Establishing a regional bone density program: lessons from the Manitoba experience

Among people on osteoporosis medication, loss of appendicular or total body lean mass is an independent risk factor for hip and major osteoporotic fractures

People with prior lean mass loss had a ~ 10% higher risk of MOF and ~ 22-26% higher risk of hip fracture, and the results were similar in people on anti-osteoporosis medications. Loss of lean mass is associated with increased fracture risk. Patients should be encouraged to pursue strategies to prevent loss of lean mass.

BACKGROUND

Sarcopenia increases fracture risk. If the risk persists after starting osteoporosis medication, patients may need to be encouraged to pursue strategies to prevent loss of lean mass.

OBJECTIVE

To estimate the effects of loss in appendicular lean mass (ALM) or total body lean mass (TBLM) on subsequent fracture risk and effect modification with anti-osteoporosis medication use.

METHODS

We conducted a registry-based cohort study linked to population-based data. We identified individuals ≥ 40 years of age with two DXA assessments ≥ 1 year apart and minimum 0.5 years of observation. ALM and TBLM were estimated from weight, sex, and percent fat from DXA (R2 = 0.91 and 0.84 vs total body DXA, respectively). We report hazard ratios (HR) from Cox regression models estimating time to first incident major osteoporotic fracture (MOF) and hip fracture, adjusted for fracture risk; osteoporosis medication was included as an interaction term and used to stratify analyses.

RESULTS

We included 21,249 individuals (mean 67 [SD 10] years, 95% female, 37% on osteoporosis medication). The mean follow-up was 7 years (SD 4). A total of 1868 and 548 people had incident MOF and hip fracture, respectively. People with prior ALM loss (HR per SD 1.09, 95% CI 1.04-1.15) or TBLM loss (HR per SD 1.09, 95% CI 1.42-1.14) had a higher risk of MOF. Hip fracture risk was greater in people with prior ALM loss (HR per SD 1.22, 95% CI 1.12-1.33) and TBLM loss (HR per SD 1.26, 95% CI 1.16-1.38). There were no interactions with anti-osteoporosis medication use (all p > 0.3). When restricted to people on anti-osteoporosis medication, each SD in ALM or TBLM loss was associated with 8-9% increased MOF risk and 18-23% increased hip fracture risk.

CONCLUSIONS

Loss of lean mass is associated with increased fracture risk among individuals on anti-osteoporosis medication. Patients should be encouraged to pursue strategies to prevent sarcopenia.

Treatment reclassification in Canada from the Osteoporosis Canada 2023 clinical practice guidelines: the Manitoba BMD Registry

Osteoporosis Canada 2023 clinical practice guidelines increase the number of individuals recommended or suggested for anti-osteoporosis pharmacotherapy by refining treatment guidance for those who fell within the 2010 guidelines' moderate-risk category.

PURPOSE

In 2023, Osteoporosis Canada updated its 2010 clinical practice guidelines based upon consideration of fracture history, 10-year major osteoporotic fracture (MOF) risk, and BMD T-score in conjunction with age. The 2023 guidelines eliminated risk categories, including the moderate-risk group that did not provide clear treatment guidance. The current study was performed to appreciate the implications of the shift from 2010 risk categories to 2023 treatment guidance.

METHODS

The study population consisted of 79,654 individuals age ≥ 50 years undergoing baseline DXA testing from January 1996 to March 2018. Each individual was assigned to mutually exclusive categories based on 2010 and 2023 guideline recommendations. Treatment qualification, 10-year predicted and 10-year observed MOF risk were compared.

RESULTS

Treatment reclassification under the 2023 guidelines only affected 33.8% of individuals in the 2010 moderate-risk group, with 13.0% assigned to no treatment, 14.4% to suggest treatment, and 6.4% to recommend treatment. During the mean follow-up of 7.2 years, 6364 (8.0%) individuals experienced one or more incidents of MOF. The observed 10-year cumulative incidence of MOF in the study population was 10.5% versus the predicted 10.7% (observed to predicted mean calibration ratio 0.98, 95% CI 0.96-1.00). Individuals reclassified from 2010 moderate risk to 2023 recommend treatment were at greater MOF risk than those in the 2010 moderate-risk group assigned to 2023 suggest treatment or no treatment, but at lower risk than those in the 2010 high-risk group.

CONCLUSIONS

Osteoporosis Canada 2023 clinical practice guidelines affect individuals within the 2010 moderate-risk category, increasing the number for whom anti-osteoporosis pharmacotherapy is recommended or suggested. Increased treatment could reduce the population burden of osteoporotic fractures, though moderate-risk individuals now qualifying for treatment have a lower predicted and observed fracture risk than high-risk individuals recommended for treatment under the 2010 guidelines.

Bone Mineral Density Monitoring in Contiguous versus Non-Contiguous Lumbar Vertebrae: The Manitoba BMD Registry

INTRODUCTION

Only change in bone mineral density (BMD) on repeat DXA that exceeds the 95% least significant change (LSC) should be considered clinically meaningful. Frequently lumbar spine DXA must be reported after omitting vertebrae with localized structural artifact, which reduces measurement precision. Previous reports have raised concerns of higher least significant change (LSC) when spine BMD is based on non-contiguous rather than contiguous vertebrae. The current study was performed to compare lumbar spine LSC and BMD response to intervening anti-osteoporosis medication use from non-contiguous versus contiguous vertebrae.

METHODOLOGY

LSCs for lumbar spine DXA based on L1-L4 and all combinations of non-contiguous and contiguous vertebrae were calculated using 879 scan-pairs from the Manitoba BMD Program. We compared BMD change from these regions, overall and in relation to intervening anti-osteoporosis medication use, in 11,722 patients who had 2 DXA examinations.

RESULTS

LSCs were slightly greater when calculated from combinations of fewer than 4 vertebrae, but there was no meaningful difference between contiguous versus non-contiguous vertebrae. There were consistently high correlations between lumbar spine BMD change from L1-L4 and all combinations of continuous and non-contiguous vertebrae (all Pearson r≥ 0.9, p<0.001). Percentage changes in spine BMD and the fraction with treatment-concordant change exceeding the LSC were similar using contiguous or non-contiguous vertebrae.

CONCLUSIONS

Lumbar spine BMD change can be assessed from 2 or 3 non-contiguous vertebrae when clinically necessary, and precision in such cases is similar to using contiguous vertebrae. Non-contiguous vertebrae can detect treatment-concordant changes similar in spine BMD to contiguous vertebrae.

Effect of abdominal tissue thickness on trabecular bone score and fracture risk in adults with diabetes: the Manitoba BMD registry

Individuals with type 2 diabetes have lower trabecular bone score (TBS) and increased fracture risk despite higher bone mineral density. However, measures of trabecular microarchitecture from high-resolution peripheral computed tomography are not lower in type 2 diabetes. We hypothesized that confounding effects of abdominal tissue thickness may explain this discrepancy, since central obesity is a risk factor for diabetes and also artifactually lowers TBS. This hypothesis was tested in individuals aged 40 years and older from a large DXA registry, stratified by sex and diabetes status. When DXA-measured abdominal tissue thickness was not included as a covariate, men without diabetes had lower TBS than women without diabetes (mean difference -0.074, P < .001). TBS was lower in women with versus without diabetes (mean difference -0.037, P < .001), and men with versus without diabetes (mean difference -0.007, P = .042). When adjusted for tissue thickness these findings reversed, TBS became greater in men versus women without diabetes (mean difference +0.053, P < .001), in women with versus without diabetes (mean difference +0.008, P < .001), and in men with versus without diabetes (mean difference +0.014, P < .001). During mean 8.7 years observation, incident major osteoporotic fractures were seen in 7048 (9.6%). Adjusted for multiple covariates except tissue thickness, TBS predicted fracture in all subgroups with no significant diabetes interaction. When further adjusted for tissue thickness, HR per SD lower TBS remained significant and even increased slightly. In conclusion, TBS predicts fractures independent of other clinical risk factors in both women and men, with and without diabetes. Excess abdominal tissue thickness in men and individuals with type 2 diabetes may artifactually lower TBS using the current algorithm, which reverses after accounting for tissue thickness. This supports ongoing efforts to update the TBS algorithm to directly account for the effects of abdominal tissue thickness for improved fracture risk prediction.

Assessing Change in Spine Bone Density from Different Numbers and Combinations of Lumbar Vertebrae: The Manitoba BMD Registry

INTRODUCTION

Change in bone mineral density (BMD) is considered significant when it exceeds the 95 % least significant change (LSC) derived from that facility's precision study. The lumbar spine is often affected by structural artifact such that not all four lumbar vertebrae are evaluable. Guidelines suggest using a site-matched LSC when omitting vertebrae from the BMD measurement. The current study describes significant BMD change related to intervening anti-osteoporosis treatment for different numbers and combinations of lumbar vertebrae using site-matched LSC values.

METHODOLOGY

We identified 10,526 untreated adult women mean age 59.6 years with baseline and repeat spine BMD testing (mean interval 4.7 years) where all 4 lumbar vertebrae were evaluable. Change in spine BMD for different combinations of lumbar vertebrae was assessed in relation to intervening anti-resorptive treatment, contrasting women with high treatment exposure (medication possession ratio, MPR ≥ 0.8) versus women who remained untreated. Site-matched LSC values were derived from 879 test-retest precision measurements.

RESULTS

There was consistent linear trend between increasing MPR and BMD change exceeding the LSC for all lumbar vertebral combinations, positive with BMD increase and negative with BMD decrease (all p-trend <0.001). In the high treatment exposure group, mean percent increases in spine BMD were similar for all vertebral combinations, from L1-4 to a single vertebra. In untreated women, mean percent decreases in spine BMD were also similar for all vertebral combinations. The net treatment response (proportion of women with treatment-concordant changes minus proportion with treatment-discordant changes exceeding the LSC) was 29.7 % for 4 vertebrae, 27.5-30.0 % for 3 vertebrae, 22.4-28.5 % for 2 vertebrae, and 18.1-21.9 % for a single vertebra.

CONCLUSIONS

All numbers and combinations of lumbar vertebrae, when used in conjunction with site-matched LSC values, can provide clinically meaningful follow-up in treated and untreated patients, even when spine BMD is based on a single vertebral body.

Fracture Risk Associated with Different Numbers and Combinations of Lumbar Vertebrae: The Manitoba BMD Registry

Bone mineral density (BMD) is widely used for assessment of fracture risk. For the lumbar spine, BMD is typically measured from L1-L4 as it provides the largest area for assessment with the best measurement precision. Structural artifact often confounds spine BMD in clinical practice, and the International Society for Clinical Densitometry (ISCD) recommends removing vertebrae with artifact when reporting spine BMD. In its most recent position statements, the ISCD recommended against the use of a single vertebra when reporting spine BMD but stated that further studies should be done. The current analysis was performed to compare the performance of BMD from different numbers and combination of vertebral levels on fracture prediction in a large clinical registry of DXA tests for the Province of Manitoba, Canada. The study population comprised 39,727 individuals aged 40 years and older (mean age 62.7 years, 91.0 % female) with baseline DXA after excluding those with evidence of structural artifact. Mean follow-up for ascertaining fracture outcomes was 8.7 years. Area under the curve (AUC) for incident fracture risk stratification was statistically significant regardless of the BMD measurement site or fracture outcome. AUC differences with the various numbers and combinations of vertebral levels including a single vertebral body were small (less than or equal to 0.01). More substantial AUC differences were seen for femoral neck and total hip BMD versus L1-L4 BMD, approaching 0.1 for hip fracture stratification. In summary, we found that using combinations of fewer than 4 vertebrae including individual lumbar vertebrae predicted incident fractures. Importantly, differences between these different combinations were small when compared with L1-L4. Spine BMD was a better predictor of incident spine fracture compared to the hip, whereas the hip was better for hip fracture and overall fracture prediction.

