Bone Turnover Marker Interpreter
Bone Turnover Marker Interpreter
Read P1NP and CTX as a change from baseline on osteoporosis treatment — with the least significant change, the drug class that decides which direction to expect, and the fasting morning sample that makes a CTX interpretable at all.
P1NP and CTX on treatment
Drug class + both markers + samplingThree months after starting an oral bisphosphonate. P1NP essentially unchanged from baseline; CTX has fallen, but by about 30%. Both samples were taken fasting before 10 a.m.
Two markers, two directions, one sampling rule
Anabolics — teriparatide, abaloparatide: P1NP rises, early and steeply.
Romosozumab: P1NP rises and resorption falls, at the same time.
Adequate response on the IOF/ECTS criteria: a fall of more than 38% in P1NP or more than 56% in CTX, at three months.
- The reference markers
- “procollagen type I N propeptide (PINP) and β isomerized C-terminal telopeptide of type I collagen (β-CTX-I)”, designated by the IOF and IFCC as the reference markers of formation and resorption so that studies can be compared
- Least significant change
- two different figures are in circulation and they answer different questions. The IOF/ECTS adherence thresholds — 38% for P1NP, 56% for CTX — come from a treated cohort and answer “has this drug worked?”. The 2025 consensus’s reference change values — −19.9% for PINP, −30.8% for β-CTX-I, from within-subject variations of 8.8% and 15.1% — answer “has this number really moved?”. This page uses the first pair
- Why a reference interval is not enough
- within-subject variation is 8.8% for PINP and 15.1% for β-CTX-I, and the between-subject spread is wider still, so an individual’s result can move substantially inside the reference interval or sit outside it without anything having changed. The comparison that carries information is with the patient’s own baseline
- The CTX sampling rule
- collect “between 7:30 and 10 am and after overnight fasting”. β-CTX-I peaks around 5 a.m. and bottoms out around 2 p.m., falls about 21% an hour through the day in untreated patients, and “food consumption can reduce serum β-CTX-I by up to 40%”. P1NP is far less affected by either
Worked example
Three months after starting an oral bisphosphonate. P1NP essentially unchanged from baseline; CTX has fallen, but by about 30%. Both samples were taken fasting before 10 a.m.
Three months is the recommended interval — the IOF/ECTS strategy is baseline and three months, because markers move long before densitometry can
The sampling was correct, so the CTX result can be used. Had it been a fed or afternoon sample the number would have been uninterpretable, since food alone can lower CTX by up to 40%
P1NP unchanged: the threshold for an adequate response is a fall of more than 38%, so this fails
CTX down about 30%: the threshold is a fall of more than 56%, so this fails too — even though 30% looks like a substantial drop
Neither marker meets the criterion, so this is not an adequate response, and the published instruction is to reassess for "problems with the treatment, mainly low adherence"
With an oral bisphosphonate that means asking exactly how the tablet is taken, because bioavailability is under 1% even when it is taken correctly and a tablet with breakfast is very nearly a tablet not taken
Then vitamin D status, malabsorption, a secondary cause of high turnover, and whether the baseline was a genuine pre-treatment sample — before concluding the drug does not work
Which way each drug class moves the markers
| Treatment | P1NP (formation) | CTX (resorption) | When to measure |
|---|---|---|---|
| Oral bisphosphonate | Falls — 48 to 63% in the TRIO cohort | Falls — 68 to 81% in the TRIO cohort | Baseline and 3 months (CTX), 6 months for P1NP has also been suggested |
| Intravenous bisphosphonate | Falls | Falls | CTX at 1 month, P1NP at 3 months |
| Denosumab | Falls | Falls, rapidly and profoundly | Baseline and ~3 months; a rise later is a warning that a dose is overdue |
| Teriparatide / abaloparatide | Rises, immediately, peaking at 1 to 3 months | Rises later, secondarily | 1 to 3 months, to catch the peak |
| Romosozumab | Rises — peak at month 1 in one cohort, still well above baseline at 3 months | Falls at the same time | Month 1 and month 3 |
