Sage Journals: Discover world-class research

Abstract

Background

The use of bone turnover markers in clinical practice and research in younger people is limited by the lack of normative data and understanding of common causes of variation in bone turnover marker values in this demographic. To appropriately interpret bone turnover markers, robust reference intervals specific to age, development and sex are necessary. This study aimed to determine reference intervals of bone turnover markers in females aged 16–25 years participating in the Safe-D study.

Methods

Participants were recruited through social networking site Facebook and were asked to complete an extensive, online questionnaire and attend a site visit. Participants were tested for serum carboxy-terminal cross-linking telopeptide of type 1 collagen and total procollagen type 1 N-propeptide using the Roche Elecsys automated analyser. Reference intervals were determined using the 2.5th to 97.5th percentiles of normalized bone turnover marker values.

Results

Of 406 participants, 149 were excluded due to medical conditions or medication use (except hormonal contraception) which may affect bone metabolism. In the remaining 257 participants, the reference interval was 230–1000 ng/L for serum carboxy-terminal cross-linking telopeptide of type 1 collagen and 27–131 µg/L for procollagen type 1 N-propeptide. Both marker concentrations were inversely correlated with age and oral contraceptive pill use. Therefore, intervals specific to these variables were calculated.

Conclusions

We defined robust reference intervals for cross-linking telopeptide of type 1 collagen and procollagen type 1 N-propeptide in young females grouped by age and contraceptive pill use. We examined bone turnover markers’ relationship with several lifestyle, clinical and demographic factors. Our normative intervals should aid interpretation of bone turnover markers in young females particularly in those aged 16 to 19 years where reference intervals are currently provisional.

Keywords

Bone turnover markers reference intervals CTX P1NP young females Safe-D study

Introduction

Suboptimal peak bone mass substantially increases the risk of osteoporosis (and related fracture) in later life. Thus, strategies aimed at optimizing peak bone mass in younger, premenopausal females are crucial in its prevention.¹ Genetics heavily determine adult peak bone mass, but 20–40% of adult peak bone mass is believed to be influenced by lifestyle choices.² The period of late adolescence to early adulthood has recently been highlighted as an understudied transitional life stage in regards to bone acquisition.² To appropriately interpret results of studies investigating the mechanisms and determinants of peak bone mass development in the period of late adolescence to early adulthood, quality normative data are required spanning these life stages.

In younger females at the end of puberty, high bone turnover is present due to bone modelling, but subsides as bone growth halts and newly formed osteoid undergoes primary and secondary mineralization.³ High bone turnover in this age group is characterized by a tight coupling of bone resorption and formation, allowing for bone accrual and increase in bone strength. By the third decade of life, the majority of bone mass has been attained, and bone turnover reaches a plateau in healthy, premenopausal females.^4,5

Areal bone mineral density (aBMD), determined by dual energy X-ray absorptiometry (DXA), is currently the gold standard clinical measure of bone health and fracture risk. Biochemical markers of bone turnover may provide information that is complementary to aBMD.⁶ Bone turnover markers (BTMs) are able to indicate changes in bone metabolism more rapidly than DXA, which may not be able to detect a significant change in aBMD for 12–24 months in many clinical settings. Other advantages of BTM measurement are the lack of exposure to ionizing radiation and their relatively low cost. They have been previously shown to be a reliable measure of bone turnover in young females.⁷

The measurement of BTMs in the blood or urine reflects whole body rates of bone formation and resorption. Carboxy-terminal cross-linking telopeptide of type 1 collagen (CTX) is a marker of bone resorption and is released when collagen degrades. Conversely, procollagen type 1 N-propeptide (P1NP) is a marker of bone formation and is cleaved off when collagen is laid down. These two analytes have been identified as the reference standard analytes by the International Osteoporosis Foundation and International Federation of Clinical Chemistry working groups to allow focused efforts on development of harmonized age- and sex-specific reference intervals.⁸

BTMs have been shown to be predictive of fracture risk and have application in evaluating responses to osteoporosis therapy, making them clinically useful in fracture prevention strategies.^9,10 Although the overall prevalence of skeletal frailty is low in young females, groups at risk include patients with eating disorders, coeliac disease, congenital bone diseases such as osteogenesis imperfecta or haematological malignancies requiring bone marrow transplantation. These females may benefit from BTM monitoring to assess disease progression and treatment efficacy. In a research setting, the evaluation of BTMs may be useful in studies implementing lifestyle or nutritional interventions aimed at optimizing peak bone mass and/or improving bone mineral density (BMD) in adolescents and young adults. However, the clinical utility of BTM measurement in young females is severely limited by the lack of age-specific normative data in a well-described reference population.