Estimating Lumbar Spine Least Significant Change for Fewer than Four Vertebrae: The Manitoba BMD Registry

INTRODUCTION

The International Society of Clinical Densitometry recommends omitting lumbar vertebrae affected by structural artifact from spine BMD measurement. Since reporting fewer than 4 vertebrae reduces spine BMD precision, least significant change (LSC) needs to be adjusted upwards when reporting spine BMD change based on fewer than 4 vertebrae.

METHODOLOGY

In order to simplify estimating LSC from combinations of vertebrae other than L1-L4 (denoted LSCL1-4 ), we analyzed 879 DXA spine scan-pairs from the Manitoba BMD Program's ongoing precision evaluation. The additional impact on the LSC of performing the second scan on the same day vs different day was also assessed.

RESULTS

LSC progressively increased when fewer vertebrae were included, and also increased when the scans were performed on different days. We estimated that the LSCL1-4 should be adjusted upwards by 7 %, 24 % and 65 % to approximate the LSC for 3, 2, or 1 vertebral body, respectively. To additionally capture the greater LSC when the precision study was done on different days, LSCL1-4 derived from a precision study where scans were done on the same day should be adjusted upwards by 39 %, 60 % and 112 % for 3, 2, or 1 vertebral body, respectively.

CONCLUSION

LSCL1-4 derived from a precision study where scans are performed on the same day can be used to estimate LSC for fewer than 4 vertebrae and for scans performed on different days.

Effect of BMI-Discordant Abdominal Tissue Thickness on Fracture Probability: A Registry-Based Study

FRAX, which is used to assess fracture probability, considers body mass index (BMI), but BMI may not reflect individual variation in body composition and distribution. We examined the effect of BMI-discordant abdominal thickness on FRAX-derived fracture probability for major osteoporotic fracture (MOF) and hip fracture. We studied 73,105 individuals, mean age 64.2 years. During mean 8.7 years, 7048 (9.6%) individuals sustained incident MOF, including 2155 (3.0%) hip fractures. We defined abdominal thickness index (ATI) as the difference between abdominal thickness measured by dual-energy X-ray absorptiometry (DXA) and thickness predicted by BMI using sex-stratified regression. ATI was categorized from lower (<-2 cm, -2 to -1 cm) to higher (1-2 cm, >+2 cm) with referent around zero (-1 to +1 cm). Adjusted for FRAX probability, increasing ATI was associated with incident MOF and hip fracture (p < 0.001). For the highest ATI category, MOF risk was increased (hazard ratio [HR] = 1.23, 95% confidence interval [CI] 1.12-1.35) independent of FRAX probability. Similar findings were noted for hip fracture probability (HR = 1.28, 95% CI 1.09-1.51). There was significant age-interaction with much larger effects before age 65 years (HR = 1.44, 95% CI 1.23-1.69 for MOF; 2.29, 95% CI 1.65-3.18 for hip fracture). In contrast, for the subset of individuals with diabetes, there was also increased risk for those in the lowest ATI category (HR = 1.73, 95% CI 1.12-2.65 for MOF; 2.81, 95% CI 1.59-4.97 for hip fracture). Calibration plots across ATI categories demonstrated deviation from the line of identity in women (calibration slope 2.26 for MOF, 2.83 for hip fracture). An effect of ATI was not found in men, but this was inconclusive as the sex-interaction terms did not show significant effect modification. In conclusion, these data support the need to investigate increased abdominal thickness beyond that predicted by BMI and sex as a FRAX-independent risk factor for fracture. © 2023 The Authors. Journal of Bone and Mineral Research published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research (ASBMR).

Effect of Race/Ethnicity on United States FRAX Calculations and Treatment Qualification: A Registry-Based Study

Since 2008. the United States has had four race/ethnic fracture risk assessment tool (FRAX) calculators: White ("Caucasian"), Black, Asian, and Hispanic. The American Society for Bone and Mineral Research Task Force on Clinical Algorithms for Fracture Risk has been examining the implications of retaining race/ethnicity in the US FRAX calculators. To inform the Task Force, we computed FRAX scores according to each US calculator in 114,942 White, 485 Black, and 2816 Asian women (self-reported race/ethnicity) aged 50 years and older. We estimated treatment qualification based upon FRAX thresholds (3% for hip fracture, 20% for major osteoporotic fracture [MOF]). Finally, we examined measures for a hypothetical population-based FRAX calculator derived as the weighted mean for the US population based upon US Census Bureau statistics. With identical inputs, the highest FRAX measurements were found with the White FRAX calculator, lowest measurements with the Black calculator, and intermediate measurements for the Asian and Hispanic calculators. The percentage of women with FRAX scores exceeding the hip fracture treatment threshold was 32.0% for White, 1.9% for Black, and 19.7% for Asian women; the MOF treatment threshold was exceeded for 14.9% of White, 0.0% of Black, and 3.5% of Asian women. Disparities in treatment qualification were reduced after considering additional criteria (fracture history and dual-energy X-ray absorptiometry [DXA] T-score -2.5 or lower). When fracture risk was recalculated for non-White women using the White FRAX calculator, mean values for Asian women slightly exceeded those for White women but for Black women remained substantially below those for White women. When using a single population-based FRAX calculator, the mean probability of fracture and treatment qualification increased for non-White women across the age range. In summary, use of a single population-based FRAX calculator, rather than existing US race/ethnic FRAX calculators, will reduce differences in treatment qualification and may ultimately enhance equity and access to osteoporosis treatment. © 2023 The Authors. Journal of Bone and Mineral Research published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research (ASBMR).

Adjusting Trabecular Bone Score (TBS) for Level-Specific Differences Reduces FRAX®-Based Treatment Reclassification in Patients with Vertebral Exclusions: The Manitoba BMD Registry

Trabecular bone score (TBS) is a FRAX®-independent risk factor for fracture prediction. TBS values increase from cranial to caudal, with the following mean differences between TBSL1-L4 and individual lumbar vertebrae: L1 -0.093, L2 -0.008, L3 +0.055 and L4 +0.046. Excluding vertebral levels can affect FRAX-based treatment recommendations close to the intervention threshold. We examined the effect of adjusting for level-specific TBS differences in individuals with vertebral exclusions due to structural artifact on TBS-adjusted FRAX-based treatment recommendations. We identified 71,209 individuals aged ≥40 years with TBS and FRAX calculations through the Manitoba Bone Density Program. In the 24,428 individuals with vertebral exclusions, adjusting TBS using these level-specific factors agreed with TBSL1-L4 (mean difference -0.001). We compared FRAX-based treatment recommendations for TBSL1-L4 and for non-excluded vertebral levels before and after adjusting for level-specific TBS differences. Among those with baseline major osteoporotic fracture risk ≥15 %, TBS with vertebral exclusions reclassified FRAX-based treatment in 10.6 % of individuals compared with TBSL1-L4, and was reduced to 7.2 % after adjusting for level-specific differences. In 11,131 patients where L1-L2 was used for BMD reporting (the most common exclusion pattern with the largest TBS effect), treatment reclassification was reduced from 13.9 % to 2.4 %, respectively. Among individuals with baseline hip fracture risk ≥2 %, TBS vertebral exclusions reclassified 7.1 % compared with TBSL1-L4, but only 4.5 % after adjusting for level-specific differences. When L1-L2 was used for BMD reporting, treatment reclassification from hip fracture risk was reduced from 9.2 % to 5.2 %. In conclusion, TBS and TBS-adjusted FRAX-based treatment recommendations are affected by vertebral level exclusions for structural artifact. Adjusting for level-specific differences in TBS reduces reclassification in FRAX-based treatment recommendations.

FRAX® Adjustment Using Renormalized Trabecular Bone Score (TBS) from L1 Alone may be Optimal for Fracture Prediction: The Manitoba BMD Registry

Lumbar spine trabecular bone score (TBS) used in conjunction with FRAX® improves 10-year fracture prediction. The derived FRAX risk adjustment is based upon TBS measured from L1-L4, designated TBSL1-L4-FRAX. In prior studies, TBS measurements that include L1 and exclude L4 give better fracture stratification than L1-L4. We compared risk stratification from TBS-adjusted FRAX using TBS derived from different combinations of upper lumbar vertebral levels renormalized for level-specific differences in individuals from the Manitoba Bone Density Program aged >40 years with baseline assessment of TBS and FRAX. TBS measurements for L1-L3, L1-L2 and L1 alone were calculated after renormalization for level-specific differences. Corresponding TBS-adjusted FRAX scores designated TBSL1-L3-FRAX, TBSL1-L2-FRAX and TBSL1-FRAX were compared with TBSL1-L4-FRAX for fracture risk stratification. Incident major osteoporotic fractures (MOF) and hip fractures were assessed. The primary outcome was incremental change in area under the curve (ΔAUC). The study population included 71,209 individuals (mean age 64 years, 89.8% female). Before renormalization, mean TBS for L1-3, L1-L2 and L1 was significantly lower and TBS-adjusted FRAX significantly higher than from using TBSL1-L4. These differences were largely eliminated when TBS was renormalized for level-specific differences. During mean follow-up of 8.7 years 6745 individuals sustained incident MOF and 2039 sustained incident hip fractures. Compared with TBSL1-L4-FRAX, use of FRAX without TBS was associated with lower stratification (ΔAUC = -0.009, p < 0.001). There was progressive improvement in MOF stratification using TBSL1-L3-FRAX (ΔAUC = +0.001, p < 0.001), TBSL1-L2-FRAX (ΔAUC = +0.004, p < 0.001) and TBSL1-FRAX (ΔAUC = +0.005, p < 0.001). TBSL1-FRAX was significantly better than all other combinations for MOF prediction (p < 0.001). Incremental improvement in AUC for hip fracture prediction showed a similar but smaller trend. In conclusion, this single large cohort study found that TBS-adjusted FRAX performance for fracture prediction was improved when limited to the upper lumbar vertebral levels and was best using L1 alone.

Effects of Severe Lumbar Spine Structural Artifact on Trabecular Bone Score (TBS): The Manitoba BMD Registry

Trabecular bone score (TBS) is a bone mineral density (BMD)-independent risk factor for fracture. During DXA analysis and BMD reporting, it is standard practice to exclude lumbar vertebral levels affected by structural artifact. Although TBS is relatively insensitive to degenerative artifact, it is uncertain whether TBS is still useful in the presence extreme structural artifact that precludes reliable spine BMD measurement even after vertebral exclusions. Among individuals aged 40 years and older undergoing baseline DXA assessment from September 2012 to March 2018 we identified three mutually exclusive groups: spine BMD reporting performed without exclusions (Group 1, N=12,865), spine BMD reporting performed with vertebral exclusions (Group 2, N=4867), and spine BMD reporting not performed due to severe structural artifact (Group 3, N=1541). No significant TBS difference was seen for Group 2 versus Group 1 (referent), whereas TBS was significantly greater in Group 3 (+0.041 partially adjusted, +0.043 fully adjusted). When analyzed by the reason for vertebral exclusion, multilevel degenerative changes significantly increased TBS (+0.041 partially adjusted, +0.042 fully adjusted), while instrumentation significantly reduced TBS (-0.059 partially adjusted, -0.051 fully adjusted). Similar results were seen when analyses were restricted to those in Group 3 with a single reason for vertebral exclusions, and when follow up scans were also included. During mean follow-up of 2.5 years there were 802 (4.2 %) individuals with one or more incident fractures. L1-L4 TBS showed significant fracture risk stratification in all groups including Group 3 (P-interaction >0.4). In conclusion, lumbar spine TBS can be reliably measured in the majority of lumbar spine DXA scans, including those with artifact affecting up to two vertebral levels. However, TBS is significantly affected by the presence of extreme structural artifact in the lumbar spine, especially those with multilevel degenerative disc changes and/or instrumentation that precludes reliable BMD reporting.

