Background: The relationship between bone metabolism and plasma sodium levels has lately gained increasing interest as hyponatraemia has been linked to both increased risk of osteoporosis and fractures. The aim of this study was examine the frequency of hypo- and hypernatraemia in patients admitted with a fractured hip and the association with 30-day mortality in these patients.
Methods: A database of all surgically treated hip fracture patients admitted to our hospital between January 1996 and November 2013 was searched for all patients aged 60 years or above. 7755 patients were identified and a search for plasma sodium levels for these patients was conducted in the hospitals laboratory system. 7644 (98.6%) had a preoperative admission plasma sodium measurement and composed the study cohort. Comorbidity was included in the form of Charlson Comorbidity Index, which was calculated based on data from the Danish National Patient Registry. Hyponatraemia was defined as [Na+]<135 mmol/l and hypernatraemia as [Na+]>145 mmol/l.
Results: The patients had a mean age of 82.5 (S.D. 8.5) years and 76.5% (5845/7644) were female. 19.0% (1455/7644) were hyponatraemic, 1.6% (123/7644) were hypernatraemic and 79.4% (6066/7644) were normonatraemic on admission. There was an increased 30-day mortality rate for patients with hyponatraemia (12.1%, P=0.008 (χ2)) and hypernatraemia (16.3%, P=0.02 (χ2)) compared to normonatraemic patients (9.7%). The hazard ratios for 30-day mortality were 1.26 (1.06; 1.49) (unadjusted) and 1.35 (1.14; 1.60) (adjusted for sex, age and comorbidity) for hyponatraemic patients and 1.74 (1.12;2.72) (unadjusted) and 1.76 (1.13;2.78) (adjusted for sex, age and comorbidity) for hypernatraemic patients.
Conclusion: The study showed that the prevalence of hyponatraemia in hip fracture patients was high. Furthermore, patients with decreased or elevated plasma sodium levels had an increased mortality rate. Disturbances in plasma sodium levels may itself cause increased mortality but could also be a surrogate marker for frailty in these patients.
Disclosure: The authors declared no competing interests.