A recent review by Vasikaran et al.¹¹ highlighted that CTX reference intervals for females under the age of 20 years and P1NP reference intervals for females under the age of 25 years were areas of uncertainty in the field. Reference intervals have been established for healthy premenopausal females in a number of studies, but are limited in females in their third and fourth decade of life.^12–14 Provisional reference intervals exist for the age range 16–18 years, which are calculated from a single study with 30 subjects only.¹⁵

The main objectives of this cross-sectional study were to: (1) define age-specific reference intervals for CTX and P1NP in young females aged 16–25 years; (2) investigate the relationship of BTMs with several common clinical and lifestyle factors.

Methods

The study was approved by Melbourne Health Human Research and Ethics Committee, Melbourne Health, Victoria, Australia (project number 2013.007). The study was carried out in accordance with the National Statement on Ethical Conduct in Research Involving Humans (2007) produced by the National Health and Medical Research Council of Australia.

Participants

Data were obtained from participants recruited via Facebook advertising for part A of the Safe-D study, a project investigating vitamin D and health in females aged 16 to 25 years at the time of recruitment and living in Victoria, Australia.^16,17 Participants were requested to complete a comprehensive online questionnaire which covered five key health domains: (1) health behaviours including diet; (2) musculoskeletal health; (3) mental health; (4) sexual and reproductive health; (5) cardiovascular and metabolic health. They also attended a study site visit including a physical assessment and blood sampling, between April 2014 and November 2015. Pregnancy and breastfeeding were exclusion criteria for the study.

Lifestyle factors were assessed by self-report and included: smoking status, daily calcium intake,¹⁸ hormonal contraceptive use, alcohol consumption and physical activity. The time spent walking and in moderate or vigorous physical activity was recorded to give a physical activity score.¹⁹ Physical activity was then categorized into four categories based on metabolic equivalent values or MET.minutes per week: minimal (<40), low (400 to < 600), moderate (600 to < 1200) or high (1200+). Menarche, menstrual cycle length and the date of the participants’ last period were self-reported. Site visit date and date of last period were collected to estimate menstrual cycle timing.

Sampling procedures

Before attending a study site visit at the Royal Melbourne Hospital, participants were instructed to fast overnight for a minimum of 8 h. To minimize the effects of diurnal variation, fasting blood samples were collected between 8:00 and 11:00. Serum aliquots were stored at −80℃, if not tested in real-time.

BTMs

Serum CTX, total P1NP and oestradiol were measured using the Roche Cobas e411 automated analyser (Roche Diagnostics GmbH, Mannheim, Germany).²⁰ Serum CTX had a coefficient of variation (CV) of 5.5% at 342 ng/L and 4.2% at 787 ng/L. Serum P1NP had a CV of 5.2% at 40.2 µg/L and 4.5% at 222.6 µg/L. Oestradiol had a CV of 18.3% at 46.8 pmol/L, 2.9% at 376.8 pmol/L and 2.9% at 1188.2 pmol/L.

Other pathology testing

Serum 25-hydroxyvitamin D (25OHD; 25OHD₃ plus 25OHD₂) was measured using liquid chromatography tandem mass spectrometry (LC-MS/MS) at VivoPharm Laboratories (Melbourne, Australia).²¹ This method used a hexa-deuterated 25OHD₃ as an internal standard and was performed on an API 4000 QTRAP LC-MS/MS system. Concentrations of each metabolite were extrapolated by the use of an eight-point standard curve prepared using Chromsystems Vitamin D metabolite serum calibration standards (Chromsystems GmbH, Munich, Germany). Tri-Level vitamin D metabolite Quality Control samples from UTAK Laboratories (PM Separations, US) were used as quality controls samples in each assay run. The CV for 25OHD₃ at 24.74 nmol/L was 2.0%, 1.6% at 72.72 nmol/L and 1.4% at 163.33 nmol/L. The CV for 25OHD₂ was 4.9% at 21.35 nmol/L, 2.5% at 63.48 nmol/L and 2.5% at 152.47 nmol/L. Serum biochemistry was measured using Abbott ARCHITECT c16000 integrated system (Abbott Diagnostics, Abbott Park, IL). Glomerular filtration rate (eGFR) was estimated using the chronic kidney disease epidemiology collaboration formula.²² Parathyroid hormone (PTH) and thyroid-stimulating hormone (TSH) were measured using an Abbott ARCHITECT i2000 analyser. The CVs for PTH and TSH were 4.9% and 2.9%, respectively.

BMD

aBMD was measured using DXA (QDR 4500A densitometer, Hologic Inc., Bedford, USA) at the lumbar spine (L1–L4), total hip and femoral neck. The results were expressed as grams per centimetre squared (g/cm²). Values were compared with reference data to give a Z-score (SD from mean values of age-matched reference group).²³ Scans were analysed with QDR software version 9.1D. Quality control scans were performed on a daily basis according to manufacturer protocol.

