BACKGROUND:How changes in hemoglobin A1c (HbA1c) and continuous glucose monitoring (CGM) metrics track together over time is poorly understood, particularly in type 2 diabetes. We investigated the patterns of change in HbA1c and CGM metrics among older adults with type 2 diabetes. METHODS:We analyzed data from 88 Atherosclerosis Risk in Communities (ARIC) study participants (baseline age, 82 years; 28% Black race and 42% women) who had HbA1c and 14-days of CGM assessed by standardized protocols at visit 9 (2021-22) and visit 10 (2023). HbA1c, CGM mean glucose, and time in range (TIR, 70-180 mg/dL) were compared across visits. Discordance was defined as having a different direction or magnitude of change, based on an absolute HbA1c change of 0.5% and the corresponding changes in CGM metrics derived from linear mixed-effect models. RESULTS:Over a median of 1.6 (IQR, 1.3-1.8) years, HbA1c, CGM mean glucose, and TIR showed moderate to strong correlations (r ∼0.5 to 0.7) across visits, and HbA1c had the lowest within-person variability (CVw = 8.4%). Approximately one-third of the participants had discordant changes between HbA1c and CGM metrics, with percentage agreement of 68.2% between HbA1c and CGM mean glucose, and 67.0% between HbA1c and TIR. Similar results were found in subgroups by sex, race, diabetes medication use, and after excluding participants with reduced kidney function. CONCLUSIONS:Among older adults with type 2 diabetes, long-term changes in HbA1c and CGM metrics are frequently discordant. This suggests the complementary nature of using HbA1c and CGM together to monitor glucose control.

Objectives: Peripheral neuropathy (PN) is common in older adults, even in the absence of diabetes. The association of PN with mortality in older adults is poorly characterized. We aimed to assess the association of advanced PN with mortality in community-dwelling adults aged 70-95 years. Methods: We conducted a prospective cohort analysis of participants in the Atherosclerosis Risk in Communities (ARIC) Study who underwent the 5.07 Semmes–Weinstein nylon monofilament insensitivity test for PN between 2016 and 2017 (ARIC Visit 6). Advanced PN was defined as lack of sensation at one or more foot sites during the monofilament test. We used Kaplan Meier survival analyses to estimate crude differences in all-cause mortality by PN status, and Cox regression models adjusting for demographics and clinical factors to quantify the association of advanced PN with all-cause mortality, overall and stratified by diabetes status. Results: Among 3,501 participants (median age 79 years; 41% male; 21% Black; 35% with diabetes, median follow up time 4.0 years), 34% had advanced PN. Estimated 5-year mortality was 25.5% (95%CI 23.0%–28.0%) in individuals with advanced PN and 14.2% (95%CI: 12.7%–15.6%) in those without PN (P<0.001). Participants with advanced PN had higher all-cause mortality compared to participants without PN regardless of diabetes status (Log-rank P<0.001; Figure ). After risk adjustment, advanced PN was associated with a higher risk of all-cause mortality overall (HR 1.33, 95%CI 1.15–1.55) and among participants without diabetes (HR 1.53, 95%CI 1.26–1.87). However, the association of PN with mortality was no longer significant among participants with diabetes (HR 1.10, 95%CI 0.87–1.38; P-value for interaction = 0.01). Conclusions: Older adults with advanced PN detected by monofilament testing have a higher risk of all-cause mortality compared to older adults without advanced PN, particularly among patients without diabetes. Reduced foot sensation may be an important screening tool for mortality risk in older individuals regardless of their diabetes status.

