Re-analysis and meta-analysis of summary statistics from gene-environment interaction studies.

MOTIVATION: statistics from genome-wide association studies enable many valuable downstream analyses that are more efficient than individual-level data analysis while also reducing privacy concerns. As growing sample sizes enable better-powered analysis of gene-environment interactions, there is a need for gene-environment interaction-specific methods that manipulate and use summary statistics. RESULTS: We introduce two tools to facilitate such analysis, with a focus on statistical models containing multiple gene-exposure and/or gene-covariate interaction terms. REGEM (RE-analysis of GEM summary statistics) uses summary statistics from a single, multi-exposure genome-wide interaction study to derive analogous sets of summary statistics with arbitrary sets of exposures and interaction covariate adjustments. METAGEM (META-analysis of GEM summary statistics) extends current fixed-effects meta-analysis models to incorporate multiple exposures from multiple studies. We demonstrate the value and efficiency of these tools by exploring alternative methods of accounting for ancestry-related population stratification in genome-wide interaction study in the UK Biobank as well as by conducting a multi-exposure genome-wide interaction study meta-analysis in cohorts from the diabetes-focused ProDiGY consortium. These programs help to maximize the value of summary statistics from diverse and complex gene-environment interaction studies. AVAILABILITY AND IMPLEMENTATION: REGEM and METAGEM are open-source projects freely available at https://github.com/large-scale-gxe-methods/REGEM and https://github.com/large-scale-gxe-methods/METAGEM.

Self-Reported Food Security in Adolescents with Type 1 Diabetes: Association with Hemoglobin A1c and Mental Health Symptoms Independent of Household Food Security.

BACKGROUND: Typically, child exposure to food insecurity is assessed by caregiver reports of household food security. Child report has the potential for greater accuracy because it pertains only to the child whose experiences may differ from caregiver reports. OBJECTIVE: We assessed if adolescent-reported food insecurity was associated with levels of hemoglobin A1c (HbA1c), acute diabetes-related complications, depressive symptoms, and disordered eating behaviors in adolescents with type 1 diabetes, independently from household food security. METHODS: In a cross-sectional analysis of the multicenter SEARCH for Diabetes in Youth Cohort Study (phase 4, 2016-2019) including 601 adolescents aged 10-17 y with type 1 diabetes and their caregivers, household food security, and adolescent-reported food security were assessed using the 18-item Household Food Security Survey Module and the 6-item Child Food Security Assessment questionnaire. Age-stratified (10-13 and 14-17) regression models were performed to estimate independent associations, adjusting for sociodemographics, clinical factors, and household food security. RESULTS: Food insecurity was reported by 13.1% (n = 79) of adolescents and 15.6% (n = 94) of caregivers. Among adolescent-caregiver dyads, 82.5% (n = 496) of reports were concordant and 17.5% (n = 105) discordant, Cohen's κ= 0.3. Adolescent-reported food insecurity was not independently associated with HbA1c, diabetic ketoacidosis, and severe hypoglycemia, including in age-stratified analyses. Adolescent-reported food insecurity was independently associated with elevated odds of depressive symptoms [odds ratio (OR): 3.6; 95% confidence interval (CI): 1.3, 10.3] and disordered eating behaviors (OR: 2.5, 95% CI: 1.4, 4.6) compared with adolescents reporting food security; these associations remained in both age groups for disordered eating behaviors and in the older group for depressive symptoms. CONCLUSIONS: Adolescents with type 1 diabetes may experience food insecurity differently than caregivers. Adolescent-reported food insecurity was independently associated with depressive symptoms and disordered eating behaviors and thus may be an important attribute to assess in addition to household food security in adolescents with type 1 diabetes.

Properties of the Household Food Security Survey Module Scale in Young Adults with Diabetes.

BACKGROUND: The Household Food Security Survey Module (HFSSM) was not tailored to people with chronic diseases or young adults (YAs). OBJECTIVES: We aim to evaluate whether the 18-item HFSSM meets assumptions underlying the scale among YAs with diabetes. METHODS: Data from 1887 YAs with youth-onset type 1 diabetes or type 2 diabetes were used from the SEARCH for Diabetes in Youth Study, 2016-2019, and on 925 who returned for the SEARCH Food Security Cohort Study, 2018-2021, all of whom had completed the HFSSM. Guttman scaling properties (affirmation of preceding less severe items) and Rasch model properties (probability to answer an item based on difficulty level) were assessed. RESULTS: Items 3 (balanced meals) and 6 (eating less than one should) were affirmed more frequently than expected (nonmonotonic response pattern). At 1.2%-3.5%, item nonresponse was rare among type 1 diabetes but higher among type 2 diabetes (range: 3.1%-10.6%). Items 9 (not eating the whole day) and 3 did not meet the Guttman scaling properties. Rasch modeling revealed that item 3 had the smallest difficulty parameter. INFIT indices suggested that some responses to item 3 did not match the pattern in the rest of the sample. Classifying household food insecurity (HFI) based on items 1 and 2 compared with other 2-item combinations, including item 3, revealed a substantial undercount of HFI ranging from 5% to 8% points. CONCLUSIONS: Use of the HFSSM among YAs with diabetes could potentially result in biased HFI reporting and affect estimates of HFI prevalence in this population.