Statistical analysis

Statistical analysis was performed using StataSE 13 (StataCorp, TX, USA). CTX values were exported as ng/mL to two decimals places and were converted into ng/L. To determine BTM reference intervals in healthy young females, participants were excluded from analysis if they had not completed the medical history section of the questionnaire, not fasted before blood collection, if other pathology results were abnormal (eGFR < 60 ml/min/1.73 m², corrected calcium > 2.60 mmol/L or <2.10 mmol/L, PTH > 10 pmol/L, 25OHD < 25 nmol/L, TSH < 0.35 mIU/L or C-reactive protein > 10 mg/L), were previously diagnosed with medical conditions which affect bone health (e.g. osteoporosis, type 1 or 2 diabetes mellitus, hyperthyroidism, hyperparathyroidism, cystic fibrosis, coeliac disease, Crohn’s disease, pancreatitis) or if participants were taking medications which affect bone health (prednisolone, antiepileptic drugs, lithium carbonate, methotrexate, antiandrogens, antipsychotics, adalimumab and proton pump inhibitors). Participants using depot medroxyprogesterone acetate, a progesterone only pill or an etonogestrel implant, were excluded; combined oral contraceptive (COC) use was not an exclusion. Participants with a BMD Z-score < −2.0 at any skeletal site were excluded.

Both graphical (histogram and QQ-plot) and analytical (Shapiro-Wilk test) methods were used to test serum CTX and P1NP for normality. Both were found to be positively skewed, and a square root transformation was performed on CTX and P1NP before reference intervals were calculated. Values that exceeded the mean by ±3SD (outliers) were excluded. Reference intervals were determined by using 2.5th to 97.5th percentiles of each BTM concentration. Age at the time of site visit was rounded to the nearest year to allow determination of age-specific reference intervals. Participants aged 25 years at recruitment who had a birthday between consent and their site visit were recoded as 25 years old for statistical purposes (n = 4). Participants were divided into age groups as follows: 16–19, 20–22 and 23–25 years.

Descriptive statistics were used to describe both discrete and continuous variables. Univariable analysis was performed to examine the effects of age, COC use, body mass index (BMI) and season on BTMs. Wilcoxon signed-rank test was used to test whether a relationship existed between a dichotomous variable and an ordinal variable (e.g. contraception use and CTX). Spearman’s correlation was used to assess correlations between continuous variables. Multivariable regression analysis was used to identify determinants of CTX and P1NP. A P-value of less than 0.05 was considered statistically significant.

Results

Blood samples were collected from 406 participants. Figure 1 shows the distribution of CTX and P1NP before transformation.

Figure 1.

Distribution of CTX (left) and P1NP (right) before square-root transformation.

To determine reference intervals, a total of 149 participants were excluded for the following reasons (note: some participants may meet multiple exclusion criteria): BTM results exceeded the mean by three SDs (n = 11), participant reported not fasting overnight (n = 17), participant had not completed the medical history section of the questionnaire (n = 4), corrected calcium > 2.60 mmol/L (n = 2), PTH > 10 pmol/L (n = 25), 25OHD < 25 nmol/L (n = 13), TSH < 0.35 mIU/L (n = 4), C-reactive protein > 10 mg/L (n = 31), participant was taking specific medications (prednisolone [n = 2], antiepileptic drugs [n = 8], lithium carbonate [n = 1], methotrexate [n = 2], antiandrogens [n = 2], antipsychotics [n = 7], adalimumab [n = 1], proton pump inhibitors [n = 7], progesterone only pill [n = 2] the etonogestrel implant [n = 15]) or were previously diagnosed with one of the following conditions: type 1 diabetes mellitus (n = 2), hyperthyroidism (n = 2), cystic fibrosis (n = 1) and coeliac disease (n = 8). One participant was excluded as she was diagnosed with osteoporosis before commencing the study, while 45 were excluded with a Z-score < −2.0 at one or more skeletal sites.

The median (Q1, Q3) CTX value was 540 ng/L (410, 700). The median (Q1, Q3) P1NP value was 62 µg/L (47, 83). Serum CTX and P1NP were strongly correlated with each other (rho = 0.686, P < 0.001). Serum BTM concentrations were not associated with blood collection time (CTX: P = 0.501, P1NP: P = 0.222, data not shown) or storage time (CTX: P = 0.332; P1NP: P = 0.272, data not shown).

Participant characteristics are shown in Table 1. Seventy-four per cent of participants had a European ethnic background and 12% were born outside Australia. Mean (SD) age of study participants at site visit was 21.9 (2.8) years. Serum CTX and P1NP were inversely correlated with both chronological age (CTX: rho = −0.46, P < 0.001; P1NP rho = −0.39, P < 0.001; Figure 2) and years since menarche (CTX: rho = −0.42, P < 0.001; P1NP rho = −0.38, P < 0.001; Figure 3).

Table 1.

Characteristics of Safe-D study participants with BTM results available and the healthy cohort used to calculate reference intervals.

