Definition of Italian Specific DXA References for Diagnosis of Osteoporosis: Preliminary Data from the Osteoporosis Registry of Magenta Rheumatology School (OSTEOREMA)

Abstract

Background: Osteoporosis is the most common metabolic bone disease and dual energy X ray absorptiometry (DXA) scans is the gold standard to evaluate individuals at risk of osteoporosis. Up to date, studies focused on population specific DXA values among general population are lacking. Primary aim of this study was to obtain lumbar and femoral DXA values of Italy adult females based on a nationally representative sample aged 20 years and older, attempting to obtain a national specific normative bone mineral density (BMD) levels curve.

Methods: demographic and anthropometric data of females aged 20 years and older that performed their first DXA (QDR 9000 Hologic, Waltham, Mass.) in our Hospital between 2006 and 2015 were extracted from our local registry using a random sampling technique. Criteria for patient choice were: absence of known risk factors of Osteoporosis (e.g. smoke, alcohol), metabolic disease that affects bone (e.g. diabetes), a normal BMI between 18.5 to 24.5, previous fractures, any medication for treatment of osteoporosis or corticosteroids, spondylosis radiologically relevant. Our database was compared to Caucasian normal values incorporated into Hologic’s scan analysis software, that we proven were comparable in terms of BMI, gender and age.

Results: DXA scan of 15335 women were extracted and analysed. Mean age was 64.2 ± 12.8 years (range 20.8 to 90). Mean BMI was 22.4± 5.1 (range 20.1 to 24.3). Mean menopause age was 41.3±5 (range 31-54). Mean menarche age was 16.3±5 (range 11-17). The lumbar and femoral BMD were substantially constant between 25 and 35 years (test for trend using ANOVA: P =0.31); these data collected in premenopausal women (mean 1.043± 0.12 g/cm² for lumbar spine and 0.97± 0.136 g/cm² for femoral neck) were thus defined as the reference peak bone mass values, significantly lower compared to the Hologic reference values (mean 1.079±0.11 g/cm², p<0.05). The frequency of osteopenia and osteoporosis were so significantly different depending on whether you use the manufacturer criteria rather than those derived from collected data (X square p=0,01).

Conclusions: our data suggest that the reference curves for the lumbar spine and femoral neck are significantly different from the current normative data reported by the manufacturer for the Italian population.