserved.

doi: 10.1210/jc.2014-2646

jcem.endojournals.org

93

Figure 1. Distribution of BTM results in 298 healthy fasting older men. A, tOC. B, ucOC. C, ucOC/tOC. D, PINP. E, CTX-I.

which was normally distributed, each BTM was skewed to the right. Within the reference group, age was positively associated with tOC, ucOC, and CTX-I and negatively associated with ucOC/tOC (rs ⫽ 0.18 [P ⫽ .002], rs ⫽ 0.12 [P ⫽ .045], rs ⫽ 0.14 [P ⫽ .018], and rs ⫽ ⫺0.15 [P ⫽ .008], respectively). No BTM was associated with the 25OHD

concentration, eGFR, or body mass index in univariate analyses (data not shown). Therefore, further partitioning according to these factors was not required. The reference intervals for each of the BTMs are shown in Table 2. Because age was significantly associated with each BTM, the 2.5th and 97.5th quantile regressions were calculated (Figure 2). These indicated that age adjustment

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 February 2015. at 04:49 For personal use only. No other uses without permission. . All rights reserved.

94

Chubb et al

BTM Reference Intervals in Older Men

Table 2. Reference Intervals and Confidence Limits for Serum BTM in Healthy, Fasting Older Men 90% CI

tOC, ␮g/L ucOC, ␮g/L ucOC/tOC PINP, ␮g/L CTX-I, ng/L

2.5th–97.5th Percentiles

2.5th Percentile

97.5th Percentile

10.2– 41.0 5.2–21.9 0.36 – 0.73 18 –129 117–740

7.6 –10.8 4.1– 6.0 0.32– 0.37 16 –19 90 –130

38.0 – 45.0 18.4 –23.6 0.70 – 0.76 81–161 670 – 840

of the reference limits would not capture many additional men and was therefore unnecessary. Omitting the 3 men taking glucocorticoids had no impact on the reference intervals (data not shown). Factors associated with BTMs among the whole cohort The associations of various factors with BTMs across the entire cohort of 4008 men are shown in Table 3. Men receiving warfarin treatment had higher ucOC and ucOC/ tOC and lower tOC than men not receiving warfarin treatment. After adjustment for age and other factors, warfarin remained an independent predictor of ucOC, ucOC/tOC, and tOC. There was no significant association of warfarin treatment with either PINP or CTX-I in crude or age-adjusted analyses. Men taking bisphosphonates had significantly lower BTMs than those who were not (Table 3). For the bisphosphonate users, after exclusion of 4 men with Paget disease, the proportion with tOC, ucOC, PINP, and CTX-I below the lower reference limits were 34.3%, 13.4%, 55.2%, and 47.8%, respectively. Likewise, the proportions with tOC, ucOC, PINP, and CTX-I below the medians of the reference group were 83.6%, 89.6%, 77.6%, and 89.6%, respectively. When BTMs were examined by season, tOC, PINP, and CTX-I showed significant associations, with the highest values in spring and lowest in winter (Table 3). 25OHD concentrations also showed a significant association with season: median (interquartile range) 25OHD values in spring, summer, autumn, and winter were 60.7 (46.9 – 73.6), 69.5 (57.1– 85.6), 74.3 (60.4 – 88.6), and 61.7 (48.6 –76.2) nmol/L, respectively (Kruskal-Wallis test, P ⬍ .001). Nonfasting status and being overweight were associated with lower tOC, ucOC, and CTX-I, whereas renal impairment was associated with higher tOC, ucOC, PINP, and CTX-I. Oral glucocorticoid use was associated with lower ucOC, but not with other BTMs, and vitamin D status was not associated with any BTM.

