Effect of SARS-CoV-2 Infection and Infection Severity on Longer-Term Glycemic Control and Weight in People With Type 2 Diabetes

Diabetes is a commonly reported comorbidity among patients with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection, with estimates ranging from 10.9–58.0% in the U.S. (1). As we approach the third year of the coronavirus disease 2019 (COVID-19) pandemic, numerous studies have examined the effect of the pandemic on people with diabetes, from decreased diagnosis, treatment with medication, and monitoring of HbA_1c (2) to the effect of lockdowns on metabolic outcomes such as worsened glycemic control in people with type 2 diabetes (T2D) and weight gain (3–6). Recent studies have also demonstrated a diabetogenic effect of SARS-CoV-2 infection, including increased incidence of new-onset diabetes and post–COVID-19 hyperglycemia (7–11). Several studies analyzing the prevalence of new-onset diabetes after COVID-19 demonstrated a pooled prevalence of 14.4–19.7% (11,12). Proposed pathophysiological theories on the diabetogenic and hyperglycemic effects of COVID-19 suggest that the inflammatory nature of the virus may be related to insulin resistance (13). In addition, some studies also have shown a direct effect of SARS-CoV-2 on insulin-producing pancreatic islet cells (8,9). Although acute COVID-19 has been associated with hyperglycemia, there are limited data on the effect of COVID-19 on longer-term glycemic outcomes; there is only one study with a subgroup of 49 patients with diabetes who showed no change in HbA_1c in patients at a median follow up of 215 days (14). Overall, most studies on the association of COVID-19 and hyperglycemia have focused on acute glycemic changes during active SARS-CoV-2 infection (13), and there is limited evidence on the longer-term effects after infection.

Similarly, evaluation of the effect of COVID-19 on weight and nutrition has been limited to small studies with short-term follow-up after COVID-19 infection. Several studies of patients with COVID-19 have shown unintended weight loss in a high percentage of patients either during hospitalization or at post-COVID follow-up, with risk factors for malnutrition including male sex, intensive care admission, and longer disease duration (15,16). The follow-up period in these studies ranges from 3 weeks to 3 months after SARS-CoV-2 infection, and they were primarily cohort studies without comparison groups (17–19). Studies addressing longer-term effects of SARS-CoV-2 infection and the severity of infection on glycemic control and weight are limited.

As we transition from pandemic to endemic COVID-19, it is important to understand the longer-term effects that SARS-CoV-2 infection has on glycemic control and weight in patients with diabetes to effectively individualize chronic care. The aim of this study is to evaluate the effects of SARS-CoV-2 infection on longer-term glycemic control and weight. Using the National COVID Cohort Collaborative (N3C) database, we evaluate the association of SARS-CoV-2 infection and severity of infection on average HbA_1c and average weight to analyze longer-term metabolic changes due to COVID-19.

We conducted a retrospective cohort study using data from the N3C, which includes electronic health record data from multiple health care systems in the U.S. that contribute to the N3C Data Enclave (20). The adult and pediatric population in the N3C has previously been characterized (21,22). We defined the index date as the date of first COVID-19 diagnosis by either positive SARS-CoV-2 PCR test or ICD-10 code for positive patients or the date of the first negative SARS-CoV-2 PCR test for control subjects. Our study population included adults age ≥18 years with an ICD-10 diagnosis of T2D prior to the index date, with at least one HbA_1c measurement 1 year prior to or up to 7 days after the index date and one HbA_1c measurement >90 days after the index date, and with at least one outpatient diabetes medication prescribed prior to the index date. The control population, defined as patients who had a negative SARS-CoV-2 PCR test and did not have any diagnosis of COVID-19 by either SARS-CoV-2 PCR test or ICD-10 diagnosis, was a comparable group of patients who would seek and have access to medical care and for whom the time period overlapped with the COVID-19 pandemic (23). We included both ambulatory and hospitalized patients in the study. Data access and analysis were conducted with the Foundry platform (2021) (Palantir, Denver, CO), a secure analytics enclave housing the N3C data. N3C data transfer to National Center for Advancing Translational Sciences is performed under Johns Hopkins University Reliance protocol no. IRB00249128 or individual site agreements with the National Institutes of Health. The N3C Data Enclave is managed under the authority of the National Institutes of Health; information can be found at https://ncats.nih.gov/n3c/resources. The Stony Brook University Office of Research Compliance determined that the study did not constitute human subjects research.