A Population-Based Registry Study of Extended Bisphosphonate Use: Minimal Shift After Landmark Publications About Shorter Treatment Duration

Optimal duration of bisphosphonate therapy was unknown until the FLEX study was published in 2006 showing a 5-year course to be adequate for most women. In 2008, a link between long-term bisphosphonate and atypical femoral fractures was reported and confirmed in later studies. We hypothesized these landmark observations should have led to a decrease in use of bisphosphonates for >5 or 10 years, from 2010 onward. The Manitoba Bone Mineral Density (BMD) Registry with linkage to provincial pharmacy data was used to determine the percentage of long- and very long-term bisphosphonate users from therapy start. The cohort comprised women aged >50 years with BMD between 1995 and 2018 with oral bisphosphonate first prescribed for >90 days with adherence >75% in the first year. For each calendar year of continued therapy, the percentage of patients and medication possession rate was tabulated. The percentage of users beyond 5 years was compared among patients who started therapy in 1998-2004 (those taking 5 years of therapy still finish before 2010) versus 2005-2012 (all new therapy starts overlap 2010 in those taking ≥5 years of treatment). The cohort included 2991 women with mean follow-up 8.8 (1.3) years, 64.9% of whom took continuous oral bisphosphonate for >5 years and 41.9% for >10 years. In the earlier versus later era, there were 74.4% versus 70.2% who completed 5 years. With respect to longer treatment, there were 68.0% and 60.5% of patients treated for 6 or more years (p < 0.0001) and 46.6% versus 33.5% treated for >10 years (p = 0.08). Medication possession rate was >79% in every year of therapy. Landmark studies leading to more limited bisphosphonate courses may have slightly reduced longer-term treatment, but up to one-third of adherent patients in the modern era still receive continuous bisphosphonate therapy for >10 years. © 2023 The Authors. Journal of Bone and Mineral Research published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research (ASBMR).

Estimating the prevalence of osteoporosis using ranked-based methodologies and Manitoba's population-based BMD registry

Osteoporosis is a metabolic bone disorder that is characterized by reduced bone mineral density (BMD) and deterioration of bone microarchitecture. Osteoporosis is highly prevalent among women over 50, leading to skeletal fragility and risk of fracture. Early diagnosis and treatment of those at high risk for fracture is very important in order to avoid morbidity, mortality and economic burden from preventable fractures. The province of Manitoba established a BMD testing program in 1997. The Manitoba BMD registry is now the largest population-based BMD registry in the world, and has detailed information on fracture outcomes and other covariates for over 160,000 BMD assessments. In this paper, we develop a number of methodologies based on ranked-set type sampling designs to estimate the prevalence of osteoporosis among women of age 50 and older in the province of Manitoba. We use a parametric approach based on finite mixture models, as well as the usual approaches using simple random and stratified sampling designs. Results are obtained under perfect and imperfect ranking scenarios while the sampling and ranking costs are incorporated into the study. We observe that rank-based methodologies can be used as cost-efficient methods to monitor the prevalence of osteoporosis.

Trabecular Bone Score Vertebral Exclusions Affect Risk Classification and Treatment Recommendations: The Manitoba Bmd Registry

Lumbar spine trabecular bone score (TBS), a texture measure derived from spine dual-energy x-ray absorptiometry (DXA) images, is a bone mineral density (BMD)-independent risk factor for fracture. Lumbar vertebral levels that show structural artifact are excluded from BMD measurement. TBS is relatively unaffected by degenerative artifact, and it is uncertain whether the same exclusions should be applied to TBS reporting. To gain insight into the clinical impact of vertebral exclusion on TBS, we examined the effect of lumbar vertebral exclusions in routine clinical practice on tertile-based TBS categorization and TBS adjusted FRAX-based treatment recommendations. The study population consisted of 71,209 individuals aged 40 years and older with narrow fan-beam spine DXA examinations and retrospectively-derived TBS. During BMD reporting, 34.3% of the scans had one or more vertebral exclusions for structural artifact. When TBS was derived from the same vertebral levels used for BMD reporting, using fixed L1-L4 tertile cutoffs (1.23 and 1.31 from the McCloskey meta-analysis) reclassified 17.9% to a lower and 6.5% to a higher TBS category, with 75.6% unchanged. Reclassification was reduced from 24.4% overall to 17.2% when level-specific tertile cutoffs from the software manufacturer were used. Treatment reclassification based upon FRAX major osteoporotic fracture probability occurred in 2.9% overall, but in 9.6% of those with baseline risk ≥15%. For treatment based upon FRAX hip fracture probability, reclassification occurred in 3.4% overall, but in 10.4% in those with baseline risk ≥2%. In summary, lumbar spine TBS measurements based upon vertebral levels other than L1-L4 can alter the tertile category and treatment recommendations based upon TBS-adjusted FRAX calculation, especially for those close to or exceeding the treatment cut-off. Manufacturer level-specific tertile cut-offs should be used if vertebral exclusions are applied.

FRAX Adjustment by Trabecular Bone Score with or Without Bone Mineral Density: The Manitoba BMD Registry

Trabecular bone score (TBS), a texture measure derived from spine dual-energy x-ray absorptiometry (DXA) images, is a FRAX®-independent risk factor for fracture. The TBS adjustment to FRAX assumes the presence of femoral neck BMD in the calculation. However, there are many individuals in whom hip DXA cannot be acquired. Whether the TBS-adjustment would apply to FRAX probabilities calculated without BMD has not been studied. The current analysis was performed to evaluate major osteoporotic fracture (MOF) and hip fracture risk adjusted for FRAX with and without femoral neck BMD. The study cohort consisted of 71,209 individuals (89.8% female, mean age 64.0 years). During mean follow-up 8.7 years, 6743 (9.5%) individuals sustained one or more incident MOF, of which 2037 (2.9%) sustained a hip fracture. Lower TBS was significantly associated with increased fracture risk when adjusted for FRAX probabilities, with a slightly larger effect when BMD was not included. Inclusion of TBS in the risk calculation gave a small but significant increase in stratification for fracture probabilities estimated with and without BMD. Calibration plots showed very minor deviations from the line of identity, indicating overall good calibration. In conclusion, the existing equations for incorporating TBS in FRAX estimates of fracture probability work similarly when femoral neck BMD is not used in the calculation. This potentially extends the range of situations where TBS can be used clinically to those individuals in whom lumbar spine TBS is available but femoral neck BMD is not available.

Contributions of Clinical and Technical Factors to Longitudinal Change in Trabecular Bone Score and Bone Density: A Registry-Based Individual-Level Analysis

Lumbar spine trabecular bone score (TBS), a gray-level texture measure derived from spine dual-energy X-ray absorptiometry (DXA) images, is a bone mineral density (BMD)-independent risk factor for fracture. An unresolved question is whether TBS is sufficiently responsive to change over time or in response to widely used osteoporosis therapy at the individual level to serve as a useful biomarker. Using the Manitoba DXA Registry, we identified 11,643 individuals age 40 years and older with two fan-beam DXA scans performed on the same instrument within 5 years (mean interval 3.2 years), of whom 6985 (60.0%) received antiresorptive osteoporosis medication (majority oral bisphosphonate) between the scans. We examined factors that were associated with a change in lumbar spine TBS, lumbar spine BMD, and total hip BMD exceeding the 95% least significant change (LSC). Change exceeding the LSC was identified in 23.0% (9.3% increase, 13.8% decrease) of lumbar spine TBS, 38.2% (22.1% increase, 16.1% decrease) lumbar spine BMD, and 42.5% (17.6% increase, 24.9% decrease) total hip BMD measurement pairs. From regression models, the variables most strongly associated with significant change in TBS (decreasing order) were tissue thickness change, acquisition mode change, weight change, and spine percent fat change. Consistent with the insensitivity of TBS to oral antiresorptive therapies, use of these agents showed very little effect on TBS change. In contrast, for both spine BMD change and total hip BMD change, osteoporosis medication use was the most significant variable, whereas tissue thickness change, acquisition mode change, and weight change had relatively weak effects. In summary, change in spine TBS using the present algorithm appears to be strongly affected by technical factors. This suggests a limited role, if any, for using TBS change in untreated individuals or for monitoring response to antiresorptive treatment in routine clinical practice with the current version of the TBS algorithm. © 2023 The Authors. Journal of Bone and Mineral Research published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research (ASBMR).

Vertebral Level Variations in Trabecular Bone Score and Effect on Fracture Prediction: The Manitoba BMD Registry

Trabecular bone score (TBS), a texture measure derived from spine dual-energy x-ray absorptiometry (DXA) images, is a bone mineral density (BMD)-independent risk factor for fracture. TBS is reportedly insensitive to degenerative changes, and it is uncertain whether the same rules for excluding lumbar vertebral levels from BMD measurement should be applied to TBS. The current analysis was performed to explore inter-vertebral variation in TBS measurements from L1 to L4, how this relates to clinically identified structural artifact resulting in vertebral level exclusion from BMD reporting, and area under the curve (AUC) for incident fracture. The study population comprised 70,762 individuals aged 40 years and older at the time of baseline spine DXA assessment (mean age 64.1 years, 89.7% female), among whom 24,289 (34.3%) had one or more vertebral exclusions. Both TBS and BMD showed a similar cranial/caudal inter-vertebral gradient. Compared with L1-4, TBS from L1 alone was lower (mean difference -0.096; -7.6%) while TBS from L4 alone was 0.046 (3.6%) greater, similar in those without and with visual structural artifact. During mean follow-up of 8.7 years, 6744 (9.5%) individuals sustained incident major osteoporotic fractures. TBS from L1 alone gave significantly higher AUC for incident fracture than L1-4, which was in turn significantly higher than L2, L3 and L4 alone, seen in those without and with visual structural artifact. In contrast, AUCs for BMD showed minimal variation from L1 to L4, and was higher for L1-4 than for any individual lumbar vertebral level. In summary, we found inter-vertebral TBS variations within the lumbar spine are overall similar to BMD but are relatively unaffected by visual structural artifact. Fracture outcomes showed the strongest association with TBS measured from L1 alone. Further investigation is need to understand the cause and clinical application of these differences.

Bone density and trabecular bone score to predict fractures in adults aged 20-39 years: a registry-based study

Trabecular bone score (TBS) enhances fracture risk assessment in older adults; whether this is true in younger people is uncertain. In this registry-based study of adults aged 20-39 years, low BMD, but not low TBS, predicted fracture.

PURPOSE

Trabecular bone score (TBS), a bone texture measurement, is associated with fracture risk independent of bone mineral density (BMD) in older adults. In adults aged 20-40 years, TBS remains stable and its role in fracture risk assessment is unclear. We utilized the Manitoba Bone Density Registry to explore the relationship of fracture risk with BMD and TBS in younger adults.

METHODS

Women and men aged 20-39 years referred for DXA testing were studied. Incident major and any fractures were captured from health records. Categories based on WHO BMD T-score classification and TBS tertile were considered using Cox regression models to estimate covariate-adjusted (including sex) hazard ratios (aHR, 95%CI) for incident fracture by category, and each SD decrement in BMD and TBS.