Characteristic	Overall cohort n = 406 Mean (SD) (%[n/ntotal])	Included population n = 257 Mean (SD) (%[n/ntotal])
Age (years)	22.1 (2.8)	21.9 (2.8)
Weight (kg)	66.9 (15.6)	65.8 (11.2)
Height (m)	1.66 (0.06)	1.67 (0.06)
Body mass index (kg/m²)	24.1 (5.2)	23.6 (3.5)
Daily dietary calcium consumption (mg/day)	806 (313)	814 (325)
Physical activity (MET.minutes per week)^a
Minimal (<40)	4.8 (19/399)	5.1 (13/256)
Low (40–599)	25.5 (102/399)	21.5 (55/256)
Moderate (600–1199)	23.3 (93/399)	23.4 (60/256)
High (1200+)	46.4 (185/399)	50.0 (128/256)
Using hormonal contraception
Non-user	52.7 (212/402)	51.4 (132/257)
Oral contraceptive pill	42.0 (169/402)	47.1 (121/257)
Etonogestrel implant	3.7 (15/402)	–
Ethinyl estradiol and etonogestrel vaginal ring	0.3 (1/402)	0.4 (1/257)
Levonorgestrel intrauterine system	0.8 (3/402)	1.1 (3/257)
Progesterone only pill	0.5 (2/402)	–
Current smoker^b	9.5 (38/401)	7.0 (18/256)
Binge drank in the last 12 months^c	23.6 (91/386)	24.7 (61/247)
Calcium supplement use	3.5 (14/401)	2.3 (6/256)
Vitamin D supplement use	8.7 (35/401)	7.8 (20/256)
Multivitamin use	16.7 (67/401)	17.6 (45/256)

Note: numbers may not add up to 406 and 257, respectively, due to missing data.

Statistically significant different between overall cohort and included population (P = 0.032).

Participants who smoked daily or were social smokers were defined as current smokers.

Binge drinking was defined as consuming more than 4 standard drinks on a particular occasion.

Figure 2.

Scatterplot of CTX (left) and P1NP (right) by age.

Figure 3.

Scatterplot of CTX (left) and P1NP (right) by years since menarche.

Serum CTX concentrations were 17% lower in participants who used the COC pill compared with non-users (488 ng/L vs. 591 ng/L, Z = 4.26, P < 0.001) and P1NP concentrations were 22% lower (56 µg/L vs. 72 µg/L, respectively, Z = 4.71, P < 0.001). In participants not using any form of hormonal contraception (n = 132), serum CTX and P1NP concentrations were not associated with menstrual cycle timing (CTX: rho = −0.06, P = 0.481; P1NP: rho = 0.12, P = 0.190).

BTM values were significantly different between age groups (P < 0.001 for both markers). BTM values were significantly different between the 16–19 and 20–22 age group (P < 0.001 for both comparisons) as well as between the 20–22 and 23–25 age group (CTX: P < 0.001; P1NP: P = 0.002). COC use was significantly different between groups (P = 0.005). Approximately 36% used COC among 16–19-year-old participants, 43% in 20–22 year olds and 55% in 23–25 year olds. Age- and COC-stratified reference intervals are reported in Table 2.

Table 2.

Reference intervals for CTX and P1NP group by age and COC use.

Bone marker	COC use	Age	Sample size	Median (IQR)	Reference interval^a
CTX (ng/L)	Non-user	16–19	39	720 (560–810)	400–1220
		20–22	45	580 (510–720)	282–1166
		23–25	48	495 (360–630)	169–1018
		All ages	132	585 (460–740)	253–1064
	User	16–19	29	620 (510–740)	330–1140
		20–22	34	515 (410–660)	240–920
		23–25	62	410 (320–529)	211–772
		All ages	125	480 (370–620)	230–979
	Non-users	16–19	68	695 (550–770)	359–1162
	+ users	20–22	79	550 (470–700)	270–980
		23–25	110	435 (340–560)	200–850
		All ages	257	540 (410–700)	230–1000
P1NP (µg/L)	Non-user	16–19	39	84.2 (62.2–105.0)	33.9–143.0
		20–22	45	68.4 (57.8–91.7)	38.8–131.2
		23–25	48	60.4 (46.4–73.2)	23.7–112.9
		All ages	132	67.4 (57.0–91.8)	33.2–135.4
	User	16–19	29	75.2 (52.3–98.7)	19.1–109.0
		20–22	34	59.7 (42.0–68.6)	10.2–120.0
		23–25	62	48.7 (33.0–58.9)	19.2–115.9
		All ages	125	54.2 (39.9–75.2)	20.1–116.8
	Non-users	16–19	68	78.1 (61.1–102.5)	28.7–141.6
	+ users	20–22	79	64.5 (53.1–83.7)	28.0–127.0
		23–25	110	55.0 (40.2–68.6)	25.3–110.9
		All ages	257	61.7 (46.8–83.3)	27.2–131.1

COC: combined oral contraceptive; CTX: C-terminal telopeptide of type 1 collagen; P1NP: total procollagen type 1 N-propeptide.

Reference intervals were calculated based on normally distributed transformed data and converted back to original values.

