A large prospective study suggests that a routine creatinine blood test may help flag people at higher risk of osteoporosis and fractures, revealing a narrow “sweet spot” where muscle and kidney health align to protect bone strength.
Study: Serum creatinine, genetic susceptibility, and the risk of osteoporosis and fracture: a prospective cohort study from the UK Biobank. Image credit: Nanci Santos Iglesias/Shutterstock.com
Osteoporosis (decreased bone mineralization) poses a great challenge to healthy aging. Yet there are still few risk biomarkers capable of evolving to meet this growing need. A recent study published in the journal Frontiers in Endocrinology attempts to fill this gap.
The search for affordable biomarkers to predict bone loss
Osteoporosis is a health condition characterized by reduced bone mass and weakening of the skeletal microarchitecture, making bones fragile and more prone to fractures. It is common in older adults.
In 2019, there were about 178 million fractures worldwide. While not directly analyzed in the present study, Chinese estimates for 2050 indicate nearly 6 million fractures, highlighting the growing global burden of bone fragility.
Osteoporosis risk is currently estimated using tools such as the widely used WHO Fracture Risk Assessment Tool (FRAX). These integrate dual-energy X-ray absorptiometry (DXA) with clinical and sociodemographic risk factors, age, sex, smoking, body mass index (BMI), and medication use.
However, these require sophisticated equipment, limiting their accessibility for primary care and driving up costs. Their accuracy also varies across populations. Affordable and accessible quantitative biomarkers are needed to enhance early detection and risk prediction, enabling timely prevention.
Serum creatinine originates largely from muscle and is cleared by the kidneys. It is commonly used to examine muscle and kidney health. However, it may also reflect fracture risk.
For instance, low muscle mass is linked to low creatinine levels. Improving muscle strength and mass via resistance exercise and optimal nutrition is associated with lower fracture risk. Conversely, elevated creatinine levels occur in chronic kidney disease (CKD). Through multiple secondary mechanisms, it is associated with bone and mineral disorders, and thus with a higher fracture risk. Despite this, the role of serum creatinine in predicting osteoporosis risk remains under-researched.
Genome-wide association studies (GWAS) have identified several single-nucleotide polymorphisms (SNPs) associated with altered bone mineralization. These form the basis of polygenic risk scores (PRS), which quantitatively estimate genetic predisposition to osteoporosis and can improve predictive accuracy. The current study further examined whether associations between creatinine and osteoporosis or fractures were consistent across different levels of genetic risk, rather than assuming genetic risk plays no role.
A large UK Biobank analysis of creatinine and bone risk
The current prospective study aimed to uncover associations between baseline serum creatinine values and new-onset osteoporosis and fractures, further stratified by PRS. This could help position creatinine as a practical and convenient part of a dose-dependent risk assessment tool.
The study was embedded in the UK Biobank cohort, comprising over half a million participants of middle age or older. Participants completed surveys and underwent physical measurements and laboratory testing, in addition to genotyping. Health records were used for longitudinal follow-up.
Baseline creatinine was treated as both a continuous and a categorical variable. For categorical analyses, it was classified into six categories (G1–G6), with G4 (70–80 µmol/L) used as the reference group. The researchers applied diverse analytical approaches to ensure robust conclusions.
These included Cox proportional hazards models and restricted cubic splines (RCS) to map adjusted dose–response curves. Sex- and age-stratified associations were examined. PRS tertiles were analyzed within each creatinine category to assess osteoporosis risk and to test for potential interaction effects.
Low and high creatinine signal different fracture pathways
The study included 385,576 participants. Patients with and without osteoporosis had a mean age of approximately 61 years and 56 years, respectively. Women predominated among new osteoporosis cases (83 %), and 92 % of cases occurred in White participants.
There were 26,474 participants with fractures, with a mean age of approximately 59 years. About 54 % were men, and 91 % were White.
RCS analysis demonstrated a U-shaped association between serum creatinine and osteoporosis risk. Osteoporosis risk decreased by 2.0 % per 1-µmol/L increase in creatinine within the range of 25–80 µmol/L. Conversely, beyond this range (80–130 µmol/L), the risk increased by 1.1 % per 1-µmol/L increase in creatinine.
This biphasic curve may reflect two distinct biological pathways. The left arm of the curve (lower creatinine) may be associated with low muscle mass, malnutrition, and reduced mechanical loading. These states are plausibly linked to inflammation, impaired anabolic signaling, and reduced myokine activity, contributing to increased bone resorption and fracture risk.
Conversely, the right arm of the curve (higher creatinine) may reflect impaired kidney function and related disturbances in mineral metabolism, including CKD-associated bone and mineral disorders, which negatively affect bone density.
Fracture risk followed a J-shaped curve. At lower creatinine levels, fracture risk declined by 1.0 % per 1-µmol/L increase in creatinine between 25 and 80 µmol/L. Beyond this inflection point, fracture risk increased by 0.3 % per 1-µmol/L increase up to 130 µmol/L, although this increase was modest and not consistently statistically significant across all models.
When stratified by sex, broadly similar trajectories were observed, but important differences emerged. In men, osteoporosis risk declined by 4.6 % within the 25–80 µmol/L range, with little evidence of a clear increase at higher levels in categorical analyses. In patients with impaired kidney function, the protective effects of male sex and higher BMI may fail to offset declining bone density.
In women, osteoporosis risk declined by 1.8 % within the lower creatinine range and increased by 1.7 % at higher creatinine levels. Although women exhibited higher overall rates of osteoporosis and fractures, the authors emphasize that creatinine thresholds should be interpreted primarily within each sex, given the markedly different baseline creatinine distributions. Notably, the nadir of osteoporosis risk occurred around G5 in men and G4 in women.
Genetic risk did not substantially modify these associations. While PRS was associated with overall osteoporosis risk, interaction tests showed that the relationship between creatinine and osteoporosis was broadly consistent across PRS tertiles.
For example, among participants with low PRS, osteoporosis risk in G1 was approximately 93 % higher than in G4, whereas G3 showed only a minimal, non-significant increase. Among those with high PRS, osteoporosis risk increased by about 59 % in G1 (hazard ratio HR ≈ 1.59) and by approximately 11 % in G3, relative to G4.
Across age groups, osteoporosis risk decreased with increasing baseline creatinine up to the reference range, with this pattern becoming more pronounced in older participants (>65 years).
Creatinine emerges as a promising adjunct bone risk marker
There is a significant but nonlinear association between baseline serum creatinine levels and the risk of new-onset osteoporosis and fractures. Serum creatinine levels around 80 µmol/L were associated with the lowest observed risk.
These findings suggest that serum creatinine may serve as a low-cost adjunct biomarker for clinical risk stratification and surveillance of osteoporosis, although causal inference cannot be established. Future studies should validate these findings across more diverse populations and incorporate objective measures of muscle mass and strength to further refine risk prediction.
Download your PDF copy now!
Common blood signals explain why chronic diseases cluster as we age