Nicholas J. Wareham
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Nicholas J. Wareham has a total of 1000 co-authors. Here are the top co-authors by number of shared publications:
- Claudia Langenberg
Queen Mary University of London
82 shared publications - Kay-Tee Khaw
Department of Public Health and Primary Care, University of Cambridge, Cambridge CB1 8RN, UK
70 shared publications - Nita Gandhi Forouhi
University of Cambridge, Cambridge Biomedical Campus
62 shared publications - Elio Riboli
Department of Epidemiology and Biostatistics, School of Public Health, Imperial College London, London, UK
48 shared publications - Rosario Tumino
Hyblean Association for Epidemiological Research, AIRE ONLUS Ragusa, Ragusa, Italy
46 shared publications - Kim Overvad
Department of Public Health, Aarhus University, Bartholins Allé 2, 8000 Aarhus, Denmark
44 shared publications
Latest Publications
- 20 October 2026
- Published by Wolters Kluwer Health in Journal of the American Heart Association
- Vol. 15 (20)
Omega‐6 Polyunsaturated Fatty Acids and Peripheral Artery Disease: An Individual Participant‐Level Analysis of 9 International Prospective Studies
Background: Omega‐6 polyunsaturated fatty acids (PUFAs) have previously been associated with several peripheral artery disease (PAD) risk factors. We studied the association between individual omega‐6 PUFAs and incident PAD in a multicohort analysis. Methods: This study was conducted among participants from 9 cohorts (N=137 704; PAD cases=2014) in the FORCE (Fatty Acids and Outcomes Research Consortium) who were free of clinical PAD at baseline. Omega‐6 PUFAs were measured at baseline and with a median follow‐up of 14 years. Cox proportional hazards regression analyses were conducted in individual cohorts to evaluate associations of omega‐6 PUFAs (continuous cohort‐standardized interquintile ranges [difference between the medians of the 10th and 90th omega‐6 PUFAs percentiles] and cohort‐specific quintiles) with incident clinical PAD adjusting for prespecified covariates. Inverse‐variance weighted meta‐analysis was used to calculate the combined hazard ratios (HRs) and 95% CIs. Results: Linoleic acid (LA) was measured in all 9 cohorts, arachidonic acid in 8 (N=33 541; PAD cases=1766), dihomo‐γ‐LA and γ‐LA in 7 (N=29 861; PAD cases=1686), and eicosadienoic acid in 6 (N=27 538; PAD cases=1435). Only LA was associated with a lower risk of incident PAD (quintile 5 versus quintile 1: HR, 0.75 [95% CI, 0.64–0.89]; P =0.0009). Results did not differ in a statistically significant manner by age, sex, diabetes status, or smoking status. Conclusions: Results suggest that the omega‐6 PUFA LA may be protective against PAD, and other omega‐6 PUFAs are not associated with PAD incidence. This is further evidence that omega‐6 PUFA—disease relationships should be evaluated by individual fatty acids rather than grouped as total omega‐6 PUFAs.
- 20 October 2026
- Published by Wolters Kluwer Health in Journal of the American Heart Association
- Vol. 15 (20)
Omega‐6 Polyunsaturated Fatty Acids and Peripheral Artery Disease: An Individual Participant‐Level Analysis of 9 International Prospective Studies
Background: Omega‐6 polyunsaturated fatty acids (PUFAs) have previously been associated with several peripheral artery disease (PAD) risk factors. We studied the association between individual omega‐6 PUFAs and incident PAD in a multicohort analysis. Methods: This study was conducted among participants from 9 cohorts (N=137 704; PAD cases=2014) in the FORCE (Fatty Acids and Outcomes Research Consortium) who were free of clinical PAD at baseline. Omega‐6 PUFAs were measured at baseline and with a median follow‐up of 14 years. Cox proportional hazards regression analyses were conducted in individual cohorts to evaluate associations of omega‐6 PUFAs (continuous cohort‐standardized interquintile ranges [difference between the medians of the 10th and 90th omega‐6 PUFAs percentiles] and cohort‐specific quintiles) with incident clinical PAD adjusting for prespecified covariates. Inverse‐variance weighted meta‐analysis was used to calculate the combined hazard ratios (HRs) and 95% CIs. Results: Linoleic acid (LA) was measured in all 9 cohorts, arachidonic acid in 8 (N=33 541; PAD cases=1766), dihomo‐γ‐LA and γ‐LA in 7 (N=29 861; PAD cases=1686), and eicosadienoic acid in 6 (N=27 538; PAD cases=1435). Only LA was associated with a lower risk of incident PAD (quintile 5 versus quintile 1: HR, 0.75 [95% CI, 0.64–0.89]; P =0.0009). Results did not differ in a statistically significant manner by age, sex, diabetes status, or smoking status. Conclusions: Results suggest that the omega‐6 PUFA LA may be protective against PAD, and other omega‐6 PUFAs are not associated with PAD incidence. This is further evidence that omega‐6 PUFA—disease relationships should be evaluated by individual fatty acids rather than grouped as total omega‐6 PUFAs.