The primary outcomes of the study were the postindex average HbA_1c and weight among patients who did and did not have SARS-CoV-2 infection. The preindex average HbA_1c and preindex weight were defined as the average of all HbA_1c levels and weights within the 365 days before or 7 days after the index date. For evaluation of longer-term glycemic and weight changes that were not associated with the acute infection, the postindex average HbA_1c and weight were defined as the average of all HbA_1c levels and weights >90 days to 455 days after the index date. Secondary outcomes included postindex average HbA_1c and weight among the subgroup of patients diagnosed with COVID-19 who were hospitalized or who required invasive mechanical ventilation. Postindex averages were used to analyze metabolic changes as opposed to change scores to better estimate causal effects with our observational data (24).

For this study, hospitalization was defined as an inpatient visit with a start date up to 7 days prior to or 30 days after the index date and invasive mechanical ventilation was defined by the presence of invasive ventilation or extracorporeal membrane oxygenation (ECMO) during the hospital visit. Comorbidities were defined with the Charlson Comorbidity Index and included the following categories: myocardial infarction, congestive heart failure, peripheral vascular disease, stroke, dementia, pulmonary disease, mild and severe liver disease, renal disease, cancer, and HIV. BMI was grouped using the World Health Organization classification of BMI. Outpatient medications used to treat diabetes with a record within 90 days prior to the index date were included as covariates, including metformin, dipeptidyl peptidase 4 (DPP-4) inhibitors, glucagon-like peptide 1 analogs, sodium–glucose cotransporter 2 inhibitors, sulfonylureas, thiazolidinediones, and insulin. For exclusion of patients with only inpatient medications from the analysis, medication data from drug exposure tables were filtered to exclude medications associated with inpatient encounters or medications, which on manual review, would only be administered in a hospital setting. Concepts for T2D, HbA_1c, hospitalization, invasive mechanical ventilation, Charlson Comorbidity Index, and medications were either identified with use of concept sets from the N3C Knowledge Store or developed by the Diabetes and Obesity Domain Team (21,25–27). For data quality, data from N3C sites that reported a <1% rate of death, hospitalization, or ventilation and ECMO for patients with COVID-19 were excluded from the analysis.

Statistical analysis was conducted in the N3C enclave with use of Python 3.6 and R 3.5.1 with data from the beginning of the pandemic to the release date of 22 July 2021. Propensity score matching was done with the R package MatchIt (28). A one-to-one matching algorithm was performed to match control patients in the N3C with patients who had SARS-CoV-2 infection. The matching procedure included optimal pair matching with propensity scores used for the distance calculation, with matches drawn with replacement (29). Quality of matched cohorts was assessed through evaluation of standardized mean differences of baseline confounders and standardized mean differences of all pairwise interactions. Overlap in baseline characteristics of the cohorts was evaluated with use of overlap plots of the propensity scores (30). To study the effect of COVID-19 on postindex average HbA_1c and postindex average weight, we performed a linear regression using the matched COVID-19 and control groups. In addition to the point estimates, we evaluated the difference in the distribution of change in average HbA_1c and average weight. To test whether the difference in postindex average HbA_1c and weight between matched COVID-19 and control groups was smaller than a threshold (0.5% for HbA_1c and 2.5 kg for weight), we ran an equivalence test, with small P values indicating evidence of little difference (31). We further ran sensitivity analyses to test how sensitive the equivalence test findings were to unmeasured confounding. The sensitivity analyses determined the maximum amount of unmeasured confounding that could be present without nullifying any findings (32). For the subgroup of patients who had COVID-19 infection and for the control group, multivariable linear regression was performed separately for evaluation of the effect of disease severity on postindex HbA_1c and postindex average weight. The subgroup analysis models were fully adjusted for demographic covariates, BMI or preindex weight, comorbidities, and medications. To take into account the possibility of informative (i.e., differential) censoring of the outcome due to death or loss to follow-up, we conducted additional analyses based on inverse probability of censoring weighting (33), which reweights the study sample of those with uncensored outcomes to look like the study sample of all patients regardless of whether their outcomes were censored; it thus accounts for differential censoring due to death or loss to follow-up. P values <0.05 were considered statistically significant.