The two least significant changes, and what each answers
| Figure | P1NP | CTX | The question it answers |
|---|---|---|---|
| IOF/ECTS adherence threshold (TRIO study, treated cohort) | Fall of more than 38% | Fall of more than 56% | “Has this drug worked?” — used by the selectors on this page |
| ESCEO/IOF/IFCC 2025 reference change value | −19.9% for a fall, +24.8% for a rise | −30.8% for a fall, +44.5% for a rise | “Has this number really moved, given the noise?” |
| Within-subject variation (CVI) | 8.8% | 15.1% | The input to the reference change value: LSC = √2 × 1.96 × CVI = 2.77 × CVI |
| Absolute target | Below 30 µg/L, or below the premenopausal median | Below 0.25 µg/L, or below the premenopausal median | “Has this patient reached the target?” — an alternative to a percentage change |
| Detection of a poor response | 84% using P1NP alone | 87% using CTX alone | 94.5% when both are measured and either one is allowed to signal |
A change from baseline, in the direction the drug predicts, on a sample taken properly
The International Osteoporosis Foundation and the International Federation of Clinical Chemistry designated two reference markers so that studies could be compared with one another: P1NP for bone formation and β-CTX-I for bone resorption. Their clinical use is narrower than the enthusiasm around them suggests. They do not diagnose osteoporosis, they do not decide who should be treated, and they are not a substitute for a fracture risk assessment or for densitometry — “data supporting the use of BTMs as independent diagnostic tools remain insufficient”. What they do, and what nothing else does, is show within weeks whether a treatment is working, at a point when a repeat DXA scan would still be years away from giving an answer.
Everything depends on reading them as a change from the patient’s own baseline rather than against a reference interval, because the noise is large. The within-subject variation is 8.8% for P1NP and 15.1% for β-CTX-I, and the between-subject spread is wider still, so an individual’s result can move a long way inside the reference interval, or sit outside it permanently, without anything having happened. That is what the least significant change exists for — and there are two figures in circulation, which is a recurrent source of confusion. The IOF/ECTS adherence recommendation, derived from the TRIO study in treated patients, sets an adequate response at a fall of more than 38% for P1NP or more than 56% for CTX at three months. The 2025 ESCEO/IOF/IFCC consensus gives reference change values of −19.9% for PINP and −30.8% for β-CTX-I, computed as √2 × 1.96 × the within-subject variation. Those answer different questions — the first is “has this drug worked?”, the second is “has this number really moved?” — and this page uses the first pair, because that is the question a monitoring sample is asked. Measuring both markers is worth it: detection of a poor response is 84% with P1NP alone and 87% with CTX alone, and “if variation in at least one is considered when measuring both, the level of detection is 94.5%”.
The direction to expect depends entirely on the drug, and the two markers do not always move together. An antiresorptive — a bisphosphonate or denosumab — suppresses both: falls of 10 to 81% are reported across studies, and in the TRIO cohort of oral bisphosphonates CTX fell by 68 to 81% and P1NP by 48 to 63%. An anabolic does the opposite to the formation marker: with teriparatide or abaloparatide, “P1NP levels begin to rise immediately after treatment initiation, with peak levels typically occurring within 1 to 3 months”, and increases from 18 to 444% are reported. Romosozumab breaks the dichotomy altogether, because it raises formation and lowers resorption at the same time — in one cohort P1NP “increased the most (by 183.1 ± 124.0 ng/ml) at Month 1” and was still 135.0 ± 102.6 ng/ml above baseline at three months, while the resorption marker “significantly decreased at each time point from baseline”. So the same finding means opposite things in different patients, and a falling P1NP is an excellent result on alendronate and a treatment failure on teriparatide.