RESULTS

The study included 2799 individuals (77% female, mean age 32 years). Mean (SD) minimum T-score was - 0.9 (1.1) and TBS 1.355 (0.114); 7% had osteoporosis and 13% were in the lowest TBS tertile. Incident major osteoporotic fracture (MOF) and any fracture risk was elevated in those with osteopenia (aHRs 1.20/1.45) and osteoporosis (aHRs 4.60/5.16). Fracture risk was unrelated to TBS tertile. Each SD decrement in BMD was associated with increased MOF risk (aHR 1.64) and any fracture (aHR 1.71); lower TBS was unrelated to fractures.

CONCLUSION

In young adults, low BMD, but not low TBS, was predictive of MOF and any fracture. Routine clinical TBS measurement is not recommended for young adults. Further study is indicated to evaluate whether TBS is beneficial in subsets of younger adults.

General Comorbidity Indicators Contribute to Fracture Risk Independent of FRAX: Registry-Based Cohort Study

CONTEXT

FRAX® estimates 10-year fracture probability from osteoporosis-specific risk factors. Medical comorbidity indicators are associated with fracture risk but whether these are independent from those in FRAX is uncertain.

OBJECTIVE

We hypothesized Johns Hopkins Aggregated Diagnosis Groups (ADG®) score or recent hospitalization number may be independently associated with increased risk for fractures.

METHODS

This retrospective cohort study included women and men age ≥ 40 in the Manitoba BMD Registry (1996-2016) with at least 3 years prior health care data and used linked administrative databases to construct ADG scores along with number of hospitalizations for each individual. Incident Major Osteoporotic Fracture and Hip Fracture was ascertained during average follow-up of 9 years; Cox regression analysis determined the association between increasing ADG score or number of hospitalizations and fractures.

RESULTS

Separately, hospitalizations and ADG score independently increased the hazard ratio for fracture at all levels of comorbidity (hazard range 1.2-1.8, all P < 0.05), irrespective of adjustment for FRAX, BMD, and competing mortality. Taken together, there was still a higher than predicted rate of fracture at all levels of increased comorbidity, independent of FRAX and BMD but attenuated by competing mortality. Using an intervention threshold of major fracture risk >20%, application of the comorbidity hazard ratio multiplier to the patient population FRAX scores would increase the number of treatment candidates from 8.6% to 14.4%.

CONCLUSION

Both complex and simple measures of medical comorbidity may be used to modify FRAX-based risk estimates to capture the increased fracture risk associated with multiple comorbid conditions in older patients.

Ethnicity and Fracture Risk Stratification from Trabecular Bone Score in Canadian Women: The Manitoba BMD Registry

Lumbar spine Trabecular Bone Score (TBS), a grey-level texture measure derived from spine dual-energy x-ray absorptiometry (DXA) images, is a bone mineral density (BMD)-independent risk factor for fracture. An unresolved and controversial question is whether there are ethnic differences that affect the utility of TBS for fracture risk assessment. The current analysis examined whether self-identified ethnicity (White, Asian, Black) in women age 40 years and older referred for DXA testing affected fracture risk stratification from TBS using a large clinical registry. The study population comprised 63,078 White women, 1,915 Asian women and 329 Black women (n=329) with mean follow up 9.0±5.2 years. There were between group differences in BMI (Black>White>Asian), lumbar spine fat percentage (Asian>White>Black) and lumbar spine tissue thickness (Black>White>Asian). Despite this, lumbar spine TBS was not significantly different between the subgroups, though there was a significant difference in lumbar spine and total hip BMD (Black >White>Asian). TBS provided significant stratification for MOF and any fracture for all ethnicity subgroups, and for hip fracture in White and Asian subgroups (insufficient numbers for analysis in Black women). No significant difference in White vs. Asian or White vs. Black women were identified using a Bonferroni adjusted p-value. In summary, we found that lumbar spine TBS measurements were similar among White, Asian and Black women referred for DXA assessment in Manitoba, Canada. TBS and BMD measurements significantly stratified fracture risk in all three populations without a meaningful difference between groups. This suggests that TBS does not need to be used differently in White vs. non-White populations.

Trabecular Bone Score (TBS) Cross-Calibration for GE Prodigy and IDXA Scanners

Dual-energy X-ray absorptiometry (DXA) is used for osteoporosis diagnosis, fracture prediction and to monitor changes in bone mineral density (BMD). Change in DXA instrumentation requires formal cross-calibration and procedures have been described by the International Society for Clinical Densitometry. Whether procedures used for BMD cross-calibration are sufficient to ensure lumbar spine trabecular bone score (TBS) cross-calibration is currently uncertain. The Manitoba Bone Density Program underwent a program-wide upgrade in DXA instrumentation from GE Prodigy to iDXA in 2012, and a representative a sample of 108 clinic patients were scanned on both instruments. Lumbar spine TBS (L1-L4) measurements were retrospectively derived in 2013. TBS calibration phantoms were not available at our site when this was performed. We found excellent agreement for lumbar spine BMD, without deviation from the line of perfect agreement, and low random error (standard error of the estimate [SEE] 2.54% of the mean). In contrast, spine TBS (L1-L4) showed significant deviation from the line of identity: TBS(iDXA) = 0.730 x TBS(Prodigy) + 0.372 (p<0.001 for slope and intercept); SEE 5.12% of the mean with negative bias (r=-0.550). Results were worse for scans acquired in thick versus standard mode, but similar when the population was stratified as BMI < or > 35 kg/m². In summary, it cannot be assumed that just because BMD cross-calibration is good that this applies to TBS. This supports the need for using TBS phantom calibration to accommodate between-scanner differences as part of the manufacturer's TBS software installation.

Fracture Prediction from Trabecular Bone Score is Unaffected by Anti-Resorptive Treatment: A Registry-Based Cohort Study

Trabecular bone score (TBS) predicts osteoporotic fractures independent of bone mineral density (BMD) and clinical risk factors. The aim of this study was to explore whether anti-resorptive treatment affects fracture risk prediction from TBS using a large clinical registry that includes all dual-energy X-ray absorptiometry (DXA) tests for the Province of Manitoba, Canada. Cohort 1 included 53,863 individuals aged ≥ 40 years (11.4% men; mean age 64.1 years) who had not received any anti-resorptive therapy in the year prior the baseline DXA. Cohort 2 comprised 22,917 individuals aged ≥ 40 years (6% men, mean age 66.7 years) undergoing a second DXA visit. Anti-resorptive medication was initiated in the first year after DXA in 13,439 (25%) individuals from Cohort 1 (87.9% bisphosphonates); among Cohort 2 8,864 (38.7%) had received anti-resorptive medication in the year before DXA (77.8% bisphosphonates). Incident major osteoporotic fracture (MOF), hip fracture and any fracture were identified over mean follow up 8.6 and 7.0 years for Cohorts 1 and 2, respectively. Area under the curve showed significant risk stratification for all fracture types and treatment levels, whether treatment was initiated after TBS measurement (Cohort 1) or prior to TBS measurement (Cohort 2). In Cox regression models, without and with covariate adjustment, fracture prediction from TBS was unaffected by anti-resorptive medication use (p-interaction >0.5 for all analyses). In conclusion, TBS was a robust predictor of fracture in models adjusted for clinical risk factors and BMD. The use of anti-resorptive therapy, either in the year before or following TBS measurement, did not attenuate fracture risk prediction by TBS compared to untreated individuals.

Race/ethnic differences in the prevalence of osteoporosis, falls and fractures: a cross-sectional analysis of the Canadian Longitudinal Study on Aging

Most of the published epidemiology on osteoporosis is derived from White populations; still many countries have increasing ethno-culturally diverse populations, leading to gaps in the development of population-specific effective fracture prevention strategies. We describe differences in prevalent fracture and bone mineral density patterns in Canadians of different racial/ethnic backgrounds.

INTRODUCTION

We described prevalent fracture and bone mineral density (BMD) patterns in Canadians by their racial/ethnic backgrounds.

METHODS

For this cross-sectional analysis, we used the Canadian Longitudinal Study on Aging baseline data (2011-2015) of 22,091 randomly selected participants of Black, East Asian, South or Southeast Asian (SSEA) and White race/ethnic backgrounds, aged 45-85 years with available information on the presence or absence of self-reported prevalent low trauma fractures and femoral neck BMD (FNBMD) measurement. Logistic and linear regression models examined associations of race/ethnic background with fracture and FNBMD, respectively. Covariates included sex, age, height, body mass index (BMI), grip strength and physical performance score.

RESULTS

We identified 11,166 women and 10,925 men. Self-reported race/ethnic backgrounds were: 139 Black, 205 East Asian, 269 SSEA and 21,478 White. White participants were older (mean 62.5 years) than the other groups (60.5 years) and had a higher BMI (28.0 kg/m²) than both Asian groups, but lower than the Black group. The population-weighted prevalence of falls was 10.0%, and that of low trauma fracture was 12.0% ranging from 3.3% (Black) to 12.3% (White), with Black and SSEA Canadians having lower adjusted odds ratios (aOR) of low trauma fractures than White Canadians (Black, aOR = 0.3 [95% confidence interval: 0.1-0.7]; SSEA, aOR = 0.5 [0.3-0.8]). The mean (SD) FNBMD varied between groups: Black, 0.907 g/cm² (0.154); East Asian, 0.748 g/cm² (0.119); SSEA, 0.769 g/cm² (0.134); and White, 0.773 g/cm² (0.128). Adjusted linear regressions suggested that Black and both Asian groups had higher FNBMD compared to White.

CONCLUSION

Our results support the importance of characterizing bone health predictors in Canadians of different race/ethnic backgrounds to tailor the development of population-specific fracture prevention strategies.

Effect of Discordant Hip Bone Density on Incident Fracture Risk: A Registry-Based Cohort Study

The Fracture Risk Assessment Tool (FRAX®) combines clinical risk factors and optionally femoral neck bone density to estimate major osteoporotic fracture (MOF) and hip fracture probability. Hip dual-energy X-ray absorptiometry (DXA) simultaneously measures the trochanter and total hip, but these regions are not considered by FRAX. Our aim was to determine whether discordance in trochanter and total hip bone density (defined as ≥1 T-score difference from the femoral neck) affects fracture risk adjusted for fracture probability. Using the Manitoba bone density registry, we identified 84,773 women and men age 40 years or older undergoing baseline hip DXA. The outcomes were incident MOF and hip fracture. Cox regression hazard ratios (HRs) with 95% confidence intervals (CIs) adjusted for baseline fracture probability were used to test the association between hip T-score discordance and incident fractures. Hip T-score discordance affected more than one in five subjects (trochanter lower in 3.9%, higher in 14.2%; total hip lower in 0.3%, higher in 14.9%). After mean 8.8 years there were 8444 incident MOF including 2664 hip fractures. Discordantly lower trochanter and lower total hip T-score (≥1 below femoral neck) was associated with increased risk for MOF (adjusted HRs 1.47 and 1.60) and hip fracture (HRs 1.85 and 2.12), while discordantly higher trochanter and total hip T-score (≥1 above femoral neck) was associated with lower risk for MOF (HRs 0.83 and 0.71) and hip fracture (HRs 0.79 and 0.68). In models that examined the trochanter and total hip simultaneously, discordantly lower trochanter T-score was associated with increased incident MOF and hip fracture risk (HRs 1.43 and 1.79) whereas discordantly higher total hip T-score was associated with lower risk (HRs 0.73 and 0.75). In conclusion, trochanter and total hip regions frequently show T-scores that are discordant with the femoral neck. This information strongly affects incident fracture risk independent of fracture probability scores computed with femoral neck bone density. © 2022 American Society for Bone and Mineral Research (ASBMR).