J Clin Endocrinol Metab, January 2015, 100(1):90 –99

Incident hip fracture There were 114 incident hip fractures during follow-up among the entire sample of men with BTM data. Univariate and multivariate associations with incident hip fractures are shown in Table 4. Although age, osteoporosis, glucocorticoid use, tOC, ucOC, and CTX-I were associated with incident hip fracture in the univariate analyses, when all BTMs were included, the most parsimonious logistic regression model included age, glucocorticoid use, and log10(tOC) only. In the final model, age (odds ratio [OR], 1.12 per year; 95% confidence interval [CI95], 1.07–1.17), log10(tOC) (OR, 1.20 per SD; CI95, 1.00 – 1.42), and glucocorticoid use (OR, 2.62; CI95, 0.99 –5.80) were independent risk factors for incident hip fracture. After adjustment for age and glucocorticoid use, log10(PINP) and log10(CTX-I) individually were not significantly associated with incident hip fracture (P ⱖ .17).

Discussion In this large study of BTMs in older men, we have defined reference intervals for older men in optimal health and found that tOC predicted incident hip fracture independently of age and other covariates. Of men receiving bisphosphonate treatment, 77% to 90% had concentrations of CTX-I, PINP, and OC below the median of the reference sample. In addition, warfarin treatment was associated with increased ucOC and ucOC/tOC but not CTX-I and PINP, as would be expected from its mode of action to inhibit ␥ carboxylation. Our reference interval for CTX-I is similar to those for middle-aged and older men in other studies using the same analytical system (100 –700 ng/L [10] and 69 –760 ng/L [11]). Other assay systems for CTX-I may give significantly higher reference limits (7). For PINP, our reference interval has a higher upper limit than has been reported by other authors in predominantly younger cohorts of men (10, 11). This finding is consistent with the data of Jenkins et al (10), which suggested that higher PINP concentrations were more frequent in men aged ⬎80 years. One limitation of our result is the rather wide 90% CI around the upper limit, which is related to the skewed distribution of results. The 7 men with PINP values above this reference limit had no history of conditions likely to be associated with increased bone turnover, and renal function and vitamin D status were normal. Hence, these men were retained in the reference sample. Our upper reference limit for CTX-I in men is lower than a recently proposed provisional Australian reference interval for women aged ⬎70 years (100 –1000 ␮g/L), but our upper reference limit

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 February 2015. at 04:49 For personal use only. No other uses without permission. . All rights reserved.

doi: 10.1210/jc.2014-2646

jcem.endojournals.org

95

Figure 2. Age dependence of the BTMs in 298 healthy fasting older men. A, tOC. B, ucOC. C, ucOC/tOC. D, PINP. E, CTX-I. Quantile regressions for the 2.5th and 97.5th percentiles are shown. The shaded areas indicate the reference limits.

for PINP is similar to that proposed for men and women aged ⬎70 years (15–115 ␮g/L) (23). That publication highlighted the paucity of reference interval data for BTMs in people of this age group (23). We are not aware of published reference interval data for tOC measured by the Roche assay in predominantly Caucasian subjects. Hu et al (24) used this assay to deter-

mine reference intervals in younger Chinese adults (35– 45 years old), obtaining a range for men of 5.58 to 28.62 ␮g/L, somewhat lower than our result. Their ranges for CTX-I and PINP were also lower than ours, with upper limits of 612 ng/L and 65.49 ␮g/L, respectively. Because these limits are lower than those found in younger Australian (10) and Spanish (11) men, the possibility remains

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 February 2015. at 04:49 For personal use only. No other uses without permission. . All rights reserved.

96

Chubb et al

Table 3.

BTM Reference Intervals in Older Men

J Clin Endocrinol Metab, January 2015, 100(1):90 –99

Factors Associated With BTMs in 4008 Older Men OC, ␮g/L

Factor Bisphosphonates (n ⫽ 71) Yes No Fasting status (n ⫽ 3146) Yes No Glucocorticoid use (n ⫽ 76) Yes No Overweight (n ⫽ 2617) Yes No Paget disease (n ⫽ 24) Yes No Renal impairment (n ⫽ 860) Yes No Vitamin D deficiency (n ⫽ 848) Yes No Warfarin use (n ⫽ 235) Yes No Seasonality Spring Summer Autumn Winter