Serum CTX was slightly lower in winter compared with other seasons (505 ng/L in winter compared with 568 ng/L in other seasons; Z = 2.67, P = 0.008). P1NP showed no seasonal variation (Z = 0.83, P = 0.404). Neither CTX nor P1NP concentrations were associated with 25OHD or PTH (P > 0.100 for all variables). BTM concentrations were not associated with BMI (CTX: rho = −0.11, P = 0.076; P1NP: rho = −0.096, P = 0.126).

Age, COC use and season were all significant determinants of serum CTX in a multivariate model (age: β = −0.639, P < 0.001; COC use: β = −1.959, P < 0.001; winter: β = −1.102, P = 0.022; R²= 0.27). Age and COC use were independently associated with P1NP in a multivariate model (age: β = −0.209, P < 0.001; COC user: β = −0.842, P < 0.001; R²= 0.21).

Discussion

To our knowledge, this is the first study to report reference intervals for CTX and P1NP in young Australian females spanning from late teens to mid-20s. Reference intervals currently implemented for this age range are sourced from a handful of studies which do not span across adolescence into young adulthood and generally have smaller sample sizes for calculated age-specific reference intervals in young females.^15,24 We used a novel method of recruitment in the form of a Facebook advertising campaign to recruit a broadly-representative sample of young females.¹⁷ The use of this recruitment method was a key to the success of the study since young females tend not to respond to traditional approaches such as recruitment through health-care centres or mail outs. We derived reference intervals specific to age group and COC use, which we demonstrated were both prominent determinants of BTM values.

The International Osteoporosis Foundation and International Federation of Clinical Chemistry have recommended that CTX and P1NP are used as reference analytes for bone turnover in clinical trials, as markers of bone resorption and formation, respectively. A position paper published by their bone marker standards working group listed several reference intervals for CTX and P1NP from various studies. In premenopausal females, these reference intervals were 110–630,¹³ 100–620,²⁵ 90–660¹⁴ and 70–610 ng/L²⁶ for serum CTX and 16.3–78.2,¹³ 16.2–60.9²⁵ and 14.6–63.5 µg/L²⁶ for serum P1NP.⁸

Currently, the Australian Consensus Reference Intervals for premenopausal females aged 20 to 24 years are 150–800 ng/L for serum CTX and 15–90 µg/L for P1NP based on work by Jenkins et al. using the central 90% of the BTM distribution in 20–29 year-old females (see Table 3).^11,24 Jenkins et al. did not screen for bone or metabolic disease or for medications which may affect bone turnover. In addition, samples in the female cohort were up to 15 years old when assayed which may have impacted negatively on sample stability. We measured BTMs, CTX and P1NP using the Roche Elecsys automated analyser. Using the central 95th centile, we derived reference intervals for these markers in a healthy cohort (CTX: 230–1000 ng/mL; P1NP: 27–131 µg/L). These intervals were higher than the central 95th centile-derived intervals from the Jenkins study (CTX: 100–900 ng/L; P1NP: 20–100 µg/L). The difference likely lies in our younger cohort with higher bone turnover at the end of puberty as well as sample storage time. Previously published P1NP reference intervals for females under the age of 25 years were considered provisional.²⁷ The reference intervals calculated in this study may be more relevant to females in their early 20s than previously published ranges studying slightly older females.²⁴

Table 3.

CTX and P1NP reference intervals comparison in females aged up to age 30 years using automated assays.

Study	Region	n	Age range (years)	CTX (ng/L)	P1NP (µg/L)
Present study	Australia	257	16–25	230–1000	27–131
Jenkins et al.²⁴	Australia	166	20–29	150–800	25–90
Huang et al.¹⁵	Canada	30	16–18	0–1590	49–277

CTX: C-terminal telopeptide of type 1 collagen; P1NP: total procollagen type 1 N-propeptide.

In females aged 16–18 years, Huang et al. found a reference interval of 0–1590 ng/L for CTX and 49–277 µg/L for P1NP using the same assay platform. Current provisional ranges for CTX and P1NP are based on this study alone.²⁷ These reference intervals are higher than our calculated intervals for 16–19 year olds and were derived with a similar sample size (n = 30) evaluating adolescents recruited from day-care centres, schools, religious institutions or were the children of hospital workers in Hamilton, Ontario, Canada. The differences between the two studies may be due to a higher prevalence of COC use in our cohort (Huang et al. did not report on COC use); however, their reference intervals are also higher than our intervals for 16–19 years old without COC use (400–1220 ng/L). To resolve such differences between studies, more research may be required into participants under the age of 20 years, with larger sample sizes and controlling for a number of variables such as fasting, time of day, physical activity, season, COC use and recent fracture.

