Hip fracture prevention in Swedish 70-100 year old women 2001-2022: risk factors for mortality in intervention and control groups

Hans Thulesius, Lisa Alvunger, Anna Segernäs, Anna Lindgren, Robert Eggertsen, Pär Wanby, Daniel Albertsson

Keywords: Mortality risk factors, older women, hip fracture prevention, longitudinal cohort

Background:

Hip fractures cause considerable morbidity and mortality. A prevention intervention study in primary care starting in 2001 focused on physical exercise, fall prevention at home, and pharmacological osteoporosis treatment.

Research questions:

To investigate associations between risk factors for mortality 21 years after a controlled hip fracture prevention intervention.

Method:

Prospective population-based cohort with 1247 women aged 70-100 years in three rural primary care districts in Sweden. Fracture-related risk factors collected in baseline survey. 435 women recruited to hip fracture reduction intervention and 813 women age matched controls. Risk factor models defined at baseline based on age ≥80 years, weight <60 kg, previous fragility fracture and either fall last year or inability to rise up five times from a chair without the use of arms. Associations between 11 specific risk factors and mortality and intervention group allocation were studied with Cox regression proportional hazard models.

Results:

Of 1247 participants at baseline 2001, 152 were alive January 2022 with oldest woman 105 years.
Cox regression proportional hazard models showed overall hazard ratios for mortality of 1.8 (95% confidence interval (CI) 1.6-2.1, p<0.001) for inability to rise up five times from a chair and poor self rated health of 1.6 (95% CI 1.3-1.9, p<0.001). Participation in the intervention group showed mortality hazard ratio of 0.9 (95% CI 0.7-1.0, p=0.05) only for women who were able to rise up five times from a chair without the use of arms.

Conclusions:

Age-adjusted inability to rise up five times from a chair without the use of arms and poor self rated health were the strongest factors predicting mortality in this 21-year follow up of a hip fracture prevention study. Participation in intervention group showed borderline significance for lower mortality only for women who were able to rise up five times from a chair without the use of arms.

Points for discussion:

Do you think it would be useful to know more about your older patients’ muscular capacity if it proves to be a marker for earlier death?

Why is it that poor self rated health has shown association with higher mortality in many studies?

#60