Methods
A
retrospective cohort design utilizing TriNetX, an electronic health
record database from 55 United States healthcare organizations with
>85 million patients. The study population consisted of adults aged
80 years and older who had healthcare encounters between January 1, 2018
and December 31, 2019, before the pandemic. The dehydration cohort was
identified using ICD-10 diagnosis codes and laboratory test results.
Outcome measures included one of three stroke types: intracerebral
hemorrhage, ischemic stroke, and transient ischemic attack, based on
ICD-10 codes. A sub analysis of individuals with diabetes was
undertaken.
Results
Of
3,125,610 adults, 80 and older, 563,476 were dehydrated. Individuals
with diabetes numbered 443,450 and 101,661 were dehydrated. The
dehydration cohorts in both populations had a greater percentage of
females, non-Hispanic, white individuals, and were slightly older (82.7
vs. 82.4, p <0.001). After controlling for common confounders in
propensity score matching, individuals with dehydration were 1.98-3.99
times more likely to develop stroke: intracerebral hemorrhage (OR=3.99,
95% CI=3.41-4.67), ischemic stroke (OR=1.98, 95% CI=1.9-2.07), and TIA
(OR=2.88, 95% CI=2.74-3.28). The diabetes group showed: intracerebral
hemorrhage (OR=6.76, 95% CI=4-11.42), ischemic stroke (OR=1.97, 95%
CI=1.81-2.16), and TIA (OR=2.81, 95% CI=2.33-3.39).
Conclusion
A
strong association between dehydration and stroke was found. The
largest association was with intracerebral hemorrhage. Both physiologic
changes of normal aging and medications used to treat cardiovascular
stroke risk factors increase older adults’ risk for dehydration. The
strength of this study is the propensity risk management of over 3
million older adults (≥80 years). Limitations of this study include the
retrospective nature of database evaluation. Future studies should
evaluate whether increased hydration status leads to decreased stroke.
No comments:
Post a Comment