P

ucOC, ␮g/L

P

ucOC/tOC

P

P

CTX-I, ng/L

P

11.76 18.62

⬍.001

6.8 10.2

⬍.001

0.58 0.55

.03

17.4 40.0

⬍.001

137 284

⬍.001

18.68 17.98

.018

10.3 9.8

.013

0.55 0.55

.96

36.8 36.8

.61

299 217

⬍.001

17.70 18.54

.20

9.1 10.2

.007

0.53 0.55

.28

33.5 36.9

.28

294 282

.40

18.10 19.51

⬍.001

10.0 10.4

⬍.001

0.56 0.54

⬍.001

36.8 36.8

1.00

277 295

⬍.001

38.13 18.49

⬍.001

19.3 10.1

⬍.001

0.52 0.55

.06

160.9 36.7

⬍.001

709 281

⬍.001

21.18 18.00

⬍.001

11.1 10.0

⬍.001

0.52 0.56

⬍.001

40.7 36.1

⬍.001

324 272

⬍.001

18.32 18.59

.58

10.1 10.2

.16

0.55 0.55

.50

37.0 36.8

.82

284 282

.80

14.88 18.68

⬍.001

11.9 10.1

⬍.001

0.81 0.55

⬍.001

34.8 36.9

.09

268 273

.06

19.65 18.68 18.32 18.19

⬍.001

10.4 10.1 10.1 10.2

.08

0.53 0.55 0.56 0.56

⬍.001

38.3 37.8 36.3 36.2

⬍.001

303 286 280 271

⬍.001

that ethnicity-specific reference intervals may be needed for different populations. We found significant associations between tOC, ucOC, CTX-I, and incident hip fractures in univariate analyses. In multivariate analysis, tOC remained independently associated with incident hip fracture. Results for studies of the association between fracture risk and BTMs in women Table 4.

PINP, ␮g/L

have been inconsistent, especially after adjustment for confounders such as age and bone mineral density (BMD) (4). In men, fewer studies have been carried out, but, again, their conclusions are inconsistent. In the MrOS study (25), which included 72 men with and 933 men without hip fracture followed for up to 5 years, serum CTX-I and PINP were associated with fracture independently of age but not

Univariate and Multivariate Associations of BTM With Incident Hip Fracture in Older Men

Age, y Paget disease Osteoporosis Glucocorticoid use Warfarin use Bisphosphonate use Overweight Renal impairment Vitamin D deficiency Bone turnover markers log10(tOC), ␮g/L log10(ucOC), ␮g/L ucOC/tOC log10(PINP), ␮g/L log10(CTX-I), ng/L

Fracture (n ⴝ 114)

No Fracture (n ⴝ 3896)

Univariate P

Multivariate Model P

78.4 (75.1– 81.6) 0 (0) 9 (7.9) 7 (6.1) 7 (6.1) 3 (2.6) 69 (60.5) 28 (24.5) 22 (19.3)

76.3 (74.2–79.1) 24 (0.6) 148 (3.8) 70 (1.8) 228 (5.9) 68 (1.7) 2548 (65.4) 832 (21.4) 826 (21.2)

⬍.001 .40 .03 .007 .90 .47 .32 .41 .62

⬍.001

21.2 (16.1–27.0) 11.4 (8.9 –14.2) 0.54 (0.48 – 0.61) 40.9 (28.8 –53.1) 319 (226 – 454)

18.5 (14.3–23.9) 10.1 (8.0 –13.0) 0.55 (0.49 – 0.62) 36.7 (27.8 – 49.5) 281 (198 –393)

.002 .010 .18 .10 .02

.030 NR

.082 .050

NR NR

Abbreviation: NR, not retained in backwards stepwise regression. Results are median (interquartile range) or n (%).