- Open Access
- 24 September 2026
- Published by Elsevier in Ophthalmology
Prevention of Retinal Detachment in Type 2 Stickler Syndrome
- 24 September 2026
- Published by Elsevier in Ophthalmology
Open AccessPrevention of Retinal Detachment in Type 2 Stickler Syndrome
- 9 September 2026
- Published by American Association for the Advancement of Science (AAAS) in Science Translational Medicine
- Vol. 18 (866)
Proteomics identify disease-associated variants in patients with rare diseases undiagnosed after genome sequencing
Despite the introduction of genome sequencing (GS) for rare disease diagnostics, a genetic cause is not identified in most patients. Here, we explored the potential of proteomics to improve the diagnostic yield in 424 patients with rare diseases from the 100,000 Genomes Project (100kGP) without a genetic diagnosis. Serum proteomic profiling was performed using the Olink Explore 1536 assay ( N = 1463 proteins). For 13 patients without genetic diagnoses, detection of lower serum protein “outliers” ( z -score < −2) led to confirmed genetic diagnoses by resolving variants of uncertain significance or prioritizing genes for targeted GS reanalysis. For 23 additional patients without genetic diagnoses (64% of findings), we identified candidate gene-disease links and variants through convergent evidence from lower protein outliers and variants ranked through the variant prioritization tool Exomiser. For example, we identified a candidate heterozygous missense variant [Genome Aggregation Database (gnomAD) minor allele frequency = 0.006%] in tyrosine kinase with immunoglobulin-like and epidermal growth factor homology domains 1 ( TIE1 ) that was only present in a patient with lower TIE1 serum abundance ( z -score = −5.12) and their father, both of whom were affected by the same monogenic cardiac disorder, but in no other individuals from the 100kGP. Missense (52.5%) and splice region (27.5%) variants accounted for most diagnostic or candidate variants prioritized. This proof-of-principle study demonstrated that serum proteomics can support rare disease diagnosis and identify disease-causing genes in patients undiagnosed after GS, although successful implementation will likely depend on tissue specificity of protein expression, detectability in blood, proteomic platform coverage, and sensitivity.
- 9 September 2026
- Published by American Association for the Advancement of Science (AAAS) in Science Translational Medicine
- Vol. 18 (866)
Proteomics identify disease-associated variants in patients with rare diseases undiagnosed after genome sequencing
Despite the introduction of genome sequencing (GS) for rare disease diagnostics, a genetic cause is not identified in most patients. Here, we explored the potential of proteomics to improve the diagnostic yield in 424 patients with rare diseases from the 100,000 Genomes Project (100kGP) without a genetic diagnosis. Serum proteomic profiling was performed using the Olink Explore 1536 assay ( N = 1463 proteins). For 13 patients without genetic diagnoses, detection of lower serum protein “outliers” ( z -score < −2) led to confirmed genetic diagnoses by resolving variants of uncertain significance or prioritizing genes for targeted GS reanalysis. For 23 additional patients without genetic diagnoses (64% of findings), we identified candidate gene-disease links and variants through convergent evidence from lower protein outliers and variants ranked through the variant prioritization tool Exomiser. For example, we identified a candidate heterozygous missense variant [Genome Aggregation Database (gnomAD) minor allele frequency = 0.006%] in tyrosine kinase with immunoglobulin-like and epidermal growth factor homology domains 1 ( TIE1 ) that was only present in a patient with lower TIE1 serum abundance ( z -score = −5.12) and their father, both of whom were affected by the same monogenic cardiac disorder, but in no other individuals from the 100kGP. Missense (52.5%) and splice region (27.5%) variants accounted for most diagnostic or candidate variants prioritized. This proof-of-principle study demonstrated that serum proteomics can support rare disease diagnosis and identify disease-causing genes in patients undiagnosed after GS, although successful implementation will likely depend on tissue specificity of protein expression, detectability in blood, proteomic platform coverage, and sensitivity.