There were 6,003 COVID-19 patients and 6,003 matched control subjects across 14 sites who were eligible for inclusion in the analysis. The demographic and clinical characteristics of the cohort and control group are shown in Table 1. In the COVID-19 cohort, 52.7% were female and mean (SD) age was 61.7 (13.6) years. The cohort was 52.9% White, 21.2% Black or African American, and 25.5% Hispanic or Latino. The rates of hospitalization and invasive ventilation or ECMO were 38.2% (n = 2,293) and 3.2% (n = 194) respectively, with a mean (SD) length of stay (LOS) of 6.6 (SD 22.3) days. For patients in each group, the mean number of pre- and postindex HbA_1c measurements was 2.36 and 1.60, respectively, with a mean of 84 days between measurements in the COVID-19 group, and 2.25 and 1.60 with a mean of 93 days between measurements in the control group. The mean number of pre- and postindex weight measurements was 13.5 and 8.6, with a mean of 17 days between measurements in the COVID-19 group, and 12.7 and 8.8, with a mean of 20 days between measurements in the control group. The pre- and postindex mean (SD) HbA_1c measurements were 7.5% (1.8%) and 7.3% (1.7%) and pre- and postindex weight were 92.0 (24.4) kg and 90.7 (24.6) kg.

After propensity score matching, the standardized mean differences of baseline covariates and pairwise interactions between COVID-19 and control groups were <0.1, indicating high-quality matches (Supplementary Fig. 1). The largest standardized mean difference across all pairwise interactions of baseline variables between the COVID-19 and control groups was 0.024. The overlap in the COVID-19 and control groups was sufficient (Supplementary Fig. 2). For evaluation of the preindex average HbA_1c as a continuous variable, values were compared with empirical cumulative distribution function plots and showed good matching across the entire distribution of values (Supplementary Fig. 3).

The results from the linear regression comparing patients with COVID-19 with the control group did not show a statistically significant difference in the postindex average HbA_1c (coefficient 0.008, 95% CI −0.053 to 0.069, P = 0.807) or the postindex average weight (coefficient −0.65, 95% CI −1.526 to 0.227, P = 0.146). The results from the equivalence test indicated strong evidence that the difference in mean postindex average HbA_1c between the COVID-19 and control groups was <0.5% (P < 0.001) and the difference in mean postindex average weight between COVID-19 and control groups was <2.5 kg (P < 0.001). For the sensitivity analysis of equivalence test results to unmeasured confounding, the γ-value for postindex average HbA_1c was 2.195 with an α of 0.05 and a δ of 0.5; this implies that the equivalence test results would be unchanged were there unmeasured confounding strong enough to make COVID-19 patients 2.195 times more likely to have COVID-19 than their matched control pairs, indicating moderate to high robustness and insensitivity of findings to unmeasured confounding. For the sensitivity analysis for postindex average weight, the γ-value was 1.19 with an α of 0.05 and a δ of 2.5 kg, indicating moderate sensitivity of the results to unobserved confounding variables. The analysis accounting for informative censoring due to death or loss to follow-up did not show any statistically significant difference in postindex average HbA_1c (coefficient 0.06, 95% CI −0.05 to −1.6, P = 0.29) or weight (coefficient 0.62, 95% CI −1.09 to 2.32, P = 0.46). A comparison of the distribution of change from the preindex to postindex average HbA_1c and average weight across both the COVID-19 and control groups is shown in Fig. 1.

The results of the multivariable linear regression with the adjusted coefficients for postindex HbA_1c in the COVID-19 and control subgroups are shown in Fig. 2. For the subgroup of patients with COVID-19, there was a statistically significant decrease in postindex average HbA_1c of −0.43% (95% CI −0.61 to −0.25, P < 0.001) for patients who were on mechanical ventilation. There was no statistically significant difference in postindex average HbA_1c for patients with COVID-19 who were hospitalized or for patients in the control group who required hospitalization or mechanical ventilation. There was a small but statistically significant increase in postindex average HbA_1c for COVID-19 patients who were Hispanic or Latino of 0.15% (95% CI 0.05–0.24, P = 0.002) or on DPP-4 inhibitors prior to the index date of 0.21% (95% CI 0.10 – 0.32, P < 0.001). For both COVID-19 and control groups there was a statistically significant increase in postindex average HbA_1c in patients who were treated with insulin or a sulfonylurea. The analysis accounting for informative censoring due to death or loss to follow-up is shown in Supplementary Fig. 4.