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 February 2015. at 04:49 For personal use only. No other uses without permission. . All rights reserved.

doi: 10.1210/jc.2014-2646

of BMD. However, in other studies, these BTMs were not associated with any type of incident fracture after adjustments that included age and BMD (26 –28). OC was not associated with incident fracture in 3 studies performed in men (14, 27, 28). Small sample size and low fracture incidence may have limited the power of several of these studies, but a recent meta-analysis found modest but significant associations between PINP, CTX-I, and incident osteoporotic fractures (5). Our observations, therefore, contrast with these other studies and suggest that OC may be more informative in older men than CTX-I and PINP. Significant associations between carboxylation isoforms of OC and incident fracture risk have been reported in men and women (12–14). However, only 2 of these studies included adjustment for age and BMD, so the robustness of these observations is difficult to determine. In our study, ucOC but not ucOC/tOC was associated with hip fracture in unadjusted analysis, but this association was lost after adjustment. We found that most men who reported bisphosphonate use had BTMs below the reference group medians. The mean of the premenopausal female BTM reference interval has been suggested as a treatment target for effective antiresorptive therapy; for the methods used in the current study, these were 38.0 ␮g/L for PINP and 297 ng/L for CTX-I (7), close to the medians for our reference sample of men. Although we cannot verify the compliance of the men in our study, these results suggest that the majority are being treated effectively. Further studies are needed to establish whether the reference group medians would be appropriate treatment targets in older men. The prevalence of vitamin D deficiency, defined as 25OHD of ⬍50 nmol/L, in our sample of older men was 21.2%. This is lower than the value reported in a sample of all Australian older men (27.6%) (29) and in older men living in Sydney, Australia (43%) (30). A number of social and cultural factors may influence 25OHD concentrations, and differences in these may have contributed to the differences in observed prevalences. In addition, survivor bias may have affected this result. Nevertheless, we did not see any association between vitamin D status and BTM concentrations, consistent with other authors (7). Our data showed significant seasonal variation for tOC, PINP, and CTX-I, with the peak values in spring and the lowest values in winter. This has also been observed in several cross-sectional studies, with peak values in spring or summer and nadirs in autumn or winter (12, 31–33). In those publications, this variation was ascribed to seasonal variations in sunlight exposure and vitamin D status (12, 32). We also observed significant seasonal variation in 25OHD concentration, with the nadir 25OHD in winter and spring, ahead of the peak bone turnover in spring. The

jcem.endojournals.org

97

magnitude of the seasonal differences in BTM is small compared to the 90% CIs of the BTM reference limits and will not have a significant impact on the proposed reference intervals. Men taking warfarin had significantly lower ucOC and ucOC/tOC values, as expected from the mode of action of this drug (34). We also observed that these men had lower tOC. In one study, warfarin treatment had no effect on tOC concentrations (34). We are not aware of other highquality evidence to support or explain this observation, and the possibility that it has arisen by chance remains. The limitations of our study include the absence of BMD measurements in this cohort of men; hence, we cannot ascertain the prevalence of osteopenia and cannot adjust the associations of BTMs with fracture incidence for this factor. The men in this study were almost exclusively Caucasian, and, therefore, the reference intervals may not apply to men of other ethnicities. Because of the low prevalence of vitamin D deficiency, they may not apply to men living in areas with higher rates of vitamin D deficiency. We cannot comment on reference intervals in younger men nor in women. The samples in this study had been stored at ⫺80°C for 8 to 11 years. However, the design of the OC and PINP assays (35, 36) and the innate stability of PINP and CTX-I to freeze-thaw cycles (37, 38) suggest that major deterioration during storage would not be expected. Data on glucocorticoid use were collected from questionnaires completed by the men and did not include dose information; this limits our ability to explain the small effect of oral glucocorticoid use on BTMs. The incidence of hip fracture in the entire cohort of men was low, so our ability to define an association of fracture risk and BTM is limited; this may reflect survivor bias in the HIMS W2 cohort of men. Our reference intervals for ucOC and ucOC/tOC may not be generalizable to other laboratories because the binding of OC to hydroxyapatite is not fully specific for ucOC and hydroxyapatite from different suppliers and different batches may have variable affinity for OC (20). Nevertheless, our mean ucOC/tOC of 0.54 is comparable to the means of 0.37 to 0.40 given by other authors (20, 39). The strengths of our study include having enough men to define the reference limits with reasonable precision and the fact that they were well characterized with respect to their health status. The methods used for tOC, PINP, and CTX-I are automated and widely available. The ranges proposed here may therefore be transferable to other similar populations of older men in which BTMs are measured by these methods. This is the largest community-based cohort of older men in which all 4 BTMs have been analyzed and included 114 incident hip fractures, which provides a robust foundation for analysis compared with

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 February 2015. at 04:49 For personal use only. No other uses without permission. . All rights reserved.