During skeletal growth in young females, bone mass is accrued up until the third decade of life.^4,5 For example, Teegarden et al. showed 99% of BMD is attained by the age of 22 years with peak bone mineral content attained by age 26 years. In our study, CTX and P1NP showed variation with chronological age and years since menarche as expected. Previous studies have also consistently found associations between BTMs and age.^{15,24–26,28} In a Chinese population aged 15 to 110 years, Li et al.²⁹ also demonstrated relatively higher concentrations of CTX and P1NP in females between the age of 15 and 19 years, which thereafter declined with age. Nomura et al.²⁸ found significant decreases in serum CTX with each age group (mean CTX concentration for 10–19: 817 ng/L; early 20s: 546 ng/L; late 20s: 385 ng/L) and a significant decrease between the groups in their 20s and 30s for P1NP.²⁸ Both BTMs reached nadirs in participants aged in their 30s and early 40s. In a regression model, Huang et al.¹⁵ also found age had a significant effect on BTMs in 16–18-year-old females (P < 0.001). Walsh et al.³⁰ found serum CTX and P1NP were lower in young females aged 30–32 years compared with 16–18 year olds (CTX: 310 vs. 670 ng/L; P1NP: 38.7 vs. 102.8 ng/ml; P < 0.01 for both variables).

BTMs were found to be lower in COC users compared with non-users, which is consistent with previous literature.^13,14,31,32 Mean CTX was 17% lower in users than in non-users, while mean P1NP concentrations were 22% lower (P < 0.001 for both markers). In premenopausal females aged 20–50 years, Adami et al.²⁶ found P1NP concentrations 26% lower in COC users compared with non-users (P < 0.001). De Papp et al.¹⁴ and Garnero et al.³² found median concentrations of CTX and P1NP were lower in COC users compared with non-users. Glover et al.¹³ found that P1NP concentrations were 12.1% lower in females taking the COC compared with non-users.¹³ In 16–18 year olds, Walsh et al.³⁰ found COC users had lower serum CTX concentrations compared with non-users (570 vs. 710 ng/L; P < 0.05) which is in agreement with our data. Therefore, it may be more appropriate for reference intervals to be specific to contraception status, as well as age and sex, and COC use should be considered when interpreting BTM results. Further research is needed on the effects of other forms of hormonal contraception on BTM values.

Previous literature on the seasonality of BTMs has shown mixed results.^33–35 We found CTX values were 11% lower in winter compared with other seasons, but found no seasonal variation in P1NP. Jenkins et al. also found no seasonal variation in P1NP, but found increased concentrations of CTX in winter, which they postulated may in part be due by subclinical vitamin D deficiency and secondary hyperparathyroidism.²⁴ Longitudinal studies with larger sample sizes than previously studied are necessary to further explore the effect of season in younger premenopausal females.^33,36

Previous literature has suggested a relationship between bone health and adiposity with increasing adiposity potentially increasing the risk of fracture.^37,38 We found no association between BTMs and BMI. Adami et al.²⁶ found CTX and P1NP were negatively correlated with BMI (r = −0.177, r = −0.353, respectively, P < 0.001). In 25–45-year-old females, Evans et al. found CTX concentrations were significantly lower in obese adults (P = 0.024).³⁹ Other studies have found no association between BMI and BTMs.^12,13,15

The main strengths of this study were the relatively large sample size spanning the important and understudied age range of 16–25 years, and the collection of an extensive range of health, lifestyle and clinical information both through online questionnaires and the study site visit.¹⁶ This information enabled us to calculate reference intervals in a healthy cohort of young females by excluding participants with abnormal pathology results or medical conditions and/or medications that may adversely affect bone turnover. In this study, blood was collected between the hours of 8:00 and 11:00 to reduce biological variation.^40,41 Most BTMs increase at night, reaching a peak between 2:00 and 8:00, after which they decrease rapidly, reaching a nadir between 13:00 and 23:00.^36,42 We found no relationship between collection time and BTM values in our study. Another strength of the study is that samples were taken in fasted participants as CTX and P1NP have previously been shown to be lower in subjects in a fed state.⁴⁰

This study has several limitations. Lifestyle and clinical data such as medical history were by self-report, so may be affected by recall bias. Menstrual cycle timing was not taken into account when booking site visits, although the change in BTMs over the menstrual cycle is believed to be small and we found no association with menstrual cycle timing in this study.³⁶ Participants were not asked to refrain from exercising 24 h prior to their site visit, which may increase BTMs.^43,44

This study demonstrates the desirability of further research into BTM reference intervals in other specific age ranges and across both sexes to enable development of additional robust intervals that are age- and sex specific. Consideration should be taken to control for circadian rhythm, fasting status and acute exercise. Other factors that should be considered when assessing BTMs include fracture and medical history, renal function, vitamin D status, smoking, alcohol use, sample storage and assay method.

In conclusion, reference intervals of CTX and P1NP were determined in a well-characterized cohort of healthy young females across the age range 16 to 25 years taking into consideration COC use. Values were higher than the Australian Consensus Reference Intervals for premenopausal females. Importantly, we studied a single cohort whose ages were distributed across this age range, when skeletal maturation is completed in females. These findings have important application in bone health research and should prove useful in the clinical assessment of bone metabolism in young females.