Which brings the discussion to the pre-analytics, and to CTX in particular, because a CTX taken carelessly is worse than no CTX at all. β-CTX-I has a large circadian rhythm and falls sharply after eating. It peaks “very early in the morning (around 5 am)” and reaches its nadir “in the afternoon at around 2 pm”; measured directly in untreated patients, “CTX decreased with 21% per hour” through the day; and “food consumption can reduce serum β-CTX-I by up to 40%”. Forty per cent is the same order of magnitude as a genuine treatment effect, which means a fed afternoon sample can manufacture a response that is not there, or, if the baseline was the fed sample, conceal one that is. The published requirement is therefore specific: collection “between 7:30 and 10 am and after overnight fasting”, ideally into EDTA plasma and centrifuged promptly, because that is the most stable preparation. One asymmetry is worth knowing, because it tells you where the discipline matters most: the diurnal swing is largely a feature of the untreated skeleton, and in patients already established on an antiresorptive one study found “no association between P1NP and CTX and hour of collection”. It is the baseline whose timing is critical, and that is usually the sample taken with the least ceremony. P1NP, by contrast, “is minimally affected by food and circadian rhythm”, which is why it is the more robust marker in ordinary clinic conditions and the one to fall back on when the sampling cannot be controlled.
When the markers do not move as expected, the first explanation is almost always adherence, and the published instruction says so: “if no decrease occurs, the clinician should reassess to identify problems with the treatment, mainly low adherence”. With oral bisphosphonates the regimen is unforgiving — on waking, on an empty stomach, with plain water only, upright and nothing else for at least 30 minutes — and bioavailability is under 1% even when all of that is done correctly, so a tablet taken with breakfast is functionally a tablet not taken. Monitoring is itself part of the treatment here: one analysis found that “monitoring the subjects under treatment increased adherence to therapy by 57% compared to unmonitored subjects”. Beyond adherence, look for vitamin D deficiency — the vitamin D status interpreter and the vitamin D loading dose calculator cover that — malabsorption, an untreated secondary cause of high turnover, and a baseline that was never a true pre-treatment value. One specific alarm deserves naming: a rising CTX in a patient on denosumab whose next injection is overdue reflects the rebound in turnover that follows discontinuation, and that rebound carries a recognised risk of multiple vertebral fractures. Finally, neither marker can be used in chronic kidney disease — both “are not suitable in the setting of CKD, as they accumulate with kidney dysfunction”. For the unit arithmetic, the CTX converter and the P1NP converter are the tools, and for the density measurement these markers complement rather than replace, the BMD T-score and Z-score calculator.
Frequently asked questions
Why does CTX have to be taken fasting in the morning?
Because β-CTX-I has a large circadian rhythm and falls after eating, and both effects are big enough to swamp a treatment response. It peaks “very early in the morning (around 5 am)” and reaches its nadir “in the afternoon at around 2 pm”; in untreated patients CTX was measured falling by about 21% per hour through the day; and “food consumption can reduce serum β-CTX-I by up to 40%”. A 40% fall is the same order as a real drug effect. The published requirement is collection “between 7:30 and 10 am and after overnight fasting”, ideally into EDTA plasma centrifuged promptly. P1NP is minimally affected by food or time of day, which makes it the more robust marker when sampling cannot be controlled.
How much should P1NP and CTX fall on treatment?
On the IOF/ECTS adherence recommendation, an adequate response at three months is a fall of more than 38% in P1NP or more than 56% in CTX. Those thresholds come from the TRIO study of oral bisphosphonates, where the average falls were 48 to 63% for P1NP and 68 to 81% for CTX. An alternative target is absolute rather than relative: returning the marker to the lower half of the premenopausal reference interval, defined by its median, or in one fracture cohort P1NP below 30 µg/L and CTX below 0.25 µg/L. Note that a different and smaller pair of figures — reference change values of −19.9% for PINP and −30.8% for β-CTX-I — is used to answer a different question, namely whether a number has moved beyond analytical and biological noise.
What is the least significant change for bone turnover markers?