98

Chubb et al

BTM Reference Intervals in Older Men

other, smaller studies. The follow-up of the cohort is essentially complete because all hospital admissions to public and private hospitals are captured by WADLS, and few men of this age migrate interstate (16). WADLS is robust for hospital admissions, with accuracy comparable to adjudication of medical records (40). In summary, we have examined factors associated with bone remodeling in older men and defined reference intervals for serum tOC, ucOC, ucOC/tOC, PINP, and CTX-I in a sample of older men in optimal health. Approximately 85% of men taking bisphosphonates had BTM values below the reference group median, and this may be a useful target for treatment. CTX-I, PINP, and OC were associated with incident fracture risk, but only OC remained associated after adjustments including age, suggesting that OC may be more informative in older men. Further studies are required to clarify the utility of reference intervals for these BTMs in the management of the increasing numbers of aging men at risk of osteoporosis.

Acknowledgments We thank Professor Caren Gundberg, Yale School of Medicine, for assistance with establishing the ucOC assay and George Koumantakis of Roche Diagnostics Australia, for the supply of reagent kits; the staff of PathWest Laboratory Medicine, Fremantle, Royal Perth, and Sir Charles Gairdner Hospitals, Western Australia, the Data Linkage Unit, Health Department of Western Australia, and ANZAC Research Institute, Sydney, New South Wales for excellent technical assistance; the staff and management of Shenton Park Hospital for their support of the study; and especially all of the men and staff who participated in the Western Australian Abdominal Aortic Aneurysm Programme and the Health In Men Study. Address all correspondence and requests for reprints to: Dr Paul Chubb, Biochemistry Department, PathWest Laboratory Medicine Western Australia, Royal Perth Hospital, GPO Box X2213, Perth, WA 6001, Australia. E-mail: [email protected]. This work was supported by the National Heart Foundation, Australia (Grant-in-Aid G11P 5662). The Health In Men Study was funded by Project Grants 279408, 379600, 403963, 513823, 634492, 1045710, and 1060557 from the National Health and Medical Research Council of Australia. J.G. is supported by funding from the National Health and Medical Research Council of Australia and the Queensland Government. The funding sources had no involvement in the planning, analysis, or writing of the manuscript. Authors’ roles were as follows: S.A.P.C., L.M., L.F., and B.B.Y. designed the study; S.A.P.C., J.P.B., L.F., E.B., and B.B.Y. conducted the study; S.A.P.C., E.B., J.P.B., and L.F. collected data; S.A.P.C., L.M., and B.B.Y. analyzed data; S.A.P.C., L.M., P.R.E., S.D.V., J.G., and B.B.Y. interpreted data; S.A.P.C. drafted the manuscript; S.A.P.C., E.B., L.M., J.P.B., P.R.E., S.D.V., J.G., L.F., and B.B.Y. revised the manuscript for critical intellectual content; and S.A.P.C., E.B., L.M., J.P.B., P.R.E.,

J Clin Endocrinol Metab, January 2015, 100(1):90 –99

S.D.V., J.G., L.F., and B.B.Y. approved the final version of the manuscript. Disclosure Summary: The authors have nothing to disclose.