Footnotes

Acknowledgements

The authors thank the participants who took part in Safe-D study. The authors also thank the Safe-D chief investigators Associate Professor Marie Pirotta, Professor Anthony Jorm, Associate Professor Shanton Chang and Professor George Varigos along with associate investigator Professor Kim Bennell and study coordinator Ms Adele Rivers. The authors thank Young Female Health Initiative (YFHI) associate investigators Dr Yasmin Jayasinghe, Dr Catherine Segan and Dr Asvini Subasinghe. The authors thank the following people for assisting with various study components: Adrian Bickerstaffe, Maria Bisignano, Lauren Gilbert, Dr Ashwini Kale, Stefanie Koneski, Skye Maclean, Anna Scobie, Marjan Tabesh, Oktay Tacar, Miaowen Zhou.

Declaration of conflicting interests

The Safe-D study has received in-kind support from Swisse Wellness who provided vitamin D supplements for part B of the Safe-D study. Swisse Wellness did not play a role in study design nor the implementation or interpretation of these studies.

Funding

The Safe-D study was funded by National Health and Medical Research Council (NHMRC) project grant APP1049065.

Ethical approval

The study was approved by Melbourne Health Human Research and Ethics Committee, Melbourne Health, Victoria, Australia (project number 2013.007).

Guarantor

JDW.

Contributorship

JDW conceived the study. ETC participated in the design of the study data collection and drafted the manuscript. JDW, SMG and AG are study investigators and were involved in study design, study co-ordination and helped draft the manuscript. CC helped to interpret study data and drafting the manuscript. AG is the study statistician and was involved in sample size and power calculations and also advised on the statistical analysis of the data. All authors read, contributed to and approved the final manuscript.

References

Hansen

Overgaard

Riis

. Role of peak bone mass and bone loss in postmenopausal osteoporosis: 12 year study. BMJ 1991; 303: 961–961.

Weaver

Gordon

Janz

. The National Osteoporosis Foundation’s position statement on peak bone mass development and lifestyle factors: a systematic review and implementation recommendations. Osteoporosis Int 2016; 27: 1–106.

Eastell

. Role of oestrogen in the regulation of bone turnover at the menarche. J Endocrinol 2005; 185: 223–234.

Teegarden

Proulx

Martin

. Peak bone mass in young. J Bone Miner Res 1995; 10: 711–715.

Hansen

. Assessment of age and risk factors on bone density and bone turnover in healthy premenopausal women. Osteoporosis Int 1994; 4: 123–128.

Meier

Nguyen

Center

. Bone resorption and osteoporotic fractures in elderly men: the dubbo osteoporosis epidemiology study. J Bone Miner Res 2005; 20: 579–587.

Weaver

Peacock

Martin

. Quantification of biochemical markers of bone turnover by kinetic measures of bone formation and resorption in young healthy females. J Bone Miner Res 1997; 12: 1714–1720.

Vasikaran

Eastell

Bruyère

. Markers of bone turnover for the prediction of fracture risk and monitoring of osteoporosis treatment: a need for international reference standards. Osteoporosis Int 2011; 22: 391–420.

Garnero

Sornay-Rendu

Claustrat

. Biochemical markers of bone turnover, endogenous hormones and the risk of fractures in postmenopausal women: the OFELY study. J Bone Miner Res 2000; 15: 1526–1536.

10.

Chapurlat

Garnero

Breart

. Serum type I collagen breakdown product (serum CTX) predicts hip fracture risk in elderly women: The EPIDOS study. Bone 2000; 27: 283–286.

11.

Vasikaran

Chubb

Ebeling

. Harmonised Australian reference intervals for serum PINP and CTX in adults. Clin Biochem Rev 2014; 35: 237–237.

12.

Eastell

Garnero

Audebert

. Reference intervals of bone turnover markers in healthy premenopausal women: results from a cross-sectional European study. Bone 2012; 50: 1141–1147.

13.

Glover

Gall

Schoenborn-Kellenberger

. 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.

14.

de Papp

Bone

Caulfield

. A cross-sectional study of bone turnover markers in healthy premenopausal women. Bone 2007; 40: 1222–1230.

15.

Huang

Eapen

Steele

. Establishment of reference intervals for bone markers in children and adolescents. Clin Biochem 2011; 44: 771–778.

16.

Callegari

Reavley

Garland

. Vitamin D status, bone mineral density and mental health in young Australian women: the Safe-D study. J Public Health Res 2015; 4: 594–594.

17.

Fenner

Garland

Moore

. Web-based recruiting for health research using a social networking site: an exploratory study. J Med Internet Res 2012; 14: e20–e20.

18.

Giles

Ireland

. Dietary questionnaire for epidemiological studies (version 2), Melbourne: Cancer Council Victoria, 1996.

19.

Brown

Burton

Marshall

. Reliability and validity of a modified self-administered version of the Active Australia physical activity survey in a sample of mid-age women. Aust N Z J Public Health 2008; 32: 535–541.