It depends which question you are asking, and two sets of figures are published. For “has this drug worked?”, the IOF/ECTS thresholds derived from a treated cohort are a fall of more than 38% for P1NP and more than 56% for CTX. For “has this number really moved, given the noise?”, the 2025 ESCEO/IOF/IFCC consensus gives reference change values of −19.9% for a fall in PINP (+24.8% for a rise) and −30.8% for a fall in β-CTX-I (+44.5% for a rise), computed as √2 × 1.96 × the within-subject coefficient of variation, which is 8.8% for PINP and 15.1% for β-CTX-I. Neither set is wrong; they are answering different questions, and the larger thresholds are the ones used for treatment monitoring.
Do bone turnover markers go up or down on treatment?
It depends entirely on the drug, and the markers do not always move together. Antiresorptives — bisphosphonates and denosumab — suppress both P1NP and CTX, with published decreases from 10% to 81%. Anabolic agents do the opposite to the formation marker: with teriparatide or abaloparatide, P1NP begins to rise immediately and peaks at one to three months, and anabolic increases from 18% to 444% are reported. Romosozumab is neither: it raises formation and suppresses resorption at the same time, so P1NP rises while the resorption marker falls in the same patient. A falling P1NP is an excellent result on alendronate and a treatment failure on teriparatide.
Can bone turnover markers diagnose osteoporosis?
No. They are monitoring tools, not diagnostic ones — “data supporting the use of BTMs as independent diagnostic tools remain insufficient” — and they do not identify who has osteoporosis or who should be treated. Their value is that they change within weeks of starting or stopping a drug, long before a repeat densitometry scan could show anything, so they answer the question “is this treatment working?” rather than “does this patient need treatment?”. Diagnosis rests on bone mineral density and fracture history; the BMD T-score and Z-score calculator covers that side.
What does it mean if the markers have not fallen on a bisphosphonate?
Adherence first. The published instruction is that “if no decrease occurs, the clinician should reassess to identify problems with the treatment, mainly low adherence”, and with an oral bisphosphonate that means going through the administration in detail — on waking, empty stomach, plain water only, upright, nothing else for at least 30 minutes — because bioavailability is under 1% even when it is done correctly. Then check vitamin D status, malabsorption, an untreated secondary cause of high turnover such as hyperthyroidism or hyperparathyroidism, whether the baseline was a genuine pre-treatment sample taken under the same conditions, and renal function, since both markers accumulate in chronic kidney disease and are not suitable there. Changing drug before excluding non-adherence usually just produces a second drug that is not taken.
Related calculators
References
- Bhattoa HP, Vasikaran S, Trifonidi I, et al. Update on the role of bone turnover markers in the diagnosis and management of osteoporosis: a consensus paper from the European Society for Clinical and Economic Aspects of Osteoporosis, Osteoarthritis and Musculoskeletal Diseases (ESCEO), the International Osteoporosis Foundation (IOF) and the International Federation of Clinical Chemistry and Laboratory Medicine (IFCC). Osteoporos Int. 2025;36:579–608. “procollagen type I N propeptide (PINP) and β isomerized C-terminal telopeptide of type I collagen (β-CTX-I)” as reference markers; within-subject variation 8.8% for PINP and 15.1% for β-CTX-I; reference change values “−19.9% for decreases and at +24.8% for increases” (PINP) and “−30.8% for decreases and at+44.5% for increases” (β-CTX-I); the least significant change “defned as √2×1.96× CVI=2.77× CVI”; collection “between 7:30 and 10 am and after overnight fasting”; “food consumption can reduce serum β-CTX-I by up to 40%”; a circadian “peak very early in the morning (around 5 am) and a nadir in the afternoon at around 2 pm”; EDTA plasma preferred; antiresorptive therapy showing “a decrease of BTMs ranging from − 10 to − 81%” and anabolic therapy “an increase from 18 to 444%”; the target of returning markers “to the lower half of the reference interval for premenopausal women, defined by the normal median”; “Monitoring the subjects under treatment increased adherence to therapy by 57% compared to unmonitored subjects”; and that the markers “are not suitable in the setting of CKD, as they accumulate with kidney dysfunction”.