- 4 September 2026
- Published by American Association for the Advancement of Science (AAAS) in Science Advances
- Vol. 12 (36)
Glycemic and bodyweight effects of GIPR coding variation reflect differences in surface expression and intrinsic functional impairment
The glucose-dependent insulinotropic polypeptide receptor (GIPR) is a major therapeutic target in type 2 diabetes and obesity. Missense variation in GIPR could confer phenotypic effects through alterations to constitutive activity or functional responses to GIP or pharmacological agonists. In this study, we aimed to provide a deep understanding of the molecular mechanisms that underpin the cellular and physiological impacts of GIPR coding variation by studying 30 GIPR coding variants in cellular models and pancreatic islets. Many variants showed impaired GIP-induced cyclic adenosine monophosphate responses, and population-based association analysis highlighted that these loss-of-function variants decrease body mass index but increase glycemia. In many cases, reduced function was partly driven by reduced expression at the cell surface due to impaired stability and redirection toward proteasomal degradation. Molecular dynamics simulations suggest distinct variant-induced perturbations in inter- and intrahelical interactions, which interfere with receptor stability. This study highlights the mechanisms and consequences of GIPR coding variation, which may have implications for the therapeutic targeting of this receptor.
- 4 September 2026
- Published by American Association for the Advancement of Science (AAAS) in Science Advances
- Vol. 12 (36)
Glycemic and bodyweight effects of GIPR coding variation reflect differences in surface expression and intrinsic functional impairment
The glucose-dependent insulinotropic polypeptide receptor (GIPR) is a major therapeutic target in type 2 diabetes and obesity. Missense variation in GIPR could confer phenotypic effects through alterations to constitutive activity or functional responses to GIP or pharmacological agonists. In this study, we aimed to provide a deep understanding of the molecular mechanisms that underpin the cellular and physiological impacts of GIPR coding variation by studying 30 GIPR coding variants in cellular models and pancreatic islets. Many variants showed impaired GIP-induced cyclic adenosine monophosphate responses, and population-based association analysis highlighted that these loss-of-function variants decrease body mass index but increase glycemia. In many cases, reduced function was partly driven by reduced expression at the cell surface due to impaired stability and redirection toward proteasomal degradation. Molecular dynamics simulations suggest distinct variant-induced perturbations in inter- and intrahelical interactions, which interfere with receptor stability. This study highlights the mechanisms and consequences of GIPR coding variation, which may have implications for the therapeutic targeting of this receptor.
- Open Access
- 29 June 2026
- Published by Springer Nature in Nature Aging
- Vol. 6 (7) , 1-15
Associations of proteomic age clocks with lifestyle risk factors, incident chronic diseases and mortality in two European cohorts
Assessment of biological aging using proteomic clocks may enhance risk prediction and elucidate the molecular links between aging and chronic diseases. Here, among 17,473 participants of the European Prospective Investigation into Cancer and Nutrition, we examined associations of plasma SomaScan-based proteomic clocks, including organ-specific clocks, with risk factors, 24 incident chronic diseases and all-cause mortality, over up to 28 years of follow-up. Replication was conducted in the Whitehall II study. We show that the global age gap, an age acceleration score combining proteomic clocks, was associated with smoking, alcohol consumption, physical inactivity and higher risk of mortality, cardiovascular diseases, dementia and cancers of the liver, upper aero-digestive tract, lung and kidney. Lung, kidney and stomach cancers were more strongly associated with related organ-specific age gaps. Predictive performance of proteomic clocks for mortality was comparable to that of classical lifestyle risk factors. In summary, proteomic clocks appear promising biomarkers of generalized age-related disease risk.
