New data show COVID-19 triggers spike in new diabetes cases

By Tarun Sai LomteReviewed by Susha Cheriyedath, M.Sc.Jun 18 2025

New research reveals that people infected with COVID-19 in California prisons face a lasting, higher risk of diabetes, even months after recovery, underscoring the urgent need for targeted healthcare strategies in vulnerable populations.

Study: COVID-19 Increases the Rate of Incident Diabetes: A Case-Control Cohort Time-to-Event Study. Image Credit: Anatoliy Cherkas / Shutterstock

*Important notice: medRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.

A recent study posted to the medRxiv preprint* server reported that coronavirus disease 2019 (COVID-19) increases the rate of incident diabetes mellitus (DM). Most COVID-19 cases have not been fatal worldwide, albeit many involved symptomatic illness, hospitalization, and significant comorbidity.

The long-term effects of COVID-19 are poorly understood. Long COVID refers to a diverse range of conditions and symptoms lingering for over three months after acute COVID-19. Studies suggest an increase in the incidence of chronic conditions post-COVID-19, including neurological and cardiovascular conditions and DM.

However, the evidence base for increased DM incidence post-COVID-19 has various limitations. That is, despite the high rates of both DM and COVID-19 in the United States, minorities, especially non-Hispanic Black and Hispanic populations, are underrepresented in many studies, limiting generalizability. Further, some studies were not designed to exclude and ascertain preexisting DM or confounding diagnoses or medications that may increase blood glucose.

About the study

In the present study, researchers in California, U.S., examined the extent to which the first COVID-19 diagnosis increases DM risk. A key strength of this research is its large, racially and ethnically diverse population, which addresses a significant limitation in prior studies. First, they identified individuals who were continuously incarcerated in 31 California state prisons between January 2019 and February 2020 (pre-pandemic period) and who were at least incarcerated for one day during the COVID-19 pandemic (beginning March 2020).

Individuals with prior DM, history of confounding medication/diagnosis, those lacking measurements of body mass index and blood glucose before the pandemic, and those without COVID-19 and DM testing during the pandemic were excluded. The primary outcome was the time to incident DM, and the exposure was severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. All participants contributed a non-exposure observation time.

Subjects with a positive COVID-19 test before DM diagnosis started exposure time at day 31 after the positive sample collection date. Censoring events were death, release from prison, positive COVID-19 test, confounding medication/diagnosis, and termination of observation (March 1, 2023). Cox proportional hazards models assessed the relationship between COVID-19 and incident DM, adjusted for gender, age, ethnicity, baseline clinical characteristics, and healthcare system use patterns.

Sensitivity analyses were performed to evaluate the robustness of estimates by modifying the definition of the exposure period to 0, 61, and 91 days after testing SARS-CoV-2-positive. Additional analyses were performed to determine whether variations in DM diagnostic intensity following a positive COVID-19 test would explain the effect of SARS-CoV-2 on the DM incidence rate.

Findings

The Figure shows the predicted cumulative hazard of incident diabetes mellitus over time (lines) and associated 95% bootstrapped confidence intervals (shaded regions) of these predictions under two scenarios: a) no one in the study cohort was infected with COVID-19 (dashed line, blue shaded region); b) everyone in the study cohort was infected with COVID-19 (solid line, red shaded region).

A total of 29,470 individuals were included in the analyses. Of these, 58.3% tested positive for SARS-CoV-2 before the diagnosis of DM or censoring. These subjects contributed to over 15.8 million days and 8.9 million days of pre-exposure and exposure periods, during which 690 and 429 individuals were diagnosed with DM, respectively. This corresponded to crude DM rates of 17.6 and 15.9 per 1,000 person-years in COVID-19 and uninfected populations, respectively.

The COVID-19 population was older and had lower testing rates than the overall population. COVID-19 was associated with a significantly increased rate of incident DM, with a hazard rate ratio of 1.17. COVID-19 increased the cumulative hazard of DM at one-year and two-year follow-up periods.

Specifically, the study projected that if the entire cohort had been infected with COVID-19, the 2-year cumulative risk of a new diabetes diagnosis would have been 3.2%, compared to 2.7% if none had been infected. Modifying the exposure period definition to 61 or 91 days after a positive test yielded consistent findings.

However, while the magnitude and direction were similar for the zero-day definition, the association was no longer significant. Besides, while modifying the definition of the exposure period changed the number of people with COVID-19, the crude DM rates with COVID-19 were still greater than 17 per 1,000 person-years. Further, confounding due to variations in DM diagnostic intensity post-COVID-19 implied that the effect would be halved, with an estimated adjusted HRR between 1.08 and 1.10.

Conclusions

In summary, the findings suggest that the first COVID-19 infection is associated with an increased rate of incident diabetes mellitus. This increase in incident DM due to COVID-19 lingers for months, even after excluding the three months after acute infection.

These findings have specific health policy implications and raise questions for future research. The authors suggest that clinical systems providing care to incarcerated populations should be aware of the elevated DM risk post-COVID and emphasize the importance of linking individuals to health insurance and accessible healthcare upon their release. It will be crucial to determine whether SARS-CoV-2 variants have similar effects on the incidence of DM and whether prior infection and vaccination may modify the elevated DM incidence.

The authors also highlight important caveats regarding the study's setting. Generalizing findings from a carceral population to the non-incarcerated population is not straightforward, as incarcerated individuals experience unique stressors, different nutritional standards, and higher rates of premature aging.

Furthermore, the impact of the pandemic was more severe in prisons, with infection and mortality rates 3-5 times higher than in the general community. These factors may indicate that the magnitude of COVID-19's impact on diabetes risk may differ in other populations.

*Important notice: medRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.