Extreme heat was associated with increased acute-care presentations for renal and fluid-balance disorders, dermatologic conditions, multiple sclerosis, cannabinoid-related disorders, and a range of injury-related diagnoses among adults in the Chicago area, according to a study published in Science Advances.
Researchers analyzed 916,904 emergency department and urgent care encounters among 372,140 adults across 4 Chicago-area health care systems from May through September between 2011 to 2023. Eligible patients were aged 18 years or older, had geocoded residences within Cook County, and had valid diagnostic codes. Because the study spanned the US transition from International Classification of Disease, Ninth Revision (ICD-9) to ICD-10, investigators mapped earlier codes to ICD-10, yielding 1,803 distinct ICD-10 diagnosis categories.
Daily maximum temperatures were linked to patients’ residential census tracts. Extreme heat was defined as the warm-season 95th percentile of daily maximum temperature, at least 33.67 °C.
Researchers used a 2-stage analytic framework. Stage 1 used diagnosis-specific quasi-Poisson models of daily visit counts and the 3-day moving average of maximum temperature, with adjustment for calendar time, day of week, federal holidays, and same-day Cook County particulate matter measuring 2.5 μm or less. Forty-four of 1,803 diagnosis categories met the prespecified screening criteria. The researchers described stage 1 as an aggregate-level ecological screening step used for dimension reduction rather than individual-level inference.
In stage 2, researchers evaluated the 44 diagnoses using distributed lag nonlinear models within a time-stratified case-crossover design. Each acute-care visit served as a case day, with control days matched by day of the week within the same calendar month and year. The models characterized nonlinear temperature-response associations over lags of 0 to 3 days and adjusted for same-day Cook County particulate matter. Thirty-three diagnoses met the numerical reporting criteria, including 10 disease-related diagnoses and 23 injury, poisoning, or external-cause diagnoses.
Compared with the diagnosis-specific minimum risk temperature, 33.67 °C was associated with approximately 11 times the cumulative odds of glomerular disorders in diseases classified elsewhere and approximately 4 times the cumulative odds of varicose veins of the lower extremities over 0 to 3 days. The same temperature was associated with more than 3 times the cumulative odds of edema, pressure ulcer, and acute kidney failure over 0 to 3 days. Associations were also identified for nonspecific skin eruption, volume depletion, symptoms involving food and fluid intake, multiple sclerosis, and cannabinoid-related mental and behavioral disorders.
“Among disease-related diagnoses, renal and fluid-balance signals—volume depletion, acute kidney failure, edema, and glomerular disorders—were prominent, consistent with prior evidence of dehydration and renal stress during heat exposure,” wrote lead author Hyojung Jang of the Division of Biostatistics and Informatics, Department of Preventive Medicine, Feinberg School of Medicine, Northwestern University in Chicago, Illinois, and colleagues.
Among injury, poisoning, and external-cause diagnoses, the highest cumulative estimates were reported for assault by other specified means, accidental firearm discharge, lower-limb burns and corrosion, and open wounds of the upper extremities. Other diagnoses meeting the reporting criteria included superficial injuries, fractures, sprains, object-related injuries, contact with sharp glass, nonvenomous insect bites or stings, and traffic accidents. The investigators cautioned that estimates for uncommon external-cause diagnoses were imprecise.
Sensitivity analyses produced different reporting sets when researchers modified some stage 1 screening criteria. When researchers restricted both analytic stages to June through August, volume depletion was the only diagnosis retained. They also noted that several findings from the May-to-September analysis, particularly those involving injuries and external causes, were sensitive to the seasonal exposure window and should be viewed as exploratory.
The study was limited to adults using 4 health care systems in Chicago. The electronic health record cohort also overrepresented Black and West Side patients and underrepresented White and Northwest Side patients, which the investigators noted may limit generalizability. Residential ambient temperature may not have reflected patients' actual heat exposure outside the home or in indoor and workplace settings. Socioeconomic variables were not included, raising the potential for residual confounding. The study also lacked external validation because both analytic stages used the same underlying cohort. As a result, the researchers noted, the second-stage estimates could not provide confirmatory inference.
The study was supported by an American Heart Association Predoctoral Fellowship and Northwestern University’s Buffett Institute for Global Affairs Defusing Disasters working group. Researchers reported no competing interests.
Source: Science Advances