- 29 June 2026
- Published by Springer Nature in Nature Aging
- Vol. 6 (7) , 1-15
Open AccessAssociations of proteomic age clocks with lifestyle risk factors, incident chronic diseases and mortality in two European cohorts
Assessment of biological aging using proteomic clocks may enhance risk prediction and elucidate the molecular links between aging and chronic diseases. Here, among 17,473 participants of the European Prospective Investigation into Cancer and Nutrition, we examined associations of plasma SomaScan-based proteomic clocks, including organ-specific clocks, with risk factors, 24 incident chronic diseases and all-cause mortality, over up to 28 years of follow-up. Replication was conducted in the Whitehall II study. We show that the global age gap, an age acceleration score combining proteomic clocks, was associated with smoking, alcohol consumption, physical inactivity and higher risk of mortality, cardiovascular diseases, dementia and cancers of the liver, upper aero-digestive tract, lung and kidney. Lung, kidney and stomach cancers were more strongly associated with related organ-specific age gaps. Predictive performance of proteomic clocks for mortality was comparable to that of classical lifestyle risk factors. In summary, proteomic clocks appear promising biomarkers of generalized age-related disease risk.
- Open Access
- 22 June 2026
- Published by Elsevier in Computers, Environment and Urban Systems
- Vol. 129
Assessment of individualised dynamic environmental exposures within The World Avatar
- 22 June 2026
- Published by Elsevier in Computers, Environment and Urban Systems
- Vol. 129
Open AccessAssessment of individualised dynamic environmental exposures within The World Avatar
- Open Access
- 30 April 2026
- Published by Springer Nature in Nature Medicine
- Vol. 32 (6) , 2117-2127
Data-driven prioritization of high-risk individuals for weight loss interventions
New obesity medications have demonstrated efficacy in trials, but their real-world deployment is partly limited by the absence of approaches that identify individuals for treatment based on risks for obesity-related complications. Here we present a risk prediction model to guide prioritization of high-risk individuals. In a population-based sample of ~200,000 individuals with a body mass index (BMI) exceeding 27 kg m − 2 , our machine learning framework identified the 20 most informative features, from among thousands tested, that predict future onset of 18 complications of obesity, providing information beyond BMI. An integrated model (OBSCORE) successfully stratified individuals into risk groups based on incidence over 10 years: for example, 5.7%, 1.8%, 0.9%, 0.4% and 0.1% for cardiovascular mortality. We demonstrate generalizability of the model in independent populations of European and non-European ancestry and, in SURMOUNT-1 trial participants, show that weight loss was similar across baseline OBSCORE risk groups and that predicted risks decreased following treatment with tirzepatide. In summary, OBSCORE provides a framework for prioritizing high-risk individuals with overweight or obesity based on their risk of obesity-related complications, complementing BMI-based frameworks.
- 30 April 2026
- Published by Springer Nature in Nature Medicine
- Vol. 32 (6) , 2117-2127
Open AccessData-driven prioritization of high-risk individuals for weight loss interventions
New obesity medications have demonstrated efficacy in trials, but their real-world deployment is partly limited by the absence of approaches that identify individuals for treatment based on risks for obesity-related complications. Here we present a risk prediction model to guide prioritization of high-risk individuals. In a population-based sample of ~200,000 individuals with a body mass index (BMI) exceeding 27 kg m − 2 , our machine learning framework identified the 20 most informative features, from among thousands tested, that predict future onset of 18 complications of obesity, providing information beyond BMI. An integrated model (OBSCORE) successfully stratified individuals into risk groups based on incidence over 10 years: for example, 5.7%, 1.8%, 0.9%, 0.4% and 0.1% for cardiovascular mortality. We demonstrate generalizability of the model in independent populations of European and non-European ancestry and, in SURMOUNT-1 trial participants, show that weight loss was similar across baseline OBSCORE risk groups and that predicted risks decreased following treatment with tirzepatide. In summary, OBSCORE provides a framework for prioritizing high-risk individuals with overweight or obesity based on their risk of obesity-related complications, complementing BMI-based frameworks.