FREM predicts 10-year incident fracture risk independent of FRAX® probability: a registry-based cohort study

The Danish Fracture Risk Evaluation Model (FREM) was found to predict fracture risk independent of 10-year fracture probability derived with the FRAX® tool including bone mineral density from DXA.

INTRODUCTION

FREM was developed from Danish public health registers without DXA information to identify high imminent risk of major osteoporotic fracture (MOF) and hip fracture (HF), while FRAX® estimates 10-year fracture probability from clinical risk factors and femoral neck bone mineral density (BMD) from DXA. The FREM algorithm showed significant 1- and 2-year fracture risk stratification when applied to a clinical population from Manitoba, Canada. We examined whether FREM predicts 10-year fracture risk independent of 10-year FRAX probability computed with BMD.

METHODS

Using the Manitoba BMD Program registry, we identified women and men aged ≥ 45 years undergoing baseline BMD assessment. We calculated FREM and FRAX scores, and identified incident fractures over 10 years. Hazard ratios (HRs) for incident fracture were estimated according to FREM quintile, adjusted for FRAX probability. We compared predicted with observed 10-year cumulative fracture probability estimated with competing mortality.

RESULTS

The study population comprised 74,446 women, mean age 65.2 years; 7945 men, mean age 67.5 years. There were 7957 and 646 incident MOF and 2554 and 294 incident HF in women and men, respectively. Higher FREM scores were associated with increased risk for MOF (highest vs middle quintile HRs 1.49 women, 2.06 men) and HF (highest vs middle quintile HRs 2.15 women, 2.20 men) even when adjusted for FRAX. Greater mortality with higher FREM scores attenuated its effect on 10-year fracture probability. In the highest FREM quintile, observed slightly exceeded predicted 10-year probability for MOF (ratios 1.05 in women, 1.49 in men) and HF (ratios 1.29 in women, 1.34 in men).

CONCLUSIONS

Higher FREM scores identified women and men at increased fracture risk even when adjusted for FRAX probability that included BMD; hence, FREM provides additional predictive information to FRAX. FRAX slightly underestimated 10-year fracture probability in those falling within the highest FREM quintile.

Apparent "Rapid Loss" After Short-Interval Bone Density Testing in Menopausal Women Is Usually a Measurement Artifact

CONTEXT

Medication may be considered when bone mineral density (BMD) loss is reported as "excessive."

OBJECTIVE

We hypothesized that the rate of BMD change between 2 serial tests demonstrates higher random variability at shorter vs longer intervals, misclassifying some women as "rapid losers."

METHODS

This retrospective observational cohort study in Manitoba, Canada included women aged > 55 years without osteoporosis medications or glucocorticoids. Using paired baseline (1998-2016) and repeat (2001-2018) BMD measurements, we estimated the distribution of annualized change (first to second BMD) at spine, hip, and femoral neck stratified by testing interval (2-2.9, 3-3.9,...9-9.9, ≥ 10.0 years). "Rapid annual bone loss" was defined as exceeding the 95th percentile for decreases from all measurement pairs. Odds ratios (OR) for rapid loss were estimated using regression models adjusted for age and clinical covariates.

RESULTS

From 7126 paired BMD measurements, mean annualized change was constant yet standard deviations in BMD change were > 2-fold greater with intervals of 2 to 2.9 years vs ≥ 10 years(P < 0.001). "Rapid annual loss" was seen in ~10% of short-interval tests vs < 1% of long-interval tests. ORs for "rapid loss" progressively declined with increasing testing interval (spine 15.3 [4.8-48.9], total hip 9.3 [4.4-19.5], femoral neck 18.7 [6.8-51.3] for a 2- to 2.9-year testing interval; referent ≥ 10 years).

CONCLUSION

There is a wider apparent range in annualized BMD loss with short-interval testing which greatly attenuates over longer intervals. BMD reports of "rapid loss" across shorter testing intervals likely reflect an artifact of BMD measurement error and should not be used as an indication for antifracture medication initiation.

The Effect of Fracture Recency on Observed 10-Year Fracture Probability: A Registry-Based Cohort Study

FRAX estimates 10-year fracture major osteoporotic fracture (MOF) and hip fracture probability from multiple risk factors. FRAX does not consider prior fracture site or time since fracture. Fracture risk is greater in the initial 2-year post-fracture period (imminent risk), implying that FRAX may underestimate risk in this setting. We used the population-based Manitoba Bone Mineral Density (BMD) Program registry to examine the effect of fracture recency and site on incident fracture risk predictions using FRAX. We identified women aged 40 years or older with baseline BMD and FRAX scores. Observed fracture outcomes to 10 years were compared with predicted 10-year fracture probability stratified by prior fracture status: none, recent (<2 years [median 0.3 years]), and remote (≥2 years [median 10.6 years]). For women with recent fractures, we also examined proposed multipliers to adjust FRAX for the effect of fracture recency and site. The cohort comprised 33,465 women aged 40 to 64 years (1897 recent fracture, 2120 remote fracture) and 33,806 women aged ≥65 years (2365 fracture, 4135 remote fracture). Observed fracture probability was consistent with predicted probability in most analyses. In women aged 40 to 64 years, there was a significant effect of recent vertebral and humerus fracture on MOF (observed to predicted 1.61 and 1.48, respectively), but these effects were still lower than the proposed multipliers (2.32 and 1.67, respectively). No significant effect of fracture recency was found after hip or forearm fracture in either age group. Our findings contribute to accumulating evidence of the importance of recent fracture. The effect of fracture recency was not consistent across fracture sites and with a lower magnitude than previously reported. Further quantification of effect size and specificity in additional independent cohorts is warranted to validate and refine recent-fracture multipliers in fracture risk assessment. © 2022 American Society for Bone and Mineral Research (ASBMR).

Divergent Patterns of Antifracture Medication Use Following Fracture on Therapy: A Population-Based Cohort Study

CONTEXT

Fracture on therapy should motivate better antifracture medication adherence.

OBJECTIVE

This study aimed to describe osteoporosis medication adherence in women before and following a fracture.

METHODS

This retrospective cohort analysis of antifracture medication possession ratios (MPR) among women in the Manitoba BMD Registry (1996-2013) included menopausal women who started antifracture drug therapy after a dual-energy x-ray absorptiometry (DXA)-BMD assessment with follow-up for 5 years during which a nontraumatic fracture occurred at least 1 year after starting treatment. Linked prescription records determined medication adherence (estimated by MPR) in 1-year intervals. The variable of interest was MPR in the year before and after the year in which the fracture occurred, with subgroup analyses according to duration of treatment pre-fracture. We chose an MPR of ≥ 0.50 to indicate minimum adherence needed for drug efficacy.

RESULTS

There were 585 women with fracture on therapy, 193 (33%) had hip or vertebral fracture. Bisphosphonates accounted for 82.2% of therapies. Median MPR the year prior to fracture was 0.89 (IQR, 0.49-1.0) and 0.69 (IQR, 0.07-0.96) the year following the year of fracture (P < 0.0001). The percentage of women with MPR ≥ 0.5 pre-fracture was 73.8%, dropping to 57.3% post-fracture (P < 0.0001); when restricted to hip/vertebral fracture, results were similar (58.2% to 33.3%; P < 0.002). Among those with pre-fracture MPR < 0.5, only 21.7% achieved a post-fracture MPR ≥ 0.5.

CONCLUSIONS

Although fracture on therapy may motivate sustained/improved adherence, MPR remains low or even declines after fracture in many. This could reflect natural decline in MPR with time but is paradoxical to expectations. Fracture on therapy represents an important opportunity for clinicians to reemphasize treatment adherence.

Prediction of imminent fracture risk in Canadian women and men aged 45 years or older: external validation of the Fracture Risk Evaluation Model (FREM)

The Fracture Risk Evaluation Model (FREM) identifies individuals at high imminent risk of major osteoporotic fractures. We validated FREM on 74,828 individuals from Manitoba, Canada, and found significant fracture risk stratification for all FREM scores. FREM performed better than age alone but not as well as FRAX® with BMD.

INTRODUCTION

The FREM is a tool developed from Danish public health registers (hospital diagnoses) to identify individuals over age 45 years at high imminent risk of major osteoporotic fractures (MOF) and hip fracture (HF). In this study, our aim was to examine the ability of FREM to identify individuals at high imminent fracture risk in women and men from Manitoba, Canada.

METHODS

We used the population-based Manitoba Bone Mineral Density (BMD) Program registry, and identified women and men aged 45 years or older undergoing baseline BMD assessment with 2 years of follow-up data. From linked population-based data sources, we constructed FREM scores using up to 10 years of prior healthcare information.

RESULTS

The study population comprised 74,828 subjects, and during the 2 years of observation, 1612 incident MOF and 299 incident HF occurred. We found significant fracture risk stratification for all FREM scores, with AUC estimates of 0.63-0.66 for MOF for both sexes and 0.84 for women and 0.65-0.67 for men for HF. FREM performed better than age alone but not as well as FRAX® with BMD. The inclusion of physician claims data gave slightly better performance than hospitalization data alone. Overall calibration for 1-year MOF prediction was reasonable, but HF prediction was overestimated.

CONCLUSION

In conclusion, the FREM algorithm shows significant fracture risk stratification when applied to an independent clinical population from Manitoba, Canada. Overall calibration for MOF prediction was good, but hip fracture risk was systematically overestimated indicating the need for recalibration.

Diminishing Value from Multiple Serial Bone Densitometry in Women Receiving Antiresorptive Medication for Osteoporosis

CONTEXT

The value of serial bone mineral density (BMD) monitoring while on osteoporosis therapy is controversial.

OBJECTIVE

We determined the percentage of women classified as suboptimal responders to therapy with antiresorptive medications according to 2 definitions of serial BMD change.

METHODS

This was a cohort study using administrative databases at a single-payer government health system in Manitoba, Canada. Participants were postmenopausal women aged 40 years or older receiving antiresorptive medications and having 3 sequential BMD measures. Women stopping or switching therapies were excluded. The percentage of women whose spine or hip BMD decreased significantly during the first or second interval of monitoring by BMD was determined. Suboptimal responder status was defined as BMD decrease during both monitoring intervals or BMD decreased from baseline to final BMD.

RESULTS

There were 1369 women in the analytic cohort. Mean BMD monitoring intervals were 3.0 (0.8) and 3.2 (0.8) years. In the first interval, 3.2% and 6.5% of women had a decrease in spine or hip BMD; 8.0% and 16.9% had decreases in the second monitoring interval; but only 1.4% showed repeated losses in both intervals. Considering the entire treatment interval, only 3.2% and 7.4% showed BMD loss at spine or hip. Results may not apply to situations of poor adherence to antiresorptive medication or anabolic therapy use.

CONCLUSION

Among women highly adherent to antiresorptive therapy for osteoporosis, a very small percentage sustained BMD losses on repeated measures. The value of multiple serial BMD monitoring to detect persistent suboptimal responders should be questioned.