Estimating incidence of type 1 and type 2 diabetes using prevalence data: the SEARCH for Diabetes in Youth study.

BACKGROUND: Incidence is one of the most important epidemiologic indices in surveillance. However, determining incidence is complex and requires time-consuming cohort studies or registries with date of diagnosis. Estimating incidence from prevalence using mathematical relationships may facilitate surveillance efforts. The aim of this study was to examine whether a partial differential equation (PDE) can be used to estimate diabetes incidence from prevalence in youth. METHODS: We used age-, sex-, and race/ethnicity-specific estimates of prevalence in 2001 and 2009 as reported in the SEARCH for Diabetes in Youth study. Using these data, a PDE was applied to estimate the average incidence rates of type 1 and type 2 diabetes for the period between 2001 and 2009. Estimates were compared to annual incidence rates observed in SEARCH. Precision of the estimates was evaluated using 95% bootstrap confidence intervals. RESULTS: Despite the long period between prevalence measures, the estimated average incidence rates mirror the average of the observed annual incidence rates. Absolute values of the age-standardized sex- and type-specific mean relative errors are below 8%. CONCLUSIONS: Incidence of diabetes can be accurately estimated from prevalence. Since only cross-sectional prevalence data is required, employing this methodology in future studies may result in considerable cost savings.

Household Food Insecurity and Cognition in Youth and Young Adults with Youth-Onset Diabetes.

Objective: We evaluated the association of household food insecurity (FI) with cognition in youth and young adults with type 1 diabetes (T1D) or type 2 diabetes (T2D). Design: In this cross-sectional study, age-adjusted scores for composite Fluid Cognition, and sub-domain scores for Receptive Language and Inhibitory Control and Attention, were modeled stratified by diabetes-type using linear regression, with FI in the past year as the predictor, controlling for covariates. Tests for processing speed, inhibitory control/attention, working memory, episodic memory, and cognitive flexibility were administered to measure composite Fluid Cognition score. The NIHT-CB Picture Vocabulary Test was used to assess Crystallized Cognition score and rapid identification of congruent versus noncongruent items were used to assess Inhibitory Control and Attention score. Setting: The SEARCH for Diabetes in Youth study, representative of 5 U.S. states. Participants: Included 1574 youth and young adults with T1D or T2D, mean age of 21 years, mean diabetes duration of 11 years, 51% non-Hispanic white, and 47% had higher HbA1c levels (>9% HbA1c). Results: Approximately 18% of the 1,240 participants with T1D and 31% of the 334 with T2D experienced FI. The food-insecure group with T1D had a lower composite Fluid Cognition score (ß= -2.5, 95% confidence interval (CI)= -4.8, -0.1) and a lower Crystallized Cognition score (ß= -3.4, CI= -5.6, -1.3) than food-secure peers. Findings were attenuated to non-significance after adjustment for demographics. Among T2D participants, no associations were observed. In participants with T1D effect modification by glycemic levels were found in the association between FI and composite Fluid Cognition score but adjustment for socioeconomic characteristics attenuated the interaction (p=0.0531). Conclusions: Food-insecure youth and young adults with T1D or T2D did not have different cognition compared to those who were food-secure after adjustment for confounders. Longitudinal research is needed to further understand relations amongst these factors.

Diabetes complications and cognitive function in young adults with youth-onset type 1 or type 2 diabetes: the SEARCH for Diabetes in Youth Study.

Aims/hypotheses: People with type 1 (T1D) or type 2 diabetes (T2D) who also have diabetes complications can have pronounced cognitive deficits. It remains unknown, however, whether and how multiple diabetes complications co-occur with cognitive dysfunction, particularly in youth-onset diabetes. Methods: Using data from the SEARCH for Diabetes in Youth study cohort, a prospective longitudinal cohort, we examined clustering of complications and their underlying clinical factors with performance on cognitive tests in young adults with youth-onset T1D or T2D. Cognition was assessed via the NIH Toolbox Cognition Battery. The main cognitive variables were age-corrected scores for composite fluid cognition and associated cognitive subdomains. Diabetes complications included retinopathy, microalbuminuria, and peripheral neuropathy (PN). Lipids, systolic blood pressure (SBP), hemoglobin A1c, and other clinical factors were included in the analyses. Clustering was applied separately to each group (T1D=646; T2D=165). A three-cluster(C) solution was identified for each diabetes type. Mean values and frequencies of all factors were compared between resulting clusters. Results: The average age-corrected score for composite fluid cognition differed significantly across clusters for each group (p<0.001). People with T1D and the lowest average fluid cognition scores had the highest frequency of self-reporting at least one episode of hypoglycemia in the year preceding cognitive testing and the highest prevalence of PN. Persons with T2D and the lowest average fluid cognition scores had the highest SBP, the highest central systolic and diastolic blood pressures, and highest prevalence of PN. Conclusions/interpretations: These findings highlight shared (PN) and unique factors (hypoglycemia in T1D; SBP in T2D) that could be targeted to potentially mitigate cognitive issues in young people with youth-onset diabetes.

