This study systematically reviews prospective and retrospective cohort studies evaluating the risk of hip fracture following stroke. Stroke survivors are at high risk of hip fracture and had a 1.5-fold increased risk compared to stroke-free men and women of the same age. Hip fracture often occurs in ageing and female stroke patients. We performed a meta-analysis to summarize evidence from prospective and retrospective cohort studies about the risk of hip fracture following stroke. We identified English and non-English publications in MEDLINE and EMBASE using stroke and fracture as keywords to 31 December 2015. The data of the incidence of hip fracture were extracted to calculate raw incident rate in stroke survivors as well as risk of hip fractures in strokes comparing populations using a random-effects model. Subgroup analyses were performed to identify the potential influence of some factors. Six prospective and seven retrospective cohort studies were included, involving 512,214 stroke patients with 22,559 hip fractures. The pooled prevalence of hip fractures was 4.87 % (95 % CI, 4.05 to 5.68 %) in stoke patients. We conducted subgroup analysis according to sex, age, duration of follow-up, study design, and region, which showed that female (vs. male) stroke patients older than 70 years (vs. those less than 70 years) and duration of follow-up more than 2 years (vs. those less than 2 years) have higher proportions of hip fractures. Four studies showed that stroke patients had a significantly higher risk of hip fracture compared with the general population, while the other study had a non-significant higher risk. The overall prevalence of hip fracture was 3.28 % (3431 of 104,646) in stroke patients and 2.83 % (36,493 of 1,287,726) in general population, respectively, and the unadjusted combined relative risk of hip fracture was 1.54 (95 % CI, 1.06-2.25). Hip fractures often occur in ageing and female stroke patients.
Keywords: Hip fracture; Meta-analysis; Risk; Stroke.