First fracture increases subsequent fracture risk in postmenopausal women

In postmenopausal women, a fracture carries an increased risk for a subsequent fracture, regardless of location of initial fracture or patient age, according to a recent study from the US.

“[A]fter adjustment for age, race/ethnicity, BMI, hormone therapy use, and other covariates, every type of initial fracture … was associated with significantly increased risk of subsequent fracture,” said the authors.

“Clinicians should be aware that initial fractures of any type in postmenopausal women, even at sites other than the hip, vertebra, or wrist, should trigger counselling regarding increased subsequent fracture risk,” they said.

Using October 1993 to March 2018 data from the Women’s Health Initiative, the researchers prospectively analysed the incidence of subsequent fracture after an initial fracture among 157,282 postmenopausal women (age 50–79 years at baseline; mean age 63.1 years, mean BMI 28.0 kg/m², 83 percent White). Women on bisphosphonates, calcitonin, and/or raloxifene at baseline were excluded. The mean follow-up period was 15.4 years, during which time 47,126 women experienced incident initial fractures.

Women with an incident lower arm/wrist fracture had a significantly increased risk of a subsequent fracture of the upper arm/shoulder, upper leg, knee, lower leg/ankle, hip/pelvis, and vertebra (hazard ratios [HRs], 3.30–6.46). This risk of subsequent fracture was highest for lower leg/ankle fracture. Compared with women who did not sustain an initial lower arm/wrist fracture, those who did had a 5.68-fold increased risk of lower leg/ankle fracture. [EClinicalMedicine 2021;35:100826]

Women with an initial upper arm/shoulder or upper leg fracture had a significantly increased risks for subsequent fracture of the hip/pelvis (HRs, 5.06 and 5.11, respectively). The risk of subsequent fracture of the hip/pelvis was also increased among women with an initial lower arm/wrist fracture (HR, 4.80), initial knee fracture (HR, 5.03), initial lower leg/ankle fracture (HR, 4.10), or initial vertebra fracture (HR, 6.69).

Those with an initial hip/pelvis fracture had a 27-fold increased risk of subsequent upper leg (non-hip) fracture. Thirty-four percent of women who experienced an initial hip/pelvis fracture experienced a subsequent non-hip fracture.

Subsequent fracture risk varied by age (50–59, 60–69, and 70–79 years at baseline) for initial fractures of the lower arm/wrist, upper arm/shoulder, upper leg, knee, lower leg, and vertebra (p_interaction<0.0001 to 0.08). However, the risk of any subsequent (non-hip/pelvis) fracture following an initial hip/pelvis fracture did not significantly differ according to age (p_interaction=0.52).

There was a pattern of higher point estimates in the youngest age group (age 50–59 years).

“Within each age group, and after each type of initial fracture, the risk of subsequent fracture was significantly higher among women with initial fracture than among women without initial fracture,” said the authors.

“Women of all ages … who have initial fracture should be counselled that they are at elevated risk of subsequent fracture,” they continued.

The risks for subsequent fracture also varied by ethnicity (p_interaction<0.0001 to 0.002). The risk for subsequent fracture after initial lower or upper extremity fracture was higher among non-Hispanic Black, Hispanic/Latina, and Asian/Pacific Islander women vs non-Hispanic White women.

Implications of the outcomes

According to the authors, the results have important clinical implications including a potential reassessing of guidelines. For example, some guidelines recommend initiation of medical treatment following a hip or vertebral fracture, though not necessarily following other types of fracture.

“It will be important to determine whether existing risk calculators can be adapted (or new calculators developed) to help refine decision-making to determine which of the women with fractures other than hip or vertebral fractures should be treated,” they said.

“Our findings will inform Fracture Liaison Service Programs, which are specifically targeted to preventing subsequent fractures after an initial fracture,” they added, calling for further study into interventions to prevent subsequent fractures in women with non-hip or non-vertebral fractures.

They noted that the inclusion of only the initial fracture at a particular site was a limitation. “For example, if a participant had a lower arm/wrist fracture, a subsequent lower arm/wrist fracture would not be captured,” they noted. As such, the results may have underestimated the associations between initial and subsequent fractures.