March 14, 2019
2 min read

BMI, bone density mediate insulin resistance and fracture risk

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact

Among older adults without diabetes, a relationship between increasing insulin resistance and lower risk for fracture is mediated by BMI and bone mineral density, suggesting that the association is explained by higher BMI and BMD levels often seen in insulin resistance, according to an analysis of data from the Health ABC study.

Nicola Napoli

“It is not clear whether the net effect of [insulin resistance] is favorable or unfavorable for the risk of fractures,” Nicola Napoli, MD, PhD, of the division of endocrinology and diabetes at the University Campus Bio-Medico di Roma in Rome, and colleagues wrote in the study background. “It is possible that, being positively associated with BMD, [insulin resistance] reduces fracture risk. On the other hand, [type 2 diabetes] is associated with higher fracture risk in spite of higher BMD, and [insulin resistance] may have a similar relationship to fracture.”

Napoli and colleagues analyzed data from 2,398 white and black adults aged 70 to 79 years at baseline participating in the Health ABC study, a prospective cohort study designed to assess whether changes in body composition act as a common pathway by which multiple diseases affect morbidity, disability and mortality (48.4% men; 41.6% black adults; mean age, 74 years). Participants with diabetes were excluded. Baseline data collected included DXA hip scans to assess areal BMD and fasting blood samples to measure fasting glucose and insulin levels. Researchers calculated insulin resistance via the homeostatic model assessment of insulin resistance (HOMA-IR). Participants self-reported fracture data during 15 years of follow-up (median follow-up, 11.9 years). Researchers used Cox proportional hazard models to assess the association between quartiles of HOMA-IR and incident fracture. Cutoff values for HOMA-IR across quartiles were 0.019 to 1.05 (reference), 1.053 to 1.54, 1.545 to 2.33 and 2.333 to 17.45.

Researchers found that total hip BMD increased across HOMA-IR quartiles. In unadjusted models, the mean difference in total hip BMD between the lowest and highest quartiles was 0.104 g/cm² (P < .001). However, after adjustment for BMI, BMD, age, sex and race, the mean difference fell to 0.007 g/m² (P = .371). Results persisted in analyses excluding adults taking oral glucocorticoid medications (n = 54).

During 23,905 person-years of follow-up, 405 participants experienced at least one non-spine fracture. In unadjusted models, researchers observed the lowest fracture risk among adults in the highest HOMA-IR quartile (HR = 0.65; 95% CI, 0.47-0.89) when compared with adults in the first quartile. However, in models adjusted again for BMI and BMD, the trend shifted toward an increased fracture risk among adults in the highest HOMA-IR quartile (P for trend = .215).


Researchers observed no evidence of a linear relationship between HOMA-IR and fracture risk in the model adjusted for BMI and BMD (P = .215).

“Instead, the pattern of fracture risk suggested a threshold effect with the increased risk limited to those in the upper two quartiles of HOMA-IR,” the researchers wrote. “When we compared fracture risk in those with HOMA-IR above the median (1.54) and those below the median, the HR was 1.16 (95% CI, 0.99-1.36).”

The researchers noted that using HOMA-IR to estimate insulin resistance is a study limitation, as the gold standard is the hyperinsulinemic-euglycemic clamp technique. Additionally, HOMA-IR, BMI and BMD were measured at baseline only and may change over time, and the design of the Health ABC study included a healthier cohort of older adults. – by Regina Schaffer

Disclosures: The authors report no relevant financial disclosures.