Patient Outcomes in the Years After a DXA-BMD Treatment Monitoring Test: Improved Medication Adherence in Some, But Too Little Too Late

The role of mid-treatment monitoring dual-energy X-ray absorptiometry-bone mineral density (DXA-BMD) for bisphosphonate-treated patients with osteoporosis remains unsettled. A common reason for such monitoring is to encourage ongoing medication adherence. We sought to determine if a DXA-BMD treatment monitoring test was associated with improved medication adherence and whether improved adherence after a DXA-BMD treatment monitoring test was associated with subsequent reduction in fracture rates. Using linked administrative databases within Manitoba, Canada, we performed a retrospective cohort study of women starting and continuing antiresorptive therapy in whom a mid-treatment DXA-BMD monitoring test was performed. From the provincial pharmacy database, we estimated medication adherence by calculating annual medication possession ratio (MPR) and determining the change in MPR with respect to change (stable/decrease) in the DXA-BMD monitoring test, in addition to fracture rates before and after the test. The cohort comprised 3418 women, 90.7% treated with oral bisphosphonate, with pharmacy data for the 3 years before and after the mid-treatment DXA-BMD. Median (interquartile range) MPR was 0.84 (0.49-0.99) in the year before DXA-BMD and 0.84 (0.48-0.99) in the year after DXA-BMD (p = 0.37). Among those whose DXA-BMD declined, MPR in the prior year was 0.54 (0.04-0.92) but improved to 0.70 (0.31-0.92) in the year after DXA-BMD (p < 0.001). Among those whose DXA-BMD monitoring test was stable/improved, the fracture rate before the monitoring DXA-BMD was 10.1 per 1000 person-years and in those whose DXA-BMD monitoring test showed a decrease, the rate was 23.7 per 1000 person-years (p < 0.001). Despite improved adherence in those with DXA-BMD decline, the post DXA-BMD fracture rate was 22.4 per 1000 person-years versus 12.9 per 1000 person-years in those who had stable DXA-BMD (p < 0.001). A mid-treatment DXA-BMD reassessment strategy may be useful to focus attention upon adherence, but for optimal fracture outcomes, treatment adherence should be specifically addressed at the commencement of therapy. © 2021 American Society for Bone and Mineral Research (ASBMR).

Assessment of site-specific X-ray procedure codes for fracture ascertainment: a registry-based cohort study

Site-specific X-ray procedure codes are a useful ancillary source of information for identifying fractures in healthcare administrative and claims data.

INTRODUCTION

Real-world evaluation of fracture epidemiology at the population level from electronic healthcare information, such as administrative data, requires comprehensive data sources and validated case definitions. Only hip fractures are routinely hospitalized, and the identification of most osteoporosis-related fractures which are non-hospitalized fractures remains challenging. Plain X-rays (radiographs) are first-line tests for fracture diagnosis and are frequently repeated to monitor fracture healing, and claims data related to radiologic procedures are available in many healthcare systems. We hypothesized that temporal clustering in plain X-ray procedure codes might be an ancillary source of fracture data.

METHODS

We identified individuals age 40 years and older in Manitoba Bone Mineral Density (BMD) Registry with a fracture diagnosis (hip, forearm, humerus, clinical vertebral) before or following a BMD test. A subset underwent detailed review of X-rays to verify an acute fracture. We examined the association between fracture diagnosis and numbers of site-specific X-ray procedures.

RESULTS

The registry cohort included 7793 individuals with a fracture in the previous 5 years and 8417 incident fractures. The X-ray review cohort included 167 radiologically-verified fractures. The number of site-specific X-ray codes was greater in those with vs without fracture (all P < 0.001). The number of days with site-specific X-rays was strongly associated with a fracture diagnosis (area under the curve 0.90 to 0.99 for all non-vertebral fractures, 0.66 to 0.75 for clinical vertebral fractures). There was good agreement between the date of fracture diagnosis and the first X-ray at all non-vertebral fracture sites (Spearman correlation range 0.65 to 0.99), but this was lower for clinical vertebral fractures (range 0.29 to 0.59).

CONCLUSIONS

Temporal clustering in site-specific X-ray procedures was associated with a corresponding fracture diagnosis in administrative medical records. Non-vertebral fracture sites were more strongly associated with X-ray procedures than clinical vertebral fractures.

Machine learning for automated abdominal aortic calcification scoring of DXA vertebral fracture assessment images: A pilot study

BACKGROUND

Abdominal aortic calcification (AAC) identified on dual-energy x-ray absorptiometry (DXA) vertebral fracture assessment (VFA) lateral spine images is predictive of cardiovascular outcomes, but is time-consuming to perform manually. Whether this procedure can be automated using convolutional neural networks (CNNs), a class of machine learning algorithms used for image processing, has not been widely investigated.

METHODS

Using the Province of Manitoba Bone Density Program DXA database, we selected a random sample of 1100 VFA images from individuals qualifying for VFA as part of their osteoporosis assessment. For each scan, AAC was manually scored using the 24-point semi-quantitative scale and categorized as low (score < 2), moderate (score 2 to <6), or high (score ≥ 6). An ensemble consisting of two CNNs was developed, by training and evaluating separately on single-energy and dual-energy images. AAC prediction was performed using the mean AAC score of the two models.

RESULTS

Mean (SD) age of the cohort was 75.5 (6.7) years, 95.5% were female. Training (N = 770, 70%), validation (N = 110, 10%) and test sets (N = 220, 20%) were well-balanced with respect to baseline characteristics and AAC scores. For the test set, the Pearson correlation between the CNN-predicted and human-labelled scores was 0.93 with intraclass correlation coefficient for absolute agreement 0.91 (95% CI 0.89-0.93). Kappa for AAC category agreement (prevalence- and bias-adjusted, ordinal scale) was 0.71 (95% CI 0.65-0.78). There was complete separation of the low and high categories, without any low AAC score scans predicted to be high and vice versa.

CONCLUSIONS

CNNs are capable of detecting AAC in VFA images, with high correlation between the human and predicted scores. These preliminary results suggest CNNs are a promising method for automatically detecting and quantifying AAC.

Bone densitometry categories as a salient distracting feature in the modern clinical pathways of osteoporosis care: A retrospective 20-year cohort study

BACKGROUND

It is unknown as to what factors typically drive osteoporosis therapy decisions in real-world practice.

METHODS

Retrospective, 20-year cohort study within the government health system of Manitoba including all women having bone densitometry (BMD) tests between 1996 and 2017. Osteoporosis prescription data was linked to registry data on fractures, clinical risk factors and BMD tests. We defined 6 possible treatment decisions by prescription data: no treatment, starting, stopping, continuing, drug hiatus and re-starting. For each decision, we tested the association between salient patient factors (age, glucocorticoid use, recent fracture, BMD hip or spine T-score ≤ -2.5, FRAX major osteoporotic fracture probability ≥20%) using multivariable logistic regression. The factors were rank-ordered by decreasing Wald χ² statistic to determine the relative importance.

RESULTS

There were 64,181 women, 33.8% of whom started osteoporosis therapy. For patients who begin therapy after a first BMD, the rank-ordered multivariable logistic regression factor most strongly associated was the T-score ≤ -2.5 [OR of 7.59(95%CI 7.19-8.01, p < 0.001)]. This was followed by glucocorticoid use [OR 2.89(95%CI 2.59-3.22, p < 0.001)]. Increasing age and recent fracture (within 2 years) were weak predictors of therapy and high FRAX score associated with reduced odds of therapy [OR 0.80 (95%CI 0.74-0.88, p < 0.001)]. T-scores were the strongest factor predicting therapy stop/continuation/re-starting; age and prior fracture had weak or no associations.

CONCLUSIONS

Despite recommendations for fracture-risk-based approach to osteoporosis therapy, BMD T-score continues to be the dominant factor in actual practice. Age, prior fracture or global fracture risk are much less associated; it is possible that BMD T-score categories are therefore acting as a clinically salient distracting factor.

Cancer patients with fractures are rarely assessed or treated for osteoporosis: a population-based study

Among 4238 cancer and 16,418 cancer-free individuals with incident major non-traumatic fractures (hip, clinical vertebral, forearm, humerus), post-fracture osteoporosis care was equally poor for both groups, whether assessed from bone mineral density (BMD) testing, initiation of osteoporosis therapy or either intervention (BMD testing and/or osteoporosis therapy).

INTRODUCTION

Most individuals sustaining a fracture do not undergo evaluation and/or treatment for osteoporosis. Cancer survivors are at increased risk for osteoporosis and fracture. Whether cancer survivors experience a similar post-fracture "care gap" is unclear. Using population-based databases, we assessed whether cancer patients are evaluated and/or treated for osteoporosis after a major fracture.

METHODS

From the Manitoba Cancer Registry, we identified cancer cases (first cancer diagnosis between 1987 and 2013) and cancer-free controls with incident major non-traumatic fractures (from provincial physician billing claims and hospitalization databases). The outcomes were performance of BMD testing (from the BMD Registry), initiation of osteoporosis therapy (from drug dispensation database) or either intervention (BMD testing and/or osteoporosis therapy) in the 12 months post-fracture.

RESULTS

There were 4238 cancer and 16,418 cancer-free individuals who sustained a fracture after the index date (cancer diagnosis) and were followed for at least 1 year post-fracture. Subsequent BMD testing was performed in 11.0% of cancer cases versus 11.5% non-cancer controls (P = 0.43), osteoporosis treatment in 22.9% cancer cases versus 21.8% non-cancer controls (P = 0.15), and either testing or treatment in 28.9% cancer cases versus 28.4% non-cancer controls (P = 0.53). Predictors of BMD testing and/or initiation of therapy were similar for non-cancer and cancer patients. Post-fracture interventions were consistently used more frequently among women, older patients (age 50 years or older), those who sustained fractures in a later calendar period, and (for treatment) after vertebral fracture. Cancer-specific variables (cancer type, years from cancer diagnosis to fracture, specialty of care provider) showed only weak and inconsistent effects.

CONCLUSIONS

A large care gap exists among cancer patients who sustain a fracture, similar to the general population, whereby the evaluation or treatment for osteoporosis is seldom conducted. Care maps may need to be developed for cancer populations to improve post-fracture care.

Fracture risk assessment in celiac disease: a registry-based cohort study

Celiac disease is associated with an increased fracture risk but is not a direct input to the FRAX® calculation. When celiac disease is considered as a secondary osteoporosis risk factor or BMD is included in the FRAX assessment, FRAX accurately predicts fracture risk.

INTRODUCTION

The fracture risk assessment tool (FRAX®) uses clinical factors and bone mineral density (BMD) measurement to predict 10-year major osteoporotic (MOF) fracture probability. The study aim was to determine whether celiac disease affects MOF risk independent of FRAX score.

METHODS

The Manitoba BMD Registry includes clinical data, BMD measurements, 10-year probability of MOF calculated for each individual using the Canadian FRAX tool and diagnosed celiac disease. Using linkage to population-based healthcare databases, we identified incident MOF diagnoses over the next 10 years for celiac disease and general population cohorts.

RESULTS

Celiac disease (N = 693) was associated with increased fracture risk adjusted for FRAX score computed without secondary osteoporosis or BMD (adjusted hazard ratio [HR] 1.43, 95% confidence interval [CI] 1.11-1.86). Celiac disease was no longer a significant risk factor for fracture when secondary osteoporosis or BMD were included in the FRAX calculation (p > 0.1). In subjects with celiac disease, each SD increase in FRAX score (calculated with and without secondary osteoporosis or BMD) was associated with higher risk of incident MOF (adjusted HR 1.66 to 1.80), similar to the general population (p-interaction > 0.2). Including celiac disease as secondary osteoporosis or including BMD in FRAX 10-year MOF probability calculations (10.1% and 8.6% respectively) approximated the observed cumulative 10-year MOF probability (10.8%, 95% CI 7.8-13.9%).

CONCLUSIONS

Celiac disease is associated with an increased risk of major osteoporotic fractures. When celiac disease is considered as a secondary osteoporosis risk factor or BMD is included in FRAX assessment, FRAX accurately predicts fracture risk.

Fracture prediction from FRAX for Canadian ethnic groups: a registry-based cohort study

We identified large between-ethnicity calibration differences in the Canadian FRAX® tool which substantially overestimated the major osteoporotic fracture (MOF) risk in Asian women and Black women, and overestimated hip fracture risk in Asian women.