- Open Access
- 16 April 2026
- Published by Springer Nature in Metabologia
- Vol. 2 (1)
Smoking, GAD65 autoimmunity, genetic susceptibility to type 1 diabetes and incident adult-onset diabetes in the EPIC-InterAct case–cohort study
Aims/hypothesis: Smoking increases the risk of type 2 diabetes, but its role in autoimmune diabetes remains unclear. We investigated whether smoking is associated with being GAD65 autoantibody (GAD65Ab) positive or the risk of progressing to diabetes. We also assessed its interaction with genetic susceptibility to type 1 diabetes. Methods: We used data from the EPIC-InterAct case–cohort study including 11,161 incident cases of adult-onset diabetes and 14,922 subcohort participants. Logistic regression was used to estimate odds ratios (ORs) for being GAD65Ab positive at baseline by smoking status. Hazard ratios (HRs) of diabetes by baseline smoking and GAD65Ab status were estimated using Prentice-weighted Cox regression. Two- and three-way interactions between smoking, GAD65Ab status and a type 1 diabetes genetic risk score (T1D-GRS) were assessed by attributable proportion due to interaction (AP). Results: There was no evidence that smoking was associated with being GAD65Ab positive (OR 1.00, 95% CI 0.82, 1.22 current vs never smoking), but among GAD65Ab-positive participants current smokers were more likely to progress to diabetes than never smokers (HR 1.40, 95% CI 1.03, 1.90) and there was an interaction between being GAD65Ab positive and heavy smoking for diabetes risk (AP 0.32, 95% CI 0.15, 0.49 ≥ 15 pack-years vs never smoking). Three-way interaction between heavy smoking, being GAD65Ab positive and the highest T1D-GRS tertile was also observed (combined HR 7.30, 95% CI 5.17, 10.31; AP 0.46, 95% CI 0.23, 0.70). Conclusions/interpretation: While smoking may not be associated with being GAD65Ab positive, it may increase the risk of progressing to diabetes in GAD65Ab-positive individuals and in those with genetic susceptibility to type 1 diabetes. Graphical
- 16 April 2026
- Published by Springer Nature in Metabologia
- Vol. 2 (1)
Open AccessSmoking, GAD65 autoimmunity, genetic susceptibility to type 1 diabetes and incident adult-onset diabetes in the EPIC-InterAct case–cohort study
Aims/hypothesis: Smoking increases the risk of type 2 diabetes, but its role in autoimmune diabetes remains unclear. We investigated whether smoking is associated with being GAD65 autoantibody (GAD65Ab) positive or the risk of progressing to diabetes. We also assessed its interaction with genetic susceptibility to type 1 diabetes. Methods: We used data from the EPIC-InterAct case–cohort study including 11,161 incident cases of adult-onset diabetes and 14,922 subcohort participants. Logistic regression was used to estimate odds ratios (ORs) for being GAD65Ab positive at baseline by smoking status. Hazard ratios (HRs) of diabetes by baseline smoking and GAD65Ab status were estimated using Prentice-weighted Cox regression. Two- and three-way interactions between smoking, GAD65Ab status and a type 1 diabetes genetic risk score (T1D-GRS) were assessed by attributable proportion due to interaction (AP). Results: There was no evidence that smoking was associated with being GAD65Ab positive (OR 1.00, 95% CI 0.82, 1.22 current vs never smoking), but among GAD65Ab-positive participants current smokers were more likely to progress to diabetes than never smokers (HR 1.40, 95% CI 1.03, 1.90) and there was an interaction between being GAD65Ab positive and heavy smoking for diabetes risk (AP 0.32, 95% CI 0.15, 0.49 ≥ 15 pack-years vs never smoking). Three-way interaction between heavy smoking, being GAD65Ab positive and the highest T1D-GRS tertile was also observed (combined HR 7.30, 95% CI 5.17, 10.31; AP 0.46, 95% CI 0.23, 0.70). Conclusions/interpretation: While smoking may not be associated with being GAD65Ab positive, it may increase the risk of progressing to diabetes in GAD65Ab-positive individuals and in those with genetic susceptibility to type 1 diabetes. Graphical
- Open Access
- 3 April 2026
- Published by Springer Nature in Research Square
Discovery of gene-alcohol interaction loci influencing blood pressure in 1.1 million individuals from multiple populations
Genetic predisposition and alcohol consumption are risk factors for increased blood pressure (BP), but their interactions influencing BP remain understudied. We conducted population-specific and cross-population meta-analyses of genome-wide gene-alcohol (GxAlc) interactions affecting BP in >1.1M individuals from multiple populations. We identified 46 GxAlc interaction loci for BP, including 21 from one-degree-of-freedom interaction tests (PGxAlc<5x10-8; or <0.05/Meff, Meff independent BP associations at P<10-5), and 25 from two-degree-of-freedom tests of main and interaction effects (PGxAlc<0.05/M2df, M2df independent 2df-associations at P2df<5x10-8), including 7 novel and 39 known BP loci. The 12q24 locus highlights the genetic effect of BRAP-rs11066001 on BP, being ~6 times larger in current drinkers than in non-drinkers. Gene prioritization with 46 GxAlc loci identified 15 genes with ≥3 lines of evidence (location, literature, druggability, functional/regulatory annotation, or pathway analyses). Several loci showed sex- and population-specific effects and revealed biological pathways of alcohol’s influence on BP, suggesting mechanisms underlying alcohol-induced hypertension.
- 3 April 2026
- Published by Springer Nature in Research Square
Open AccessDiscovery of gene-alcohol interaction loci influencing blood pressure in 1.1 million individuals from multiple populations
Genetic predisposition and alcohol consumption are risk factors for increased blood pressure (BP), but their interactions influencing BP remain understudied. We conducted population-specific and cross-population meta-analyses of genome-wide gene-alcohol (GxAlc) interactions affecting BP in >1.1M individuals from multiple populations. We identified 46 GxAlc interaction loci for BP, including 21 from one-degree-of-freedom interaction tests (PGxAlc<5x10-8; or <0.05/Meff, Meff independent BP associations at P<10-5), and 25 from two-degree-of-freedom tests of main and interaction effects (PGxAlc<0.05/M2df, M2df independent 2df-associations at P2df<5x10-8), including 7 novel and 39 known BP loci. The 12q24 locus highlights the genetic effect of BRAP-rs11066001 on BP, being ~6 times larger in current drinkers than in non-drinkers. Gene prioritization with 46 GxAlc loci identified 15 genes with ≥3 lines of evidence (location, literature, druggability, functional/regulatory annotation, or pathway analyses). Several loci showed sex- and population-specific effects and revealed biological pathways of alcohol’s influence on BP, suggesting mechanisms underlying alcohol-induced hypertension.