The results of the multivariable linear regression showing the adjusted coefficients for postindex weight in the COVID-19 and control subgroups are shown in Fig. 3. In both the COVID-19 and control groups, there was a statistically significant decrease in postindex average weight with hospitalization, which was even more pronounced in patients who required mechanical ventilation. In both groups, there was a small but significant increase in average weight in patients who were male and on insulin prior to SARS-CoV-2 infection and decrease in patients in the older age-groups. For patients who were previously on thiazolidinediones, there was a significant increase in average weight in the COVID-19 group. The analysis accounting for informative censoring due to death or loss to follow-up is shown in Supplementary Fig. 5. Corticosteroid use within 7 days prior to 30 days after the index date was not associated with a change in average HbA_1c or weight in COVID-19 or control groups.

We studied a large, multicenter cohort of U.S. patients with T2D to analyze the associations between SARS-CoV-2 infection and glycemic control and weight at 1–2 years postinfection. While prior literature has shown that COVID-19 is acutely associated with hyperglycemia and increased incidence of diabetes, there are no studies evaluating longer-term effects on people with preexisting T2D including matched control subjects. In our analysis, there was evidence of little difference in postindex average HbA_1c for patients who had COVID-19 infection compared with a matched control group, and the results appear insensitive to unmeasured confounding. Despite the literature on the diabetogenic effects of COVID-19 during acute infection, and in line with the findings from Laurenzi et al. (14) our results suggest that for people with preexisting T2D, glycemic control in the year after illness is not clinically different from that among people with T2D who did not have COVID-19. Although the time frame of our study is still relatively short given the chronic nature of T2D, these results are reassuring and suggest that the hyperglycemia seen with acute infection is not persistent. Similarly, despite the literature showing short-term weight loss after acute COVID-19 infection, our results did not demonstrate a statistically significant difference in longer-term postindex weight or the distribution of average weight change compared with a control group. For people with T2D who have had COVID-19, our findings are promising in light of the concerns about malnutrition after infection.

Given the heterogeneity in severity of COVID-19 infection, we evaluated the longer-term implications of hospitalization and mechanical ventilation regarding average HbA_1c and weight. In our subgroup analysis of patients with COVID-19, those who required mechanical ventilation had a significant decrease in HbA_1c of ∼0.4%, which is potentially clinically significant considering the impact of glycemia on longer-term complications from diabetes and which was not seen in the control group. Although it is difficult to ascertain the exact mechanism for this decrease, persistent symptoms after critical care, such as dysphagia, altered taste, and anorexia may be more pronounced with COVID-19 and could account for some diet-related changes (34,35). It is also important to consider whether the overall decrease in HbA_1c is indicative of improved glycemic control or whether patients may have more periods of hypoglycemia contributing to lower HbA_1c. Mechanical ventilation was the variable with the largest association with postindex average weight, with average weight loss being 4 kg in the COVID-19 group and 2.9 kg in the control group. These findings may also reflect the increased LOS for COVID-19 patients who required mechanical ventilation (mean [SD] 44.8 [53.4]) compared with control patients who required ventilation for other medical indications (18.3 [24.6] days). Interestingly, although those hospitalized for COVID-19 had a longer LOS (14.4 [33.5] days) than those in the control group (6.3 [7.9] days), there was a similar degree of longer-term weight loss in both groups. Although there was a statistically significant decrease in HbA_1c of ∼0.1% in the control group, and a trend toward decrease in HbA_1c within the COVID-19 group, the clinical significance is minimal and suggests that hospitalization did not affect HbA_1c to the same degree as mechanical ventilation in either group. This is consistent with findings of other studies evaluating the effects of hospitalization on longer-term HbA_1c (36).

There were several limitations in our study. Given that the sites contributing data to the N3C are primarily from academic health care systems and that the study cohort is the subset of patients at these health systems who have >2 years of data, they may represent patients with more severe illness who have regular follow-up. As with any observational study, we cannot exclude the possibility that patients in the control group may not be negative for COVID-19 if they had been diagnosed with COVID-19 outside of the participating health systems or if they had a clinically unrecognized case of COVID-19 that was not diagnosed by either a test or ICD-10 diagnosis, which we sought to mitigate with the sensitivity analysis. Additionally, we are unable to assess virus variant or vaccination status because the data were date shifted for data privacy and a quality check of the vaccine data showed significant missingness. We also only included patients with at least one HbA_1c measurement during the preindex and postindex periods, which would exclude patients seen by providers using different electronic health record systems or with poor follow-up. The availability of data on HbA_1c for patients with T2D also limited our study design. While research on glycemia in diabetes has started to focus on glycemic trajectories rather than single data points or averages that do not account for patterns of glycemic changes (37–39), the time frame for data after the COVID-19 pandemic started is still relatively short. In the N3C database, many patients with T2D had only a single measurement for HbA_1c in the preindex (36.8%) and postindex (66.2%) time periods. As more data become available for research, future studies may be needed to better elucidate the effect of COVID-19 on trajectories of glycemic control.