References 1. Drake MT, Khosla S. Male osteoporosis. Endocrinol Metab Clin North Am. 2012;41:629 – 641. 2. van Staa TP, Dennison EM, Leufkens HG, Cooper C. Epidemiology of fractures in England and Wales. Bone. 2001;29:517–522. 3. Center JR, Nguyen TV, Schneider D, Sambrook PN, Eisman JA. Mortality after all major types of osteoporotic fracture in men and women: an observational study. Lancet. 1999;353:878 – 882. 4. Vasikaran S, Eastell R, Bruyère O, et al. Markers of bone turnover for the prediction of fracture risk and monitoring of osteoporosis treatment: a need for international reference standards. Osteoporos Int. 2011;22:391– 420. 5. Johansson H, Odén A, Kanis JA, et al. A meta-analysis of reference markers of bone turnover for prediction of fracture. Calcif Tissue Int. 2014;94:560 –567. 6. Adami S, Bianchi G, Brandi ML, et al. Determinants of bone turnover markers in healthy premenopausal women. Calcif Tissue Int. 2008;82:341–347. 7. Eastell R, Garnero P, Audebert C, Cahall DL. Reference intervals of bone turnover markers in healthy premenopausal women: results from a cross-sectional European study. Bone. 2012;50:1141–1147. 8. Glover SJ, Gall M, Schoenborn-Kellenberger O, et al. Establishing a reference interval for bone turnover markers in 637 healthy, young, premenopausal women from the United Kingdom, France, Belgium, and the United States. J Bone Miner Res. 2009;24:389 –397. 9. Glover SJ, Garnero P, Naylor K, Rogers A, Eastell R. Establishing a reference range for bone turnover markers in young, healthy women. Bone. 2008;42:623– 630. 10. Jenkins N, Black M, Paul E, Pasco JA, Kotowicz MA, Schneider HG. Age-related reference intervals for bone turnover markers from an Australian reference population. Bone. 2013;55:271–276. 11. Olmos JM, Hernández JL, Martínez J, Pariente E, Llorca J, González-Macías J. Bone turnover markers in Spanish adult men. The Camargo Cohort Study. Clin Chim Acta. 2010;411:1511– 1515. 12. Szulc P, Chapuy MC, Meunier PJ, Delmas PD. Serum undercarboxylated osteocalcin is a marker of the risk of hip fracture in elderly women. J Clin Invest. 1993;91:1769 –1774. 13. Vergnaud P, Garnero P, Meunier PJ, Bréart G, Kamihagi K, Delmas PD. Undercarboxylated osteocalcin measured with a specific immunoassay predicts hip fracture in elderly women: the EPIDOS Study. J Clin Endocrinol Metab. 1997;82:719 –724. 14. Luukinen H, Käkönen SM, Pettersson K, et al. Strong prediction of fractures among older adults by the ratio of carboxylated to total serum osteocalcin. J Bone Miner Res. 2000;15:2473–2478. 15. Norman PE, Flicker L, Almeida OP, Hankey GJ, Hyde Z, Jamrozik K. Cohort profile: the Health In Men Study (HIMS). Int J Epidemiol. 2009;38:48 –52. 16. Holman CD, Bass AJ, Rosman DL, et al. A decade of data linkage in Western Australia: strategic design, applications and benefits of the WA data linkage system. Aust Health Rev. 2008;32:766 –777. 17. Almeida OP, Flicker L, Norman P, et al. Association of cardiovascular risk factors and disease with depression in later life. Am J Geriatr Psychiatry. 2007;15:506 –513. 18. Alfonso H, Beer C, Yeap BB, Hankey GJ, Flicker L, Almeida OP. Perception of worsening health predicts mortality in older men: the Health In Men Study (HIMS). Arch Gerontol Geriatr. 2012;55: 363–368. 19. Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150:604 – 612. 20. Gundberg CM, Nieman SD, Abrams S, Rosen H. Vitamin K status

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 February 2015. at 04:49 For personal use only. No other uses without permission. . All rights reserved.

