Adults hospitalized with influenza in England had higher rates of newly diagnosed diabetes mellitus following discharge compared with matched community controls and patients hospitalized with sepsis, according to a retrospective cohort study published in BMJ Open.
Researchers analyzed linked electronic health records from the Clinical Practice Research Datalink and Hospital Episode Statistics databases in England. The study included 13,710 adults hospitalized with a first primary diagnosis of influenza between 2004 and 2021 who had no preexisting diabetes mellitus and no diabetes diagnosis during hospitalization. Outcomes were compared with those of 61,384 nonhospitalized controls matched by age, sex, and general practitioner practice, as well as 38,561 patients hospitalized with sepsis.
Follow-up began 1 day following discharge and continued until diabetes diagnosis, death, transfer out of practice, last practice data collection, or the end of available records. The primary outcome was incident diabetes mellitus. Researchers also assessed incident prediabetes, timing of diagnosis, testing frequency, medication use, and factors associated with postdischarge diabetes risk.
The incidence of diabetes following influenza hospitalization was 12.5 cases per 1,000 person-years compared with 6.5 cases per 1,000 person-years among matched controls and 11.7 cases per 1,000 person-years among patients hospitalized with sepsis. After adjustment for confounding factors, influenza hospitalization was associated with a 54% higher adjusted rate of diabetes diagnosis compared with matched controls and a modestly higher adjusted rate compared with sepsis hospitalization.
The increased risk was most pronounced within 90 days following discharge. Compared with matched controls, patients hospitalized with influenza had a 2.7-fold higher adjusted rate of diabetes diagnosis during that period. However, compared with patients hospitalized with sepsis, the adjusted hazard ratio during the same interval was 1.14, suggesting a numerically higher rate that was not statistically significant.
Researchers also observed higher rates of prediabetes following influenza hospitalization than in both comparator groups. After excluding patients with preexisting prediabetes, incident prediabetes occurred at a rate of 16.6 cases per 1,000 person-years in the influenza cohort compared with 11.3 cases per 1,000 person-years among matched controls and 14.7 cases per 1,000 person-years among patients hospitalized with sepsis. Adjusted analyses showed significantly higher rates of prediabetes diagnosis following influenza hospitalization compared with both comparator groups.
Several patient characteristics were associated with higher postdischarge diabetes risk, including male sex, older age, higher body mass index, hypertension, preexisting diabetes risk, oral corticosteroid prescriptions, and critical care admission during influenza hospitalization.
Among 1,666 influenza-hospitalized patients with preexisting prediabetes, 215 subsequently developed diabetes mellitus, accounting for 35% of all diabetes diagnoses in the influenza cohort. These patients had nearly fourfold higher adjusted rates of developing diabetes compared with patients without preexisting prediabetes.
The researchers also evaluated whether more intensive postdischarge testing could have contributed to higher diagnosis rates. In the first year following discharge, 35% of influenza-hospitalized patients underwent at least 1 glucose or diabetes-related test compared with 29% of matched controls and 27% of sepsis comparators. However, sensitivity analyses stratified by testing frequency still demonstrated higher diabetes diagnosis rates among influenza patients compared with matched controls. Within testing strata, diabetes rates did not differ significantly between influenza and sepsis cohorts.
Among the 610 influenza-hospitalized patients diagnosed with diabetes during follow-up, 69% subsequently required oral diabetes medications or insulin therapy, suggesting that persistent dysglycemia may extend beyond transient stress hyperglycemia associated with acute illness.
The researchers noted that acute respiratory infections may contribute to diabetes development through persistent inflammatory-driven insulin resistance and potential pancreatic beta-cell injury, although the mechanisms remain incompletely understood in humans. Prior animal studies have demonstrated impaired insulin signaling and pancreatic damage following influenza infection.
The study had several limitations, including its observational design, reliance on coding accuracy within electronic health records, and inability to assess inpatient management factors such as corticosteroid administration during hospitalization. The sepsis comparator cohort was also unmatched and differed substantially from the influenza cohort in age, sex distribution, and hospital length of stay. In addition, the study period ended in March 2021, which may limit generalizability to more recent influenza seasons and postpandemic patient populations.
Only 12 cases of type 1 diabetes were identified following influenza hospitalization, preventing subgroup analyses by diabetes type.
“The longer term effects of influenza infection strengthen the case for better influenza prevention and surveillance to identify those with new onset [diabetes mellitus],” wrote lead study author Sophie Middleton, of the University of Nottingham School of Medicine, and colleagues.
Disclosures: The study was funded by the National Institute for Health and Care Research Nottingham Biomedical Research Centre. Sophie Middleton, Frances S. Grudzinska, Tricia M. McKeever, and Charlotte E. Bolton were supported by the center. Bolton reported receiving research grant funding related to recovery from acute respiratory infection from the NIHR, University of Nottingham charitable donations, and Nottingham Hospitals Charity.
Source: BMJ Open