To our knowledge, this is the first report of the longer-term effects of SARS-CoV-2 infection and severity of infection on glycemic control and weight in individuals with preexisting T2D including matched control subjects. There was no statistically significant difference in the postindex average HbA_1c or weight in patients who had COVID-19 compared with control subjects. Mechanical ventilation was associated with a decrease in average HbA_1c and body weight after COVID-19.

Acknowledgments. The authors are thankful for contributions from the following: N3C Publication Committee, Data Access Committee, Download Request Committee, and N3C Diabetes and Obesity Domain Team. The analyses described in this publication were conducted with data or tools accessed through the National Center for Advancing Translational Sciences N3C Data Enclave (covid.cd2h.org/enclave). This research was possible because of the patients whose information is included within the data from participating organizations (covid.cd2h.org/dtas) and the organizations and scientists (covid.cd2h.org/duas) who have contributed to the ongoing development of this community resource (20). A full list of Acknowledgments can be found in the Appendix.

Funding. This work was supported by National Center for Advancing Translational Sciences (NCATS) grant U24 TR002306 (R.W., R.V., M.A.H., R.M.). J.S.T. was supported, in part, by National Institutes of Health/National Institute of Diabetes, Digestive, and Kidney Diseases Mentored Patient Oriented Research Award (K23DK116935). J.B.B. was supported in part by grants from the National Institutes of Health (NCATS grant UL1TR002489 and Foundation for the National Institutes of Health grant P30DK124723). T.S. receives investigator-initiated research funding and support as principal investigator (R01 AG056479) from the National Institute on Aging and as co-investigator (R01 HL118255, R01 MD011680), National Institutes of Health. He also receives salary support as Director of Comparative Effectiveness Research, NC TraCS Institute, UNC Clinical and Translational Science Award (UL1TR002489).

Duality of Interest. T.S. receives salary support from the Center for Pharmacoepidemiology (current members: GlaxoSmithKline, UCB BioSciences, Takeda, AbbVie, and Boehringer Ingelheim) and from a generous contribution from Dr. Nancy A. Dreyer of IQVia to the Department of Epidemiology, University of North Carolina at Chapel Hill. T.S. does not accept personal compensation of any kind from any pharmaceutical company. He owns stock in Novartis, Roche, and Novo Nordisk. J.S.T. discloses receiving investigator-initiated grant support from Novo Nordisk on behalf of the Trustees of the University of Pennsylvania. J.D.M. reports consulting fees from MannKind and Medtronic Diabetes. He receives investigator-initiated research funding from Dexcom. J.B.B.’s effort is contracted for consulting and clinical trials by the University of North Carolina by Novo Nordisk, Dexcom, Sanofi, Tolerion, and vTv Therapeutics; he has received compensation for consulting from Alkahest, Altimmune, Anji, AstraZeneca, Bayer, Biomea Fusion, Boehringer Ingelheim, CeQur, Cirius Therapeutics, Dasman Diabetes Institute and Research Center (Kuwait City, Kuwait), Eli Lilly, Fortress Biotech, GentiBio, Glycadia, Glyscend, Janssen, Lilly, MannKind, Mediflix, MedImmune, Medscape, Mellitus Health, Moderna, Pendulum Therapeutics, Praetego, Stability Health, Valo, and Zealand Pharma; and he has stock/options in Glyscend, Mellitus Health, Pendulum Therapeutics, PhaseBio, Praetego, and Stability Health. No other potential conflicts of interest relevant to this article were reported.