doi: 10.1210/jc.2014-2646

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

and bone health: an analysis of methods for determination of undercarboxylated osteocalcin. J Clin Endocrinol Metab. 1998;83: 3258 –3266. R Core Development Team. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2011. Solberg HE. The theory of reference values. Part 5. Statistical treatment of collected reference values. Determination of reference limits. J Clin Chem Clin Biochem. 1983;21:749 –760. Vasikaran SD, Chubb SA, Schneider HG. Towards optimising the provision of laboratory services for bone turnover markers. Pathology. 2014;46:267–273. Hu WW, Zhang Z, He JW, et al. Establishing reference intervals for bone turnover markers in the healthy Shanghai population and the relationship with bone mineral density in postmenopausal women. Int J Endocrinol. 2013;2013:513925. Bauer DC, Garnero P, Harrison SL, et al. Biochemical markers of bone turnover, hip bone loss, and fracture in older men: the MrOS study. J Bone Miner Res. 2009;24:2032–2038. Meier C, Nguyen TV, Center JR, Seibel MJ, Eisman JA. Bone resorption and osteoporotic fractures in elderly men: the Dubbo Osteoporosis Epidemiology Study. J Bone Miner Res. 2005;20:579 – 587. Szulc P, Montella A, Delmas PD. High bone turnover is associated with accelerated bone loss but not with increased fracture risk in men aged 50 and over: the prospective MINOS study. Ann Rheum Dis. 2008;67:1249 –1255. Yoshimura N, Muraki S, Oka H, Kawaguchi H, Nakamura K, Akune T. Biochemical markers of bone turnover as predictors of osteoporosis and osteoporotic fractures in men and women: 10-year follow-up of the Taiji cohort. Mod Rheumatol. 2011;21:608 – 620. Daly RM, Gagnon C, Lu ZX, et al. Prevalence of vitamin D deficiency and its determinants in Australian adults aged 25 years and older: a national, population-based study. Clin Endocrinol (Oxf). 2012;77:26 –35. Hirani V, Cumming RG, Blyth FM, et al. Vitamin D status among older community dwelling men living in a sunny country and associations with lifestyle factors: the Concord Health and Ageing in

jcem.endojournals.org

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

99

Men Project, Sydney, Australia. J Nutr Health Aging. 2013;17:587– 593. Midtby M, Magnus JH, Joakimsen RM. The Tromsø Study: a population-based study on the variation in bone formation markers with age, gender, anthropometry and season in both men and women. Osteoporos Int. 2001;12:835– 843. Pasco JA, Henry MJ, Kotowicz MA, et al. Seasonal periodicity of serum vitamin D and parathyroid hormone, bone resorption, and fractures: the Geelong Osteoporosis Study. J Bone Miner Res. 2004; 19:752–758. Woitge HW, Scheidt-Nave C, Kissling C, et al. Seasonal variation of biochemical indexes of bone turnover: results of a population-based study. J Clin Endocrinol Metab. 1998;83:68 –75. Menon RK, Gill DS, Thomas M, Kernoff PB, Dandona P. Impaired carboxylation of osteocalcin in warfarin-treated patients. J Clin Endocrinol Metab. 1987;64:59 – 61. Blumsohn A, Hannon RA, Eastell R. Apparent instability of osteocalcin in serum as measured with different commercially available immunoassays. Clin Chem. 1995;41:318 –319. Brandt J, Krogh TN, Jensen CH, Frederiksen JK, Teisner B. Thermal instability of the trimeric structure of the N-terminal propeptide of human procollagen type I in relation to assay technology. Clin Chem. 1999;45:47–53. Garnero P, Vergnaud P, Hoyle N. Evaluation of a fully automated serum assay for total N-terminal propeptide of type I collagen in postmenopausal osteoporosis. Clin Chem. 2008;54:188 –196. Okabe R, Nakatsuka K, Inaba M, et al. Clinical evaluation of the Elecsys ␤-CrossLaps serum assay, a new assay for degradation products of type I collagen C-telopeptides. Clin Chem. 2001;47:1410 – 1414. Booth SL, Dallal G, Shea MK, Gundberg C, Peterson JW, DawsonHughes B. Effect of vitamin K supplementation on bone loss in elderly men and women. J Clin Endocrinol Metab. 2008;93:1217– 1223. Barr EL, Tonkin AM, Welborn TA, Shaw JE. Validity of self-reported cardiovascular disease events in comparison to medical record adjudication and a statewide hospital morbidity database: the AusDiab study. Intern Med J. 2009;39:49 –53.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 February 2015. at 04:49 For personal use only. No other uses without permission. . All rights reserved.

Reference intervals for bone turnover markers and their association with incident hip fractures in older men: the Health in Men study.

Reference intervals for bone turnover markers (BTMs) and relationships between BTM and fracture risk in older men are not well characterized...
651KB Sizes 1 Downloads 6 Views