- Diez-Perez A, Naylor KE, Abrahamsen B, et al. International Osteoporosis Foundation and European Calcified Tissue Society Working Group. Recommendations for the screening of adherence to oral bisphosphonates. Osteoporos Int. 2017;28:767–774. Measurement at baseline and 3 months; an adequate response as a PINP “decrease of more than 38%” or a CTX “decrease of more than 56%”; detection of poor response 84% with PINP alone and 87% with CTX alone, and “if variation in at least one is considered when measuring both, the level of detection is 94.5%”; and “if no decrease occurs, the clinician should reassess to identify problems with the treatment, mainly low adherence”.
- Naylor KE, Jacques RM, Paggiosi M, et al. Response of bone turnover markers to three oral bisphosphonate therapies in postmenopausal osteoporosis: the TRIO study. Osteoporos Int. 2016;27:21–31. 172 postmenopausal women randomised to alendronate, ibandronate or risedronate; in women whose baseline PINP was above the reference interval the fall at 12 weeks was “−63 %” overall and “−67 %” in the good-compliance group. The source of the 38% and 56% thresholds above.
- Szulc P, Naylor K, Hoyle NR, Eastell R, Leary ET; National Bone Health Alliance Bone Turnover Marker Project. Use of CTX-I and PINP as bone turnover markers: National Bone Health Alliance recommendations to standardize sample handling and patient preparation to reduce pre-analytical variability. Osteoporos Int. 2017;28(9):2541–2556. PINP and CTX-I as “the reference markers of bone turnover for the fracture risk prediction and monitoring of osteoporosis treatment”; “Samples for CTX-I must be collected consistently in the morning hours in the fasted state”; “EDTA plasma is preferred for CTX-I for its greater sample stability”.
- Borgen TT, Solberg LB, Lauritzen T, et al. Target Values and Daytime Variation of Bone Turnover Markers in Monitoring Osteoporosis Treatment After Fractures. JBMR Plus. 2022;6(6):e10633. In patients not on an antiresorptive, “CTX decreased with 21% per hour” during the day, while “In patients using ARD, there was no association between P1NP and CTX and hour of collection of blood sample”; “P1NP <30 μg/L and CTX <0.25 μg/L yield the best discrimination between patients taking and not taking ARD and the best prediction of BMD gains after 2 years”; and an expected antiresorptive response defined as “a reduction in BTM of more than the least significant change or to a value below the mean levels of the premenopausal reference range”.
- Williams C, Anastasopoulou C, Sapra A. Biochemical Markers of Osteoporosis. In: StatPearls. Treasure Island (FL): StatPearls Publishing; updated 26 November 2024. NCBI Bookshelf NBK559306. The IOF and IFCC “recommend serum P1NP and CTX-1 as the preferred markers for bone formation and resorption, respectively”; for intravenous agents “CTX-1 and P1NP levels should be checked after 1 and 3 months of treatment, respectively” and for oral bisphosphonates “measured at 3 and 6 months, respectively”; with anabolic agents “P1NP levels begin to rise immediately after treatment initiation, with peak levels typically occurring within 1 to 3 months”; “CTX-1 exhibits circadian rhythm fluctuations that are influenced by food intake, whereas P1NP is minimally affected by food and circadian rhythm”; and “data supporting the use of BTMs as independent diagnostic tools remain insufficient”.
- Hattori K, Kanayama Y. Association between P1NP value and increases in bone mineral density in patients with postmenopausal osteoporosis treated with romosozumab. Mod Rheumatol. 2024;34(5):1047–1055. “P1NP value increased the most (by 183.1 ± 124.0 ng/ml) at Month 1”, remaining 135.0 ± 102.6 ng/ml above baseline at month 3, while “TRACP-5b significantly decreased at each time point from baseline”; romosozumab “increases bone formation and decreasing bone resorption” simultaneously. Note that the resorption marker in this study was TRACP-5b and not CTX, so the direction is established here and the CTX magnitude is not.
Medical Disclaimer: The tools and content provided here are for educational and reference purposes only. They are not intended to substitute for professional medical advice, diagnosis, or treatment. Clinical decisions should always be based on the comprehensive assessment of a qualified healthcare professional.