20.

Roche Diagnostics GmbH. Cobas e 411 analyze, Mannheim, Germany: Roche Diagnostics GmbH, 2009.

21.

Maunsell

Wright

Rainbow

. Routine isotope-dilution liquid chromatography–tandem mass spectrometry assay for simultaneous measurement of the 25-hydroxy metabolites of vitamins D2 and D3. Clin Chem 2005; 51: 1683–1690.

22.

Levey

Stevens

Schmid

. A new equation to estimate glomerular filtration rate. Ann Intern Med 2009; 150: 604–612.

23.

Kelly

. Bone mineral density reference databases for American men and women. J Bone Miner Res 1990; 5: S249–S249.

24.

Jenkins

Black

Paul

. Age-related reference intervals for bone turnover markers from an Australian reference population. Bone 2013; 55: 271–276.

25.

Glover

Garnero

Naylor

. Establishing a reference range for bone turnover markers in young, healthy women. Bone 2008; 42: 623–630.

26.

Adami

Bianchi

Brandi

. Determinants of bone turnover markers in healthy premenopausal women. Calcif Tissue Int 2008; 82: 341–347.

27.

Vasikaran

Chubb

Schneider

H-G

. Towards optimising the provision of laboratory services for bone turnover markers. Pathology 2014; 46: 267–273.

28.

Nomura

Yoshizaki

Yoshikata

. Study of the distribution by age group of serum cross-linked C-terminal telopeptide of type I collagen and procollagen type I N-propeptide in healthy Japanese women to establish reference values. J Bone Miner Metab 2013; 31: 644–651.

29.

Deng

. Chinese bone turnover marker study: reference ranges for C-terminal telopeptide of type I collagen and procollagen I N-terminal peptide by age and gender. PLoS One 2014; 9: e103841–e103841.

30.

Walsh

Paggiosi

Eastell

. Cortical consolidation of the radius and tibia in young men and women. J Clin Endocrinol Metab 2012; 97: 3342–3348.

31.

Garnero

Munoz

Sornay-Rendu

. Associations of vitamin D status with bone mineral density, bone turnover, bone loss and fracture risk in healthy postmenopausal women. The OFELY study. Bone 2007; 40: 716–722.

32.

Garnero

Sornay-Rendu

Delmas

. Decreased bone turnover in oral contraceptive users. Bone 1995; 16: 499–503.

33.

Woitge

Knothe

Witte

. Circannual rhythms and interactions of vitamin D metabolites, parathyroid hormone, and biochemical markers of skeletal homeostasis: a prospective study. J Bone Miner Res 2000; 15: 2443–2450.

34.

Blumsohn

Naylor

Timm

. Absence of marked seasonal change in bone turnover: a longitudinal and multicenter cross-sectional study. J Bone Miner Res 2003; 18: 1274–1281.

35.

Pasco

Henry

Kotowicz

. 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.

36.

Delmas

Eastell

Garnero

. The use of biochemical markers of bone turnover in osteoporosis. Osteoporosis Int 2000; 11: S2–S17.

37.

Zhang

Peterson

. Visceral adiposity is negatively associated with bone density and muscle attenuation. Am J Clin Nutr 2015; 101: 337–343.

38.

Misra

Klibanski

. Anorexia nervosa, obesity and bone metabolism. Pediatr Endocrinol Rev 2013; 11: 21–21.

39.

Evans

Paggiosi

Eastell

. Bone density, microstructure and strength in obese and normal weight men and women in younger and older adulthood. J Bone Miner Res 2015; 30: 920–928.

40.

Clowes

Hannon

Yap

. Effect of feeding on bone turnover markers and its impact on biological variability of measurements. Bone 2002; 30: 886–890.

41.

Fraser

Anderson

Chesters

Circadian rhythm studies of serum bone resorption markers: implications for optimal sample timing and clinical utility. In: Eastell

Baumann

Hoyle

(eds). Bone markers: biochemical and clinical perspectives, London: Martin Dunitz, 2001, pp. 107–118.

42.

Hannon

Eastell

. Preanalytical variability of biochemical markers of bone turnover. Osteoporosis Int 2000; 11: S30–S44.

43.

Scott

Sale

Greeves

. The effect of training status on the metabolic response of bone to an acute bout of exhaustive treadmill running. J Clin Endocrinol Metab 2010; 95: 3918–3925.

44.

Scott

Sale

Greeves

. The role of exercise intensity in the bone metabolic response to an acute bout of weight-bearing exercise. J Appl Physiol 2011; 110: 423–432.

Bone turnover marker reference intervals in young females

Abstract

Background

Methods

Results

Conclusions

Keywords

Introduction

Methods

Participants

Sampling procedures

BTMs

Other pathology testing

BMD

Statistical analysis

Results

Discussion

Footnotes

Acknowledgements

Declaration of conflicting interests

Funding

Ethical approval

Guarantor

Contributorship

References