R.W. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Author Contributions. Authorship was determined using International Committee of Medical Journal Editors recommendations. R.W. wrote the manuscript. R.V. cleaned and analyzed data. M.A.H. cleaned and analyzed data. M.V.P. contributed to the background and reviewed and edited the manuscript. C.T.B. reviewed and edited the manuscript. E.C. reviewed and edited the manuscript. S.G.J. reviewed and edited the manuscript. V.L. reviewed and edited the manuscript. J.D.M. reviewed and edited the manuscript. J.R. reviewed and edited the manuscript. M.S. reviewed and edited the manuscript. T.S. reviewed and edited the manuscript. J.S.T. reviewed and edited the manuscript. K.J.W. provided statistical expertise and reviewed and edited the manuscript. J.B.B. reviewed and edited the manuscript. J.S. reviewed and edited the manuscript. J.D.H. provided statistical expertise, contributed to the methods/results, and reviewed and edited the manuscript. R.M. analyzed data, contributed to the methods, and reviewed and edited the manuscript. R.W. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

N3C Consortium. Adit Anand, Tellen Bennett, Christopher G. Chute, Peter DeWitt, Michael Evans, Victor Garcia, Kenneth Gersing, Andrew Girvin, Melissa Haendel, Jeremy R. Harper, Janos Hajagos, Stephanie Hong, Emily Pfaff, Jacob Wooldridge, and Yun Jae Yoo.

Individual Acknowledgments for Core Contributors. The authors are thankful for contributions from the following N3C core teams: principal investigators, Melissa A. Haendel (lead), Christopher G. Chute (lead), Kenneth R. Gersing, Anita Walden; workstream, subgroup, and administrative leaders, Melissa A. Haendel (lead), Tellen D. Bennett, Christopher G. Chute, David A. Eichmann, Justin Guinney, Warren A. Kibbe, Hongfang Liu, Philip R.O. Payne, Emily R. Pfaff, Peter N. Robinson, Joel H. Saltz, Heidi Spratt, Justin Starren, Christine Suver, Adam B. Wilcox, Andrew E. Williams, and Chunlei Wu; key liaisons at data partner sites; regulatory staff at data partner sites; individuals at the sites who are responsible for creating the data sets and submitting data to N3C; Data Ingest and Harmonization Team, Christopher G. Chute (lead), Emily R. Pfaff (lead), Davera Gabriel, Stephanie S. Hong, Kristin Kostka, Harold P. Lehmann, Richard A. Moffitt, Michele Morris, Matvey B. Palchuk, Xiaohan Tanner Zhang, Richard L. Zhu; Phenotype Team (individuals who create the scripts that the sites use for submission of data, based on the COVID-19 and long COVID definitions), Emily R. Pfaff (lead), Benjamin Amor, Mark M. Bissell, Marshall Clark, Andrew T. Girvin, Stephanie S. Hong, Kristin Kostka, Adam M. Lee, Robert T. Miller, Michele Morris, Matvey B. Palchuk, Kellie M. Walters; Project Management and Operations Team, Anita Walden (lead), Yooree Chae, Connor Cook, Alexandra Dest, Racquel R. Dietz, Thomas Dillon, Patricia A. Francis, Rafael Fuentes, Alexis Graves, Julie A. McMurry, Andrew J. Neumann, Shawn T. O'Neil, Usman Sheikh, Andréa M. Volz, Elizabeth Zampino; partners from National Institutes of Health and other federal agencies, Christopher P. Austin (lead), Kenneth R. Gersing (lead), Samuel Bozzette, Mariam Deacy, Nicole Garbarini, Michael G. Kurilla, Sam G. Michael, Joni L. Rutter, Meredith Temple-O’Connor; Analytics Team (Individuals who build the Enclave infrastructure, help create code sets, variables, and help Domain Teams and project teams with their data sets), Benjamin Amor (lead), Mark M. Bissell, Katie Rebecca Bradwell, Andrew T. Girvin, Amin Manna, Nabeel Qureshi; Publication Committee Management Team, Mary Morrison Saltz (lead), Christine Suver (lead), Christopher G. Chute, Melissa A. Haendel, Julie A. McMurry, Andréa M. Volz, Anita Walden; and Publication Committee Review Team, Carolyn Bramante, Jeremy Richard Harper, Wenndy Hernandez, Farrukh M. Koraishy, Federico Mariona, Amit Saha, Satyanarayana Vedula.

