Is prediabetes associated with increased fracture risk among midlife women?

By Neha MathurMay 25 2023Reviewed by Lily Ramsey, LLM

In a recent article published in JAMA Network Open, researchers performed a cohort study using data of 1,690 midlife women without diabetes retrieved between January 6, 1996, and February 28, 2018, by another multi-center study named the Study of Women's Health Across the Nation (SWAN).

They examined whether there was a correlation between prediabetes before the menopause transition (MT) among midlife women and a subsequent increased fracture risk even when they did not develop type 2 diabetes (T2D).

Study: Prediabetes and Fracture Risk Among Midlife Women in the Study of Women’s Health Across the Nation. Image Credit: VGstockstudio/Shutterstock.com

Background

Researchers have recognized that diabetes results in 'diabetic bone disease' and fractures; however, it remains uncertain whether prediabetes is also a risk factor for such end-organ complications.

Nonetheless, there is a high possibility of an association between prediabetes and increased fracture risk among midlife women who do not eventually develop T2D, as these women have higher insulin resistance, which lowers their bone mineral density (BMD).

It also reduces indices of their hip strength, trabecular bone score, and rapid bone loss - all related to increased fracture risk.

Clinicians disagree on how aggressively to treat prediabetes because it does not always manifest as T2D causing end-organ complications. Thus, studies elucidating the clinical relevance of increased fracture risk due to prediabetes are crucial.

About the study

In the present study, researchers recruited all SWAN Bone Cohort participants who did not have T2D before the MT and did not take bone-beneficial medication.

They made one or more study visits before the MT and a minimum of one study visit after the MT, defined as a less predictable menstrual cycle at least once between three and 12 months.

The final study analysis dataset covered 1,690 midlife women aged 42 to 52 years in premenopause (unchanged menstrual cycles) or early perimenopause. The average (SD) study follow-up period was 12 (6) years though the researchers censored participants at the initiation of bone-beneficial medications or new T2D diagnosis.

Thus, the observed association of prediabetes with fracture could not be attributed to future progression to T2D.

The team used prespecified criteria for prediabetes, i.e., fasting blood glucose levels between 100 and 125 mg/dL. They treated prediabetes exposure as a continuous variable with values ranging between zero and one, which helped capture the consistency of a participant's prediabetes.

Further, the team modeled another primary exposure, i.e., prediabetes before the MT, as the proportion of visits from the SWAN baseline visit (during which they recorded Pre-SWAN fractures) till the last SWAN visit before the MT at which they ascertained the presence of prediabetes.

The primary study outcome was the first fracture after the MT, and the team ascertained its occurrence and location using standardized questions administered at all study visits.

The researchers included only traumatic and atraumatic fractures, occurring after falling from a height of <15.2cm but not due to accident, playing sports, rapid movement, or impact with large-sized objects because they are risk factors for fractures in the future.

During the first six follow-up visits, the team credited the fracture date as the midpoint between the participant's previous and index visits. After adjudication began at follow-up visit 7, it confirmed 95% of reported fractures.

Additionally, the team performed assays to measure the fasting blood glucose levels of all participants at two different SWAN laboratories, Medical Research Laboratories in Lexington, Kentucky, and the University of Michigan.

Finally, they developed an inter-laboratory calibration equation to convert glucose results obtained at the Medical Research Laboratories through SWAN follow-up visit 7 to the corresponding University of Michigan values.

Furthermore, the researchers adjusted for multiple covariates, such as age, body mass index (BMI) at the MT, and exposure to bone-depleting medications before the MT to address the first study objective; notably, all these covariates had a potential relation to fracture.

For the second study objective, they adjusted BMD for the lumbar spine (LS) or femoral neck (FN) at MT, measured by dual x-ray absorptiometry.

The team conducted two sets of primary analyses between January and May 2022. For the first, they used a Cox proportional hazards regression model, which examined the association between prediabetes before the MT and subsequent fracture risk.

Models 2A and 2B evaluated whether the predicted correlation of prediabetes with fracture did not depend on BMD. The models generated the hazard ratio (HR) for prediabetes at all pre-MT visits compared to prediabetes at none of the MT visits (exposure =1 vs. 0).

Finally, the team also conducted two sensitivity analyses, where they reran study models first using a binary (yes or no) exposure to prediabetes before the MT and then using the original prediabetes exposure continuous variable to assess only nonvertebral fractures as outcomes.

Results and conclusion

The main takeaway of this study was that prediabetes before the MT increased the risk of fractures by 120%. Whereas during the MT and after menopause (HR=2.20 [95% CI]), independent of BMD at the start of the MT, thus, confirming that prediabetes in midlife women being detrimental to bone health and might be a risk factor for future fractures.

It translates to an absolute increase of nearly three and seven fractures per 1,000 person-years for women with prediabetes at 50% of the pre-MT and all pre-MT visits, respectively.

Strikingly, the authors noted that adjusting additionally for LS or FN BMD at the start of the MT did not markedly alter the magnitude or significance of this association, with HRs of 2.24 and 2.26 for models 2A and 2B.

Most fractures in this study cohort occurred at nonvertebral sites, which increased the risk of subsequent vertebral or hip fracture two-fold. Thankfully, prediabetes before the MT is an early modifiable risk factor for fracture.

To conclude, future studies should examine whether treating prediabetes before the MT could reduce the risk of poorer fracture outcomes in later life.