PURPOSE

FRAX® is calibrated using population-specific fracture and mortality data. The need for FRAX to accommodate ethnic diversity within a country is uncertain. We addressed this question using the population-based Manitoba Bone Mineral Density (BMD) Program registry and self-reported ethnicity.

METHODS

The study population was women aged 40 years or older with baseline FRAX assessments (Canadian and other ethnic calculators), fracture outcomes, and self-reported ethnicity (White N = 68,907 [referent], Asian N = 1910, Black N = 356). Adjusted hazard ratios (HR) with 95% confidence intervals (CI) for time to MOF and hip fracture were estimated. We examined candidate variables from DXA that might contribute to ethnic differences including skeletal size, hip axis length (HAL), trabecular bone score (TBS), and estimated body composition.

RESULTS

Adjusted for baseline risk using the Canadian FRAX tool with BMD, Asian women compared with White women were at much lower risk for MOF (HR 0.46, 95% CI 0.35-0.59) and hip fracture (0.16, 95% CI 0.08-0.34). Black women were also at lower MOF risk (HR 0.58, 95% CI 0.32-1.00); there were no hip fractures. The US ethnic-specific FRAX calculators accounted for most of the between-ethnicity differences in MOF risk (86% for Asian, 92% for Black) but only partially accounted for lower hip fracture risk in Asian women (40%). The candidate variables explained only a minority of the effect of ethnicity. Gradient of risk in analyses was similar (p-interactions ethnicity*FRAX non-significant).

CONCLUSIONS

We identified significant ethnic differences in performance of the Canadian FRAX tool with fracture probability overestimated among Asian and Black women. The US ethnic calculators helped to address this discrepancy for MOF risk assessment, but not for hip fracture risk among Asian women.

Frequency of normal bone measurement in postmenopausal women with fracture: a registry-based cohort study

This registry-based cohort study assessed the percentage of women with prior or incident fracture who had normal bone defined as a normal bone mineral density T-score and normal trabecular bone score (TBS). Inclusion of TBS reduced the percentage with normal bone. Normal bone measurement is rare in women with fracture.

INTRODUCTION

Some fractures occur in women with normal BMD. We hypothesized that adding trabecular bone score (TBS) to DXA would (1) demonstrate that few women with fracture have normal bone, i.e., normal BMD T-score and TBS and (2) increase the percentage of women with fracture that have abnormal bone defined as a BMD T-score ≤ - 2.5 or low TBS.

METHODS

The public healthcare system in Manitoba, Canada, makes it possible to link clinical DXA data to population databases. This study included all women age 50+ with a first DXA from February 1999 to March 2018 with valid BMD, TBS, and fracture data. Bone status was defined as Normal = BMD T-score of the spine, femoral neck, and total femur ≥ - 1.0 AND TBS > 1.31; Abnormal = BMD T-score ≤ - 2.5 OR TBS < 1.23; and borderline = all others. Analyses were stratified by age decade.

RESULTS

Among women with prior (n = 4649) or incident (n = 2547) fracture, bone status assessed by both BMD and TBS was normal in only 6% and 4%, respectively. In women with prior or incident hip fracture, normal bone was present in < 1%. The prevalence of normal bone declined (p trend < 0.001) with age as expected. BMD T-score osteoporosis was present in 40% with any prior and 46% with any incident fracture. BMD T-score osteoporosis was present in 65% and 60% with prior and incident hip fracture, respectively. Including TBS with BMD increased the percentage of women with abnormal bone to 61% and 68% for any prior or incident fracture and to 80% and 81% for prior or incident hip fracture, respectively (all p < 0.001).

CONCLUSION

Including TBS with BMD increases identification of abnormal bone in women with fracture compared with BMD alone. Normal bone is present in < 6% of women with any fracture and < 1% of those with hip fracture. What is thought to be normal bone in women with fracture is rarely normal.

Comparison of screening tools for optimizing fracture prevention in Canada

The best screening strategy to identify treatment qualification based upon indicators of high fracture risk (low-trauma fractures of the hip, spine, or multiple fracture episodes at other sites; high fracture probability with the Canadian fracture risk assessment [FRAX®] tool major osteoporotic fracture [MOF] computed with bone mineral density [BMD] > 20%; or vertebral fracture on vertebral fracture assessment [VFA]) was FRAX-MOF without BMD using a cutoff of ≥ 10%.

PURPOSE

To inform clinical practice guidelines in Canada, we compared multiple screening tools using the population-based Manitoba BMD Program registry.

METHODS

The study populations consisted of (a) 28,906 individuals > 50 years or older, and (b) 15,429 women age > 65 years undergoing baseline BMD assessment (2010-2018). We considered two treatment qualifications: Treatment Approach 1: prior high-risk fracture, high fracture probability (FRAX-MOF with BMD > 20%), or vertebral fracture on VFA; Treatment Approach 2: Approach 1 or an osteoporotic BMD T score. Candidate screening tools were FRAX-MOF without BMD, age alone, weight alone, SCORE, ORAI, SOFSURF, OSIRIS, ABONE, and OST. Healthcare records were assessed for the presence of incident fracture diagnoses.

RESULTS

Among all individuals, FRAX-MOF without BMD demonstrated the best ability to identify those satisfying Treatment Approach 1 (area under the curve [AUC 0.863]) and was significantly better than all other screening tools (P < 0.001). For identification of individuals satisfying Treatment Approach 2, FRAX-MOF without BMD showed moderate stratification (AUC 0.735), slightly lower than OSIRIS (AUC 0.752, P < 0.05), similar to SCORE (AUC 0.739, P > 0.05) and significantly better than all other screening tools (P < 0.05). For prediction of incident MOF, FRAX-MOF without BMD achieved the highest performance (AUC 0.652), and was significantly better than all other screening tools except OSIRIS. AUCs among women age > 65 years tended to be greater with a similar ranking, and no tool outperformed FRAX-MOF without BMD. Based upon a summary score, the highest ranked strategy was FRAX-MOF without BMD using a cutoff of 10%.

CONCLUSIONS

All screening tools show some ability to identify individuals qualifying for treatment and stratify risk for incident fracture. For treatment based upon indicators of high fracture risk, the best performing strategy was FRAX-MOF without BMD using a cutoff of ≥ 10%.

Estimating Waist and Hip Circumference from Routine Clinical DXA

INTRODUCTION

Waist circumference and waist:hip ratio have body mass index-independent detrimental effects on health and mortality. Dual-energy X-ray absorptiometry (DXA) scans of the lumbar spine and hip can provide site-specific measures of soft-tissue thickness, and we hypothesized that this could be used to opportunistically estimate body circumference in patients undergoing DXA for osteoporosis assessment.

METHODOLOGY

We assessed the correlation and explained variance (as coefficient of determination, R²) between directly measured body circumference (waist circumference, hip circumference, and waist:hip circumference ratio) with DXA-derived measures of soft tissue thickness (spine DXA tissue thickness, hip DXA tissue thickness, and spine:hip tissue thickness ratio) in 214 women and 96 men (mean age 66.1 and 63.7 yr, respectively) undergoing DXA screening for osteoporosis.

RESULTS

DXA-derived spine tissue thickness explained most of the variance in measured waist circumference (female R² 0.90, male R² 0.88). Explained variance was slightly lower for measured hip circumference (female R² 0.87, male R² 0.76) and waist:hip ratio (female R² 0.68, male R² 0.72). Final models predicted waist circumference with an adjusted R² 0.91, hip circumference with R² 0.86, and waist:hip ratio with R² 0.70.

CONCLUSION

Routine clinical DXA measurements of the spine and hip can be used to estimate body circumference measurements.

Trabecular Bone Score (TBS) Predicts Fracture in Ankylosing Spondylitis: The Manitoba BMD Registry

INTRODUCTION

Ankylosing spondylitis (AS) is a chronic inflammatory disease of the spine characterized among other features by spinal boney proliferation, back pain, loss of flexibility, and increased fracture risk. Overlying bone limits the utility of bone mineral density (BMD) by dual X-ray absorptiometry (DXA) in the spine. Trabecular bone score (TBS) is a bone texture measurement derived from the spine DXA image that indicates bone quality and fracture risk independent of BMD.

METHODOLOGY

Using the Manitoba Bone Density Program database, patients with diagnosis codes for ankylosing spondylitis, baseline DXA and lumbar spine TBS were identified. Incident nontraumatic fractures (major osteoporotic [MOF], clinical spine, hip, and all fracture) were identified from population based databases. Cox-proportional hazard models are presented.

RESULTS

We identified 188 patients with diagnosed AS. TBS was lower in those with incident MOF (1.278 ± 0.126, compared to 1.178 ± 0.136, p < 0.001). Unadjusted TBS and FRAX-MOF-BMD adjusted predicted major osteoporotic fracture (N = 19) (hazard ratio [HR] 2.04, 95% confidence interval [CI]: 1.28-2.26, p = 0.003; HR 1.81, 95% CI: 1.11-2.96, p = 0.018). TBS unadjusted and FRAX-MOF-BMD adjusted also predicted clinical spine fracture (N = 7) (HR 2.50, 95% CI: 1.17-5.37; p = 0.019; HR 2.40 95% CI: 1.1-5.25; p = 0.028). Higher HRs were observed for prediction of hip fracture (N = 6), but these did not achieve statistical significance (FRAX-adjusted HR 1.74, 95% 0.73-4.17; p = 0.211). Unadjusted models show TBS was predictive of all fracture (N = 27) (HR 1.60, 95% CI: 1.08-2.39; p = 0.020), which was borderline significant after adjustment for FRAX-MOF-BMD (HR 1.51, 95% CI: 1.00-2.29; p = 0.052).

CONCLUSION

We report the first analysis of TBS for fracture prediction as an incident event in AS. TBS independently predicted major osteoporotic and clinical spine fracture in AS independent of FRAX.

Factors Associated With Bone Density Monitoring While on Antiosteoporosis Treatment in Routine Clinical Practice: A Registry-Based Cohort Study

The role for bone mineral density (BMD) monitoring while on antiosteoporosis therapy remains controversial. The current study used population-based registries to identify factors associated with BMD monitoring in women within 5 yr of receiving antiosteoporosis treatment vs treatment without monitoring in routine clinical practice. The analytical dataset consisted of women age 40 yr and older at baseline receiving antiosteoporosis therapy: 6877 with BMD monitoring (mean interval 3.2 yr) and 6747 without BMD monitoring. There was a significant negative secular trend in BMD monitoring during the study period (p < 0.001). Multivariable logistic regression demonstrated that parental hip fracture, glucocorticoid and aromatase inhibitor use, and lower baseline BMD were independently and positively associated with BMD monitoring. Individuals with increasing age, greater body mass index, smoking, rheumatoid arthritis, later calendar year, diabetes, rural residency, lower income, and greater comorbidity score were less likely to undergo monitoring. A shorter monitoring interval (<23 mo) was strongly associated with glucocorticoid and aromatase inhibitor use. In conclusion, our study identifies factors associated with BMD monitoring over 5 yr in patients receiving antiosteoporosis therapy.

Targeted bone density testing for optimizing fracture prevention in Canada

The Canadian FRAX® tool used without bone mineral density (BMD) is highly sensitive for identifying individuals qualifying for pharmacotherapy based upon an intervention threshold of 20% for major osteoporotic fracture risk (MOF) computed with BMD.

INTRODUCTION

This analysis was performed to inform initial BMD testing as part of Osteoporosis Canada's Guidelines Update for women and men at average risk, assuming a pharmacotherapy intervention threshold of 20% for FRAX® MOF computed with BMD.