Background: There is growing interest and use of continuous glucose monitoring (CGM) technology as an adjunct or substitute for HbA1c in assessing glycemic control. However, how changes in CGM metrics and HbA1c track together over time is poorly understood, particularly in type 2 diabetes. We investigated the patterns of change in HbA1c and CGM metrics among older adults with type 2 diabetes. Methods: We analyzed data from 88 older adults with type 2 diabetes from the Atherosclerosis Risk in Communities (ARIC) Study who had measurements of HbA1c and wore CGM sensors at visit 9 (2021-22) and visit 10 (2023). HbA1c, CGM mean glucose, and time in range (TIR, percent time with CGM glucose 70-180 mg/dL) were compared using Deming regression, Bland-Altman plots, and Pearson’s correlations. Variability was assessed using the within-person coefficient of variation (CV w ). Proportion of variance in HbA1c change explained by changes in CGM metrics was estimated using linear regression with splines. Discordance was defined as having different direction or magnitude of change, based on an absolute HbA1c change of 0.5% and the corresponding changes in CGM metrics derived from linear mixed-effects models. Results: At baseline, the mean age was 82 years, 42% were women, and 27% self-identified as Black. The mean CGM wear time was 13 days (SD, 2) at both visits. Over a median of 1.6 (1.3, 1.8) years, the population mean HbA1c, CGM mean glucose, and TIR did not significantly change, with baseline means of 6.9% (SD, 0.9), 138.2 mg/dL (SD, 35.8), and 77.5% (SD, 21.4), respectively. CGM mean glucose showed higher variability (CV w =18.8%) than HbA1c (CV w =8.4%) and TIR (CV w =11.7%) ( Figure 1 ). Changes in CGM mean glucose explained 52.0% of the variance in HbA1c change, while changes in TIR explained 45.6%. We found that approximately one-third of the participants had discordant changes between HbA1c and CGM metrics over time, with percent agreement of 67.0% between HbA1c and CGM mean glucose, and 65.9% between HbA1c and TIR ( Figure 2 ). Similar results were found in subgroups by sex, race, diabetes medication use, and after excluding participants with reduced kidney function. Conclusions: Among older adults with type 2 diabetes, HbA1c is less variable than CGM mean glucose or TIR, and long-term changes (>1 year) in HbA1c and CGM metrics are frequently discordant. This suggests the complementary nature of using HbA1c and CGM together to monitor glucose control.

Background: We recently showed that a DASH-style diet optimized for diabetes (DASH4D diet) significantly reduced blood pressure and improved glycemic control in a randomized trial of adults with type 2 diabetes. To better understand the generalizability of the DASH4D findings, we compared demographic and clinical characteristics and the baseline dietary intake of the trial participants with those of the general US adult population with type 2 diabetes. Methods: We analyzed pre-randomization data from 103 participants in the DASH4D trial. For the comparison population, we analyzed data from 546 participants from NHANES 2021-2023 who met the inclusion criteria of the DASH4D study (aged 18 years and older with type 2 diabetes, systolic blood pressure of 120 – 159 mmHg, and diastolic blood pressure of <100 mmHg). Participants with HbA1c >9 were excluded. Typical dietary intake was assessed using 24-hour dietary recall in both studies. We assessed dietary recall data along with demographic, physical and clinical characteristics in both study populations. Analyses of NHANES accounted for the complex survey design and were weighted to provide nationally representative estimates. Results: The DASH4D participants were more likely to be female, Black race, and have higher attained education, and have lower total cholesterol than US adults with diabetes. The prevalence of obesity was similar in the two populations (Table 1) . Participants in DASH4D and US adults had comparable intake of protein (17 vs.16 % of total energy intake), carbohydrates (44 vs. 45% of total energy intake), and total fat (39 vs. 38% of total energy intake) (Figure 1) . DASH4D participants consumed a higher average level of fiber (10 g per 1,000 kcal) compared to NHANES (8 g per 1,000 kcal). Both populations had similar intakes of calcium, magnesium, phosphorous and potassium (Figure 2) . Conclusion: The usual dietary intake of participants in the DASH4D trial is broadly representative of US adults with diabetes, suggesting findings from the DASH4D trial are likely generalizable to the US population with diabetes.

This study reports decreasing out-of-pocket costs for insulin among Medicare beneficiaries not receiving the low-income subsidy and examines the distribution of costs by state from 2019 through 2023.

