Background/Purpose:

The risk factors for osteoporotic fragility fractures are well documented in the literature and although they are associated with a diagnosis of established osteoporosis (T score < -2.5) it is known that the majority of fragility fractures occur in osteopenic individuals (T score between -1.0 and -2.5)¹. It is not known whether traditional risk factors for osteoporosis convey additional fracture risk in an osteopenic population. Objectives:To identify and examine the predictors of fracture in patients with diagnosed osteopenia.

Methods: Using a nested case control approach, patients referred for a first dual energy X-ray absorptiometry (DXA) scan at a district scanner in the North West of England between 2004 and 2012 were identified and their data included in the analysis if they were found to be osteopenic at the lumbar spine and/or femoral neck. Data collected included: age at scan, gender and indication information (including history of fracture). Univariate logistic regression was used to estimate the odds of fracture in this group using traditional risk factors and subsequently multivariate logistic models were fitted to predict any fracture for using the hip and spine bone mineral density (BMD).

Results: A total of 5021 patients were diagnosed with osteopenia at the femoral neck, of whom 80.5% were female and mean age at scan was 65.1 years (standard deviation 11.8 years). 1709 of these had a history of fracture. At the lumbar spine, 3737 patients were found to be osteopenic, of whom 83.6% were female and the mean age was 64.0 years (standard deviation 12.4 years). 1300 of these had a history of fracture.

Odds ratios for risk of fracture with the presence of a classical fracture risk factor are listed in table 1. Table 2 contains the coefficients of the multivariate logistic models for fracture risk. Standardised residuals suggest good fit for both multivariatelogistic models as they lie between 2 and -2.

	Femoral Neck	Lumbar Spine
Age (per year)	1.03 (1.02, 1.03)	1.04 (1.03, 1.04)
BMI (per unit)	1.03 (1.02, 1.05)	1.03 (1.02, 1.05)
T-score at site (per unit)	0.59 (0.51, 0.67)	0.78 (0.66, 0.91)
Female	1.18 (1.02, 1.37)	1.16 (0.96, 1.39)
Maternal Hip Fracture	1.66 (1.09, 2.55)	2.54 (1.54, 4.21)
Smoking	1.09 (0.96, 1.23)	1.15 (0.99, 1.32)
Alcohol Excess	1.37 (1.09, 1.72)	1.32 (1.01, 1.72)
Corticosteroid Therapy	0.55 (0.47, 0.64)	0.63 (0.53, 0.75)
Rheumatoid Arthritis	0.49 (0.35, 0.68)	0.56 (0.38, 0.82)
Inflammatory Bowel Disease	0.38 (0.23, 0.63)	0.42 (0.25, 0.72)
Coeliac Disease	0.33 (0.21, 0.52)	0.43 (0.27, 0.69)
Hyperthyroidism	0.89 (0.34, 2.36)	1.46 (0.54, 3.93)
Hyperparathyroidism	0.51 (0.29, 0.91)	0.52 (0.27, 1.03)

Table 1. Odds ratios for fracture.

 

Coefficients	Femoral Neck: Estimate	p-value	Lumbar Spine: Estimate	p-value
Age (per year)	0.02	<0.001	0.04	<0.001
BMI (per unit)	0.03	<0.001	0.03	<0.001
T-score at site (per unit)	-0.45	<0.001	-0.18	<0.05
Female	NS		NS
Maternal Hip Fracture	0.57	<0.05	1.01	<0.001
Smoking	NS		0.17	<0.05
Alcohol Excess	0.43	<0.001	0.37	<0.01
Corticosteroid Therapy	-0.57	<0.001	-0.45	<0.001
Rheumatoid Arthritis	-0.63	<0.001	-0.49	<0.05
Inflammatory Bowel Disease	NS		NS
Coeliac Disease	-0.90	<0.001	-0.57	<0.05
Hyperthyroidism	NS		NS
Hyperparathyroidism	-0.92	<0.01	-0.83	<0.05

Table 2.Results of multivariate logistic modelling for both sites: the femoral neck and lumbar spine. Variables that did not significantly add to the model show NS.

Conclusion:

Increasing age, reducing T-score, maternal hip fracture and excessive alcohol consumption predict fracture in osteopenic individuals irrespective of site. Other classical risk factors are unreliable predictors of fracture in this cohort, some even showing protective effects. Differences exist between predictors of fracture at the two sites. Possible unmeasured confounders could predict fractures in this cohort. Further research is needed.

References:

1. Arch Intern Med. 2004 May;164(10):1108-1112.

---

« Back to 2013 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/predictors-of-fracture-in-patients-with-osteopenia-data-from-an-observational-cohort/