Cause-specific excess mortality after hip fracture: the Norwegian Epidemiologic Osteoporosis Studies (NOREPOS)

Abstract

Background: Information on cause of death may help appraise the degree to which the high excess mortality after hip fracture reflects pre-existing comorbidities or the injury itself. We aimed to describe causes of death and cause-specific excess mortality through the first year after hip fracture.

Methods: For studying the distribution of causes of death by time after hip fracture, we calculated age-adjusted cause-specific mortality at 1, 3, 6 and 12 months in patients hospitalized with hip fracture in Norway 1999-2016. Underlying causes of death were obtained from the Norwegian Cause of Death Registry and grouped by the European Shortlist for Causes of Death. For estimating excess mortality, we performed flexible parametric survival analyses comparing mortality hazard in patients with hip fracture (2002-2017) with that of age- and sex matched controls drawn from the Population and Housing Census 2001.

Results: Of 146,132 Norwegians with a first hip fracture, a total of 35,498 (24.3%) died within one year. By 30 days post-fracture, external causes (mainly the fall causing the fracture) were the underlying cause for 53.8% of deaths, followed by circulatory diseases (19.8%), neoplasms (9.4%), respiratory diseases (5.7%), mental and behavioural disorders (2.0%) and diseases of the nervous system (1.3%). By one-year post-fracture, external causes and circulatory diseases together accounted for approximately half of deaths (26.1% and 27.0%, respectively). In the period 2002-2017, cause-specific one-year relative mortality hazard in hip fracture patients vs. population controls ranged from 1.5 for circulatory diseases to 2.5 for diseases of the nervous system in women, and correspondingly, from 2.4 to 5.3 in men.

Conclusions: Hip fractures entail high excess mortality from all major causes of death. However, the traumatic injury of a hip fracture is the most frequently reported underlying cause of death among older patients who survive less than one year after their fracture.

Keywords: Causes of death; Excess mortality; Hip fracture; Norway; Registry-based epidemiology.

Fig. 1

Age- and sex adjusted cause-specific mortality (%) by time after hip fracture. Norwegians 50 years and older, men and women combined (n = 146,132). Predicted at mean values of age at first hip fracture (80.7 years) and sex (30.3% men) from a logistic model (general linear model with ‘family = binomial’ and ‘link = logit’) regressing cause-specific death on age and sex. External: External causes of morbidity and mortality (ICD-10: V01-Y89); Circulatory: Diseases of the circulatory system (ICD-10: I00-I99); Neoplasms: Neoplasms (ICD-10: C00-D48); Respiratory: Diseases of the respiratory system (ICD-10: J00-J99); Mental/behavioral: Mental and behavioural disorders (ICD-10: F01-F99); Nervous system: Diseases of the nervous system and the sense organs (ICD-10: G00-H95); Other: Deaths caused by digestive diseases, endocrine, nutritional and metabolic diseases, genitourinary diseases, infectious and parasitic diseases, musculoskeletal diseases, immunologic disorders, diseases of the skin or subcutaneous tissue, congenital malformations, and ill-defined/unspecified causes

Fig. 2

Distribution (%) of deaths after hip fracture by seven major cause of death categories. Grouped according to the European Shortlist for Causes of Death by 30 days (n = 11,322 deaths), 3 months (n = 20,256 deaths), 6 months (n = 26,772 deaths) and 1 year post-fracture (n = 35,498 deaths)

Fig. 3

All-cause and cause-specific excess mortality (HR with 95% CI) by months after hip fracture. Panel a: all-cause deaths; Panel b: deaths from exter­nal causes; Panel c: deaths from neoplasms; Panel d: deaths from diseases of the nervous system and the sense organs; Panel e: deaths from circulatory diseases; Panel f: deaths from mental and behavioural disorders; Panel g: deaths from diseases of the respiratory system; Panel h: Deaths from other causes