In this article, we present the cgmstats package for the analysis of continuous glucose monitoring (CGM) data. The use of wearable CGMs is growing rapidly. The latest generation of CGM systems do not require fingerstick calibration, are minimally invasive, and are frequently used in research studies. CGM sensors are typically worn for up to 2 weeks and record interstitial glucose measurements every minute to every 15 minutes, depending on the sensor used. CGM systems generate hundreds of measurements per day and thousands of measurements in one person over a single wear. There is a need for tools that allow researchers to efficiently organize and summarize the wealth of data on glucose patterns produced by CGM systems. The cgmstats package generates CGM summary measures for data from a variety of CGM systems and allows the user to flexibly define ranges and generate data visualizations. In this article, we provide an overview of the cgmstats package and examples of its use. The cgmstats package supports rigorous and reproducible analyses of CGM data.

Obesity disproportionately affects low-income individuals (1). By 2023, three glucagon-like peptide-1 receptor agonists (GLP-1RAs) were approved for obesity treatment (Saxenda, Wegovy, Zepbound). Medicaid, the primary insurer for low-income patients, does not consistently cover GLP-1RAs in people with obesity but no diabetes, with only 11 states offering coverage in 2023 (2). This ecological study evaluated state-level variation in GLP-1RA use in persons with obesity but no diabetes according to obesity coverage policies.

Background: In a recent controlled feeding trial, we showed that a DASH-style diet tailored for individuals with type 2 diabetes (DASH4D diet) improved short-term glycemic control assessed by continuous glucose monitoring. In this secondary analysis of the trial, we examined the effect of the DASH4D diet on biomarkers of glycemia. Methods: In the DASH4D trial, adults with type 2 diabetes were fed four isocaloric diets in a random order: DASH4D diet or comparison diet (representative of a typical American diet), each with higher or lower sodium. Each feeding period lasted 5 weeks and was separated by ≥1 week break. The primary outcome in this analysis was fructosamine, a marker reflecting glycemia over the past 2-3 weeks. Secondary outcomes included fasting glucose (reflecting glycemia at a single moment) and HbA1c (reflecting glycemia over 2-3 months). Outcomes were assessed at the end of each feeding period. Using an intention-to-treat approach, we estimated the effect of the DASH4D (versus comparison) diet on all outcomes with linear mixed effect models. In exploratory analyses, we examined the effect of the DASH4D diet by baseline HbA1c. Results: We included 101 participants (mean age: 67 years; 65% female; 88% Black adults). At baseline, mean HbA1c was 7.0% and 55% of participants were using two or more glucose-lowering medications. Compared to the comparison diet, the DASH4D diet significantly reduced end-of-period fructosamine (adjusted difference: -5.6 umol/L, P<0.002) and fasting glucose (adjusted difference: -4.5 mg/dL; P=0.019) (Figure A1-A2) . The DASH4D diet also had a small effect on HbA1c (adjusted difference: -0.09 %-points; P=0.035) (Figure A3) ; however, the 5-week duration of the feeding periods was suboptimal for estimating the effect of the diets on HbA1c. The effect of the DASH4D diet on fructosamine and fasting glucose was larger for participants with higher baseline HbA1c (Figure B1-B3) . Conclusion: The DASH4D diet significantly lowered fructosamine and fasting glucose in adults with type 2 diabetes. These results support the inclusion of the DASH4D dietary pattern into policy and clinical guidelines to improve glycemic control in type 2 diabetes.

Objectives: Peripheral neuropathy (PN) of the lower extremities is highly prevalent among older adults, even in the absence of diabetes, and has been linked to substantial morbidity and mortality. However, the biologic pathways underlying PN remain poorly understood. We examined the proteomic profiles of community-dwelling adults aged 70–95 years to identify proteins and pathways associated with PN. Methods: We conducted a prospective cohort analysis within the Atherosclerosis Risk in Communities (ARIC) Study among participants who underwent a monofilament insensitivity test for PN between 2016 and 2017 (ARIC Visit 6). PN was defined as loss of sensation at one or more foot sites. Blood samples collected at Visit 5 (2011-2013) were assayed using the SomaScan 5k proteomics platform. Logistic regression models were used to assess associations between 4,955 proteins and PN adjusting for demographic and clinical covariates. Statistical significance thresholds were set at P < 0.05 (nominal) and P < 0.05/4,955 (Bonferroni correction). Proteins meeting Bonferroni-corrected significance were carried forward for pathway