Health Care Transition to Adult Care in Type 1 Diabetes: Associations With Student and Employment Status-The SEARCH for Diabetes in Youth Study.

Successful transition from a pediatric to adult diabetes care provider is associated with reduced ambulatory diabetes care visits and increased acute complications. This study aimed to determine whether the degree of independence in diabetes care and the rate of acute complications after transition to adult diabetes care were associated with individuals' student or employment status. Nonstudents were found to be less likely than students to be independent with diabetes care, and employed nonstudents were at lower risk of diabetic ketoacidosis than unemployed nonstudents. Additional support may be needed for young adults who are not students or are unemployed to improve independence and reduce the risk for acute complications.

Preconception Counseling in Women With Diabetes: The SEARCH for Diabetes in Youth Study.

Preconception counseling is recommended for all women with diabetes starting at puberty to convey the importance of optimal diabetes management for maternal and fetal outcomes. This study included 622 female participants from the SEARCH for Diabetes in Youth study with a mean age of 22.2 years (range 14-35 years). Only 53.7% reported ever receiving preconception counseling, which was significantly lower among women seeing pediatric providers than those seeing adult or all-age providers. Older age and history of prior pregnancy were associated with increased odds of reporting having received preconception counseling. Identification of barriers to delivering preconception counseling to young females with diabetes and strategies to overcome them are needed to reduce the risk for pregnancy complications and adverse offspring health outcomes.

A longitudinal assessment of diabetes autoantibodies in the SEARCH for diabetes in youth study.

To assess changes in diabetes autoantibodies (DAs) over time in children and young adults with diabetes and determine whether observed changes were associated with demographic characteristics, clinical parameters and diabetes complications. Participants had DAs measured at baseline (10.3 ± 7.1 months after diabetes diagnosis) and at 12, 24 months and ≥5 years after the baseline measurement. At the ≥5-year follow-up, the presence of diabetes complications was assessed. We examined the associations between change in number of positive DAs and changes in individual DA status with the participants' characteristics and clinical parameters over time. Out of 4179 participants, 62% had longitudinal DA data and 51% had complications and longitudinal DA data. In participants with ≥1 baseline positive DA (n = 1699), 83.4% remained positive after 7.3 ± 2.3 years duration of diabetes. Decrease in number of positive DAs was associated with longer diabetes duration (p = 0.003 for 1 baseline positive DA; p < 0.001 for 2 baseline positive DAs) and younger age at diagnosis (p < 0.001 for 2 baseline positive DAs). No associations were found between change in number of positive DAs in participants with ≥1 baseline positive DA (n = 1391) and HbA1c, insulin dose, acute, or chronic complications after 7.7 ± 1.9 years duration of diabetes. DA status likely remains stable in the first 7 years after diabetes diagnosis. Younger age at diabetes diagnosis and longer duration were associated with less persistence of DAs. Measuring DAs after initial presentation may aid in diabetes classification but not likely in predicting the clinical course.

Determining diagnosis date of diabetes using structured electronic health record (EHR) data: the SEARCH for diabetes in youth study.

BACKGROUND: Disease surveillance of diabetes among youth has relied mainly upon manual chart review. However, increasingly available structured electronic health record (EHR) data have been shown to yield accurate determinations of diabetes status and type. Validated algorithms to determine date of diabetes diagnosis are lacking. The objective of this work is to validate two EHR-based algorithms to determine date of diagnosis of diabetes. METHODS: A rule-based ICD-10 algorithm identified youth with diabetes from structured EHR data over the period of 2009 through 2017 within three children's hospitals that participate in the SEARCH for Diabetes in Youth Study: Cincinnati Children's Hospital, Cincinnati, OH, Seattle Children's Hospital, Seattle, WA, and Children's Hospital Colorado, Denver, CO. Previous research and a multidisciplinary team informed the creation of two algorithms based upon structured EHR data to determine date of diagnosis among diabetes cases. An ICD-code algorithm was defined by the year of occurrence of a second ICD-9 or ICD-10 diabetes code. A multiple-criteria algorithm consisted of the year of first occurrence of any of the following: diabetes-related ICD code, elevated glucose, elevated HbA1c, or diabetes medication. We assessed algorithm performance by percent agreement with a gold standard date of diagnosis determined by chart review. RESULTS: Among 3777 cases, both algorithms demonstrated high agreement with true diagnosis year and differed in classification (p = 0.006): 86.5% agreement for the ICD code algorithm and 85.9% agreement for the multiple-criteria algorithm. Agreement was high for both type 1 and type 2 cases for the ICD code algorithm. Performance improved over time. CONCLUSIONS: Year of occurrence of the second ICD diabetes-related code in the EHR yields an accurate diagnosis date within these pediatric hospital systems. This may lead to increased efficiency and sustainability of surveillance methods for incidence of diabetes among youth.