- Top Cited Papers
- 12 March 2026
- Published by Springer Nature in Osteoporosis International
- Vol. 37 (5) , 1-19
Predictive value of BMD for hip and other fractures: a meta-analysis to update FRAX
Summary The relationship between bone mineral density (BMD) at the femoral neck and fracture risk was determined in a meta-analysis of primary data of 307205 men and women from 53 cohort studies. Low BMD was an important predictor of fracture risk, particularly for hip fracture. Introduction This study aimed to quantify the relationship between DXA-measured femoral neck BMD and fracture risk and examine the effect of age, sex, time since measurement, and initial BMD value on fracture risk, with a view to updating FRAX®. Methods We studied 307,205 men and women from within 53 predominately population-based cohorts followed up for an average of 8.7 years and a total of 2,683,185 person-years. The association of BMD and fracture risk was examined using a Poisson model in each cohort separately by sex. Results were expressed as a gradient of risk (GR, hazard ratio/standard deviation decrease in BMD). The different studies were then merged using weighted coefficients. Results Most hip fractures arose in men and women with low bone mass or osteoporosis at baseline (73% and 92%, respectively) as was also the case for MOF (65% and 85%, respectively). BMD at the femoral neck strongly predicted hip fractures both in men and women with a similar gradient of risk and absolute risk at any given T-score. At the age of 65 years, the GR was 2.73 (95% CI = 2.29–3.26) in men and 2.61 (95% CI = 2.34–2.92) in women. However, the magnitude of association was significantly dependent on age, with a higher gradient of risk at younger ages. For example, at age 50 years, the GR was 3.49 (95% CI 2.51–4.84) in men and decreased to 2.14 (1.96–2.34) at age 80 years. The change in GR with age was slightly less marked in women (3.23 [2.75–3.80] and 2.11 [1.99–2.44], respectively). The GR for major osteoporotic fracture (MOF) remained unchanged with age (p = 0.12 for women and p = 0.89 for men). A significant decrease in GR for hip fracture was observed with increasing duration of follow-up, but the magnitude of the effect was modest compared with the effect of age. For other fracture outcomes, including non-hip major osteoporotic fracture, the gradient of risk was lower than for hip fracture. Conclusions Femoral neck BMD is a risk factor for fracture of substantial importance, particularly for future hip fracture. The lower magnitude of association at older age is consistent with other non-skeletal factors contributing to hip fracture risk with advancing age. Its validation on an international basis supports its use in case finding strategies. Its use should, however, take account of the variations in predictive value of BMD with age, sex, length of follow-up, and BMD.
- 12 March 2026
- Published by Springer Nature in Osteoporosis International
- Vol. 37 (5) , 1-19
Top Cited PapersPredictive value of BMD for hip and other fractures: a meta-analysis to update FRAX
Summary The relationship between bone mineral density (BMD) at the femoral neck and fracture risk was determined in a meta-analysis of primary data of 307205 men and women from 53 cohort studies. Low BMD was an important predictor of fracture risk, particularly for hip fracture. Introduction This study aimed to quantify the relationship between DXA-measured femoral neck BMD and fracture risk and examine the effect of age, sex, time since measurement, and initial BMD value on fracture risk, with a view to updating FRAX®. Methods We studied 307,205 men and women from within 53 predominately population-based cohorts followed up for an average of 8.7 years and a total of 2,683,185 person-years. The association of BMD and fracture risk was examined using a Poisson model in each cohort separately by sex. Results were expressed as a gradient of risk (GR, hazard ratio/standard deviation decrease in BMD). The different studies were then merged using weighted coefficients. Results Most hip fractures arose in men and women with low bone mass or osteoporosis at baseline (73% and 92%, respectively) as was also the case for MOF (65% and 85%, respectively). BMD at the femoral neck strongly predicted hip fractures both in men and women with a similar gradient of risk and absolute risk at any given T-score. At the age of 65 years, the GR was 2.73 (95% CI = 2.29–3.26) in men and 2.61 (95% CI = 2.34–2.92) in women. However, the magnitude of association was significantly dependent on age, with a higher gradient of risk at younger ages. For example, at age 50 years, the GR was 3.49 (95% CI 2.51–4.84) in men and decreased to 2.14 (1.96–2.34) at age 80 years. The change in GR with age was slightly less marked in women (3.23 [2.75–3.80] and 2.11 [1.99–2.44], respectively). The GR for major osteoporotic fracture (MOF) remained unchanged with age (p = 0.12 for women and p = 0.89 for men). A significant decrease in GR for hip fracture was observed with increasing duration of follow-up, but the magnitude of the effect was modest compared with the effect of age. For other fracture outcomes, including non-hip major osteoporotic fracture, the gradient of risk was lower than for hip fracture. Conclusions Femoral neck BMD is a risk factor for fracture of substantial importance, particularly for future hip fracture. The lower magnitude of association at older age is consistent with other non-skeletal factors contributing to hip fracture risk with advancing age. Its validation on an international basis supports its use in case finding strategies. Its use should, however, take account of the variations in predictive value of BMD with age, sex, length of follow-up, and BMD.