HALP score-adjusted FRAXplus tool effectively predicts hip fracture risk in elderly women with type 2 diabetes...

Elderly women with type 2 diabetes mellitus (T2DM) are at high risk of fragility fractures, particularly lethal hip fractures. Pathophysiologically, this elevated risk may be attributed to reduced nutritional reserves and dysregulation of the immune-inflammatory system. However, the fracture risk assessment tool (FRAX) often underestimates the 10-year probability of future fractures, as it does not fully account for these pathophysiological factors in this high-risk population. This study aims to determine whether the hemoglobin, albumin, lymphocyte, and platelet (HALP) score serves as an effective prognostic marker for a real-world fragility fracture endpoint, and whether incorporating the HALP score into FRAX improves the prediction of the 10-year probability of fragility fractures in this high-risk population.

This ambispective longitudinal cohort study utilized multicenter retrospective cohort data collected between January 1, 2014, and January 1, 2016 (Phase 1), followed by prospective follow-up for fragility fracture endpoints from January 1, 2016, to January 1, 2021 (Phase 2). Follow-up was conducted via outpatient services and telephone interviews every six months, with a median follow-up duration of 46.8 months. A total of 423 participants were included in the present analysis. Based on HALP score tertiles, participants were divided into low (≤ 31.98, n = 141), moderate (31.98-48.08, n = 140), and high (≥ 48.08, n = 142) groups.

Among the 423 participants, 75 (17.73%) experienced real-world fragility fractures. Spearman's partial correlation analysis revealed a significant inverse association between the HALP score and the FRAX‑estimated 10‑year probability of hip fracture (FRAX‑HF) (r = - 0.103, p = 0.037), but no significant correlations with bone mineral density (BMD) and the FRAX‑estimated 10‑year probability of major osteoporotic fracture (FRAX‑MOF). Multivariate linear regression further quantified that each one‑tertile decrease in HALP (tertile interval = 16.1) was associated with an approximate 0.31% increase in FRAX‑HF probability (calculated as 0.019 × 16.1 = 0.3059) (β = -0.019; 95% confidence interval [CI]: -0.034 to - 0.005; p = 0.007). Notably, in the lowest HALP tertile, the FRAX‑HF increased from the overall baseline of 2.30% (moderate risk, requiring only lifestyle management) to 3.00% (high risk, warranting anti‑osteoporosis therapy), representing a relative increase of 30.43%. Consequently, all non‑osteoporotic participants in this tertile, whose initial hip fracture probability (2.3%) was below the guideline-recommended intervention threshold of 3%, met criteria for initiating anti‑osteoporosis treatment, resulting in an intervention rate of 27.13% (35/129). These findings support the utility of combining the HALP score with FRAX for risk stratification to refine hip fracture prediction. Restricted cubic spline (RCS) analysis indicated dose‑dependent nonlinear relationships between HALP and overall fragility fracture endpoint events (Z‑shaped, overall p = 0.0011), as well as between HALP and FRAX‑HF (L‑shaped, overall p < 0.001). Multivariable Cox regression confirmed a significant inverse trend between HALP and the real‑world fragility fracture endpoint (p for trend = 0.002), and subgroup stratified analyses demonstrated stable predictive value across strata (all p for interaction > 0.05). Kaplan-Meier analysis illustrated a higher cumulative fracture incidence associated with lower HALP scores (log‑rank test, p = 0.00037). Receiver operating characteristic (ROC) curve analysis identified an optimal HALP cut‑off of 46.77 for fragility fractures, with an area under the curve (AUC) of 0.643 and a sensitivity of 84.0% (p < 0.001).

Our findings suggest that the HALP score may serve as an effective prognostic indicator for fragility fractures. Furthermore, the HALP score-adjusted FRAXplus tool (integrating the HALP score into FRAX) makes more elderly female patients with T2DM whose initial hip fracture probability falls below the guideline-recommended intervention threshold eligible for anti-osteoporosis therapy, thereby helping to reduce future 10-year hip fracture events.

Not applicable.