Data Partners With Released Data. Stony Brook University (U24TR002306); University of Oklahoma Health Sciences Center (U54GM104938), Oklahoma Clinical and Translational Science Institute; West Virginia University (U54GM104942), West Virginia Clinical and Translational Science Institute; University of Mississippi Medical Center (U54GM115428), Mississippi Center for Clinical and Translational Research; University of Nebraska Medical Center (U54GM115458), Great Plains IDeA-Clinical & Translational Research; Maine Medical Center (U54GM115516), Northern New England Clinical & Translational Research Network; Wake Forest University Health Sciences (UL1TR001420), Wake Forest Clinical and Translational Science Institute; Northwestern University at Chicago (UL1TR001422), Northwestern University Clinical and Translational Sciences Institute; University of Cincinnati (UL1TR001425), Center for Clinical and Translational Science and Training; The University of Texas Medical Branch at Galveston (UL1TR001439), The Institute for Translational Sciences; Medical University of South Carolina (UL1TR001450), South Carolina Clinical & Translational Research Institute; University of Massachusetts Medical School Worcester (UL1TR001453), The University of Massachusetts Center for Clinical and Translational Science; University of Southern California (UL1TR001855), The Southern California Clinical and Translational Science Institute; Columbia University Irving Medical Center (UL1TR001873), Irving Institute for Clinical and Translational Research; The George Washington University Children’s Research Institute (UL1TR001876), Clinical and Translational Science Institute at Children’s National; University of Kentucky (UL1TR001998), University of Kentucky Center for Clinical and Translational Science; University of Rochester (UL1TR002001), UR Clinical & Translational Science Institute; University of Illinois at Chicago (UL1TR002003), UIC Center for Clinical and Translational Science; Penn State Health Milton S. Hershey Medical Center (UL1TR002014), Penn State Clinical and Translational Science Institute; the University of Michigan at Ann Arbor (UL1TR002240), Michigan Institute for Clinical and Health Research; Vanderbilt University Medical Center (UL1TR002243), Vanderbilt Institute for Clinical and Translational Research; University of Washington (UL1TR002319), Institute of Translational Health Sciences; Washington University in St. Louis (UL1TR002345), Institute of Clinical and Translational Sciences; Oregon Health & Science University (UL1TR002369), Oregon Clinical and Translational Research Institute; University of Wisconsin-Madison (UL1TR002373), UW Institute for Clinical and Translational Research; Rush University Medical Center (UL1TR002389), The Institute for Translational Medicine; The University of Chicago (UL1TR002389), The Institute for Translational Medicine; University of North Carolina at Chapel Hill (UL1TR002489), North Carolina Translational and Clinical Sciences Institute; University of Minnesota (UL1TR002494), Clinical and Translational Science Institute; Children’s Hospital Colorado (UL1TR002535), Colorado Clinical and Translational Sciences Institute; The University of Iowa (UL1TR002537), Institute for Clinical and Translational Science; The University of Utah (UL1TR002538), Uhealth Center for Clinical and Translational Science; Tufts Medical Center (UL1TR002544), Tufts Clinical and Translational Science Institute; Duke University (UL1TR002553), Duke Clinical and Translational Science Institute; Virginia Commonwealth University (UL1TR002649), C. Kenneth and Dianne Wright Center for Clinical and Translational Research; The Ohio State University (UL1TR002733), Center for Clinical and Translational Science; University of Miami Leonard M. Miller School of Medicine (UL1TR002736), University of Miami Clinical and Translational Science Institute; University of Virginia (UL1TR003015), integrated Translational Health Research Institute of Virginia (iTHRIV); Carilion Clinic (UL1TR003015), iTHRIV; University of Alabama at Birmingham (UL1TR003096), Center for Clinical and Translational Science; Johns Hopkins University (UL1TR003098), Johns Hopkins Institute for Clinical and Translational Research; University of Arkansas for Medical Sciences (UL1TR003107), UAMS Translational Research Institute; Nemours (U54GM104941), Delaware CTR ACCEL Program; University Medical Center New Orleans (U54GM104940), Louisiana Clinical and Translational Sciences Center; University of Colorado Anschutz Medical Campus (UL1TR002535), Colorado Clinical and Translational Sciences Institute; Mayo Clinic (UL1TR002377), Mayo Clinic Center for Clinical and Translational Sciences; Tulane University (UL1TR003096), Center for Clinical and Translational Science; Loyola University Medical Center (UL1TR002389), The Institute for Translational Medicine; Advocate Health Care Network (UL1TR002389), The Institute for Translational Medicine; OCHIN (INV-018455), Bill & Melinda Gates Foundation grant to Sage Bionetworks.