METHODS

Women and men age 50 + without previous low-trauma fracture or high-risk medication use were identified in a BMD registry for the province of Manitoba, Canada. Fracture probability assessments with the Canadian FRAX® tool were computed without and with BMD (denoted MOF-clinical and MOF-BMD, respectively).

RESULTS

The study population consisted of 50,700 women (mean age 65.5 ± 9.4 years) and 4152 men (69.2 ± 10.0 years). FRAX MOF-clinical score was > 10% in 33.8% of women and 13.3% of men (P < 0.001). The median (interquartile range [IQR]) age for MOF-clinical to reach 10% in women was 70 (69-72) and 65 years (62-67) years in the absence and presence of additional FRAX clinical risk factors, respectively. In men, comparable ages were 83 years [82-86] and 76 [70-78] years. Using MOF-BMD of 20% as the intervention threshold, 4.3% of women and 0.7% of men qualified for treatment. MOF-clinical > 10% had high sensitivity to identify those qualifying for treatment (99.3% in women and 99.1% in men). An age-based rule ("BMD testing is indicated at age 70 if no additional FRAX clinical risk factors are present, or at age 65 if one or more clinical risk factors exists") gave similarly high sensitivity (women 99.9% and men > 99.9%).

CONCLUSIONS

FRAX without BMD offers an effective strategy to identify individuals meeting the current Canadian treatment threshold based upon FRAX® with BMD (≥ 20%). Moreover, this can be operationalized using simple age cutoffs of 70 years in the absence of additional clinical risk factors and 65 years in the presence of additional clinical risk factors.

Loss in DXA-estimated total body lean mass but not fat mass predicts incident major osteoporotic fracture and hip fracture independently from FRAX: a registry-based cohort study

During median follow-up 6.0 years in 9622 individuals, prior loss in estimated total body lean mass (TBLM), but not total body fat mass loss (TBFM), was associated with increased fracture risk, particularly for hip fracture.

INTRODUCTION

Weight loss, and especially muscle loss, adversely affects skeletal health. The FRAX® tool considers baseline body mass index, but not body composition nor changes in its components over time. Our aim was to compare the independent associations between prior loss in DXA-estimated TBLM and TBFM and subsequent fracture risk.

METHODS

We identified women and men age 40 years or older with two DXA assessments at least 1 year apart (median interval 3.3 years). TBLM and TBFM were estimated from weight, sex, and DXA of the spine and hip. Incident fractures and deaths were ascertained from linked population-based health service data after the date of the second DXA. Hazard ratios (HRs) from Cox regression models were used to study time to fracture from prior loss in TBLM and TBFM adjusted for FRAX-related covariates.

RESULTS

The study population consisted of 9622 individuals (mean age 67 [SD 10] years, 95% female). We identified 692 subjects with incident major osteoporotic fracture [MOF] and 194 with hip fracture. Mean TBLM loss was significantly greater in those with incident MOF and hip fracture (P < 0.001) while TBFM loss was only significantly greater in those with incident hip fracture (P < 0.001). Each SD greater TBLM loss was associated with 10-13% increased MOF risk and 29-38% increased hip fracture risk, adjusted for TBFM loss and other covariates. Prior TBFM loss was not associated with fractures when adjusted for TBLM loss.

CONCLUSIONS

Prior loss in total body lean mass, but not in fat mass, is associated with increased fracture risk, particularly hip fracture, independent of other risk factors. This is consistent with the hypothesis that muscle loss (sarcopenia) adversely impacts skeletal health and fracture risk.

Osteoporosis treatment considerations based upon fracture history, fracture risk assessment, vertebral fracture assessment, and bone density in Canada

Among 39,475 women, age 65 years and older, use of fracture history, major osteoporotic fracture (MOF) probability from FRAX®, vertebral fracture assessment (VFA), and bone mineral density (BMD) T-score stratified women into different levels of risk. The majority of women identified as being at high risk from fracture history, FRAX MOF-BMD > 20%, or vertebral fracture on VFA had a BMD T-score in the osteoporotic range.

PURPOSE

To inform criteria for pharmacologic treatment in women age 65 years and older, we examined subgroups defined from fracture history, MOF calculated with BMD (MOF-BMD), VFA, and BMD T-score using the population-based Manitoba BMD Program registry.

METHODS

The study population consisted of women age > 65 years was divided into mutually exclusive subgroups based upon fracture history, MOF-BMD ≥ 20%, vertebral fracture on VFA, and osteoporotic BMD T-score. Healthcare records were assessed for the presence of fracture diagnosis codes occurring after DXA assessment. For each subgroup, we estimated the proportion of individuals with BMD T-score in the osteoporotic range, predicted versus observed 10-year MOF probability, hazard ratio (HR) for MOF, and number needed to treat (NNT) for 3 years to prevent a fracture event.

RESULTS

The study population consisted of 39,475 women (median age 72 years). The majority of women (76.8%) selected as being at high risk based on fracture history, MOF-BMD > 20%, or vertebral fracture on VFA had a BMD T-score in the osteoporotic range. During a median follow-up of 8 years, 5169 (13.1%) sustained one or more incident MOF. Fracture rates and HRs generally paralleled the FRAX prediction, except in women with a positive VFA where predicted risk based upon clinical risk factors prior to VFA underestimated the observed risk. NNT differed by the risk subgroup, and showed a gradient of decreasing NNT (consistent with greater benefit) in individuals with the highest fracture risk.

CONCLUSIONS

Fracture history, fracture probability from FRAX, targeted vertebral fracture assessment (VFA), and BMD T-score can stratify older women into different levels of risk and treatment benefit. These results are expected to inform clinical practice guidelines in Canada.

Impact of spine-hip discordance on fracture risk assessment and treatment qualification in Canada: the Manitoba BMD registry

In 50,330 women undergoing bone mineral density (BMD) testing, adjusting FRAX® major osteoporotic fracture (MOF) probability for T-score differences between lumbar spine and femoral neck had a small impact on treatment qualification, with only 1.1% reclassified from below to above the 20% MOF intervention threshold.

PURPOSE

Discordance between lumbar spine and femoral neck T-score measurements are frequently encountered in clinical practice. The FRAX tool accepts femoral neck BMD as an optional input but does not consider lumbar spine BMD, though adjustments based upon T-score difference have been proposed. This analysis was undertaken to examine change in treatment qualification using the lumbar spine T-score adjustment to FRAX assuming an intervention threshold of 20% MOF.

METHODS

Women aged > 50 years with baseline FRAX MOF probability computed with femoral neck BMD of ≥ 20% but without previous high-risk fracture or high-risk medication use were identified in the province of Manitoba BMD registry. If lumbar spine T-score was lower than the femoral neck T-score, MOF probability was recalculated, and treatment qualification, reassessed.

RESULTS

The study population consisted of 50,300 women (mean age 64.2 ± 8.6 years). During mean follow-up of 9.5 years, 4550 sustained incident MOF. The baseline mean T-score difference (femoral neck minus lumbar spine) was - 0.2 ± 1.1. Recalculated MOF probability using the T-score difference significantly improved overall net reclassification index for incident MOF prediction (+ 0.017, p < 0.001). A total of 561 (1.1%) of these women were reclassified from below to above 20% MOF probability based upon the T-score difference. No individuals with MOF probability less than 15% were reclassified to the higher risk category. Risk reclassification increased with lower lumbar spine T-score (15.6% reclassification for lumbar spine T-score of - 3.5 or lower, 4.0% of the study population) and larger T-score differences (9.7% reclassification for T-score difference of 2.5 or more, < 1% of the study population).

CONCLUSIONS

Very few women showed risk reclassification based upon adjusting FRAX MOF probability for lumbar spine T-score. Reclassification occurred only in those with baseline MOF probability > 15%.

Fracture risk following high-trauma versus low-trauma fracture: a registry-based cohort study

Prior high-trauma fractures identified through health services data are associated with low bone mineral density (BMD) and future fracture risk to the same extent as fractures without high-trauma.

INTRODUCTION

Some have questioned the usefulness of distinguishing high-trauma fractures from low-trauma fractures. The aim of this study is to compare BMD measurements and risk of subsequent low-trauma fracture in patients with prior high- or low-trauma fractures.

METHODS

Using a clinical BMD registry for the province of Manitoba, Canada, we identified women and men age 40 years or older with fracture records from linked population-based healthcare data. Age- and sex-adjusted BMD Z-scores and covariate-adjusted hazard ratios (HR) with 95% confidence intervals (CI) for incident fracture were studied in relation to prior fracture status, categorized as high-trauma if associated with external injury codes and low-trauma otherwise.

RESULTS

The study population consisted of 64,428 women and men with no prior fracture (mean age 63.7 years), 858 with prior high-trauma fractures (mean age 65.1 years), and 14,758 with prior low-trauma fractures (mean age 67.2 years). Mean Z-scores for those with any prior high-trauma fracture were significantly lower than in those without prior fracture (P < 0.001) and similar to those with prior low-trauma fracture. Median observation time for incident fractures was 8.8 years (total 729,069 person-years). Any prior high-trauma fracture was significantly associated with increased risk for incident major osteoporotic fracture (MOF) (adjusted HR 1.31, 95% CI 1.08-1.59) as was prior low-trauma fracture (adjusted HR 1.55, 95% CI 1.47-1.63), and there was no significant difference between the two groups (prior trauma versus low-trauma fracture P = 0.093). A similar pattern was seen when incident MOF was studied in relation to prior hip fracture or prior MOF, or when the outcome was incident hip fracture or any incident fracture.

CONCLUSIONS

High-trauma and low-trauma fractures showed similar relationships with low BMD and future fracture risk. This supports the inclusion of high-trauma fractures in clinical assessment for underlying osteoporosis and in the evaluation for intervention to reduce future fracture risk.

Measured height loss predicts incident clinical fractures independently from FRAX: a registry-based cohort study

During median follow-up 6.0 years in 11,495 individuals, prior absolute and annualized measured height loss was significantly greater in those with subsequent incident fracture compared with those without incident fracture.

PURPOSE

FRAX® accepts baseline height and weight as input variables, but does not consider change in these parameters over time.

AIM

To evaluate the association between measured height or weight loss on subsequent fracture risk adjusted for FRAX scores, risk factors, and competing mortality.

METHODS

Using a dual-energy x-ray absorptiometry (DXA) registry for the Province of Manitoba, Canada, we identified women and men age 40 years or older with height and weight measured at the time of two DXA scans. Cox regression analyses were performed to test for a covariate-adjusted association between prior height and weight loss with incident fractures occurring after the second scan using linked population-based healthcare data.

RESULTS

The study population consisted of 11,495 individuals (average age 68.0 ± 9.9 years, 94.6% women). During median follow-up 6.0 years, records demonstrated incident major osteoporotic fracture (MOF) in 869 individuals, hip fractures in 265, clinical vertebral fractures in 207, and any fracture in 1203. Prior height loss was significantly greater in individuals with fracture compared with those without fracture, regardless of fracture site. Mortality was greater in those with prior height loss (HR per SD 1.11, 95% CI 1.06-1.17) or weight loss (HR per SD 1.26, 95% CI 1.19-1.32). Each SD in height loss was associated with increased fracture risk (MOF 12-17%, hip 8-19%, clinical vertebral 28-37%, any fracture 14-19%). Prior weight loss was associated with 21-30% increased risk for hip fracture, but did not increase risk for other fractures. Height loss of 3.0 cm or greater more than doubled the risk for subsequent fracture.

CONCLUSIONS

Prior height loss is associated with a small but significant increase in risk of incident fracture at all skeletal sites independent of other clinical risk factors and competing mortality as considered by FRAX. Prior weight loss only increases risk for subsequent hip fracture.