# Association of a polygenic risk score with low trauma fractures in people with HIV – The swiss HIV cohort study

## TL;DR

This study shows that a genetic risk score for bone density is linked to an increased risk of fractures in people with HIV, even after accounting for other risk factors.

## Contribution

The study demonstrates the independent association of a polygenic risk score with low trauma fractures in people with HIV.

## Key findings

- Participants in the top PRS quintile had a 2.30-fold increased risk of low trauma fractures.
- The PRS remained significant after adjusting for traditional and HIV-related risk factors.
- Non-genetic risk factors also showed a strong association with fracture risk.

## Abstract

Polygenic risk scores (PRS) are likely to enter routine clinical care for individual disease risk prediction in the next 10 years. We recently showed that the bone mineral density-associated gSOS-PRS is independently associated with a > 4-fold increased risk of osteoporosis in the Swiss HIV Cohort Study (SHCS). Here we investigate whether this PRS is also associated with low trauma fractures (LTF) in people with HIV in the SHCS.

Applying a case-control design, cases had a first LTF (1994–2022) and LTF-free controls were matched on age, sex and observation time. We obtained univariable odds ratios (OR) for LTF in SHCS participants of European descent, based on a genome-wide PRS built from 9413 LTF-associated single nucleotide polymorphisms (SNPs). In multivariable analysis, LTF odds ratios of the PRS were adjusted for non-genetic (traditional and HIV-related) LTF risk factor profile including potentially adverse antiretroviral exposures.

We included 277 SHCS participants with a first LTF (cases) and 796 LTF-free controls (median age 55 years; 68% male; 91% with suppressed HIV RNA). Participants with the most unfavorable genetic background (top quintile of the gSOS-PRS) had univariable and multivariable LTF-OR of 2.30 (95% confidence interval, 1.49–3.56) and 2.30 (1.43–3.72), respectively, compared to participants with the most favorable genetic background (bottom gSOS-PRS quintile). Participants with the most unfavorable non-genetic risk factor profile (top quintile) had an adjusted gSOS-OR of 7.42 (95% confidence interval [CI], 4.3–12.82), compared with participants in the bottom quintile.

In people with HIV in Switzerland, an unfavorable PRS was independently associated with LTF risk after adjustment for traditional and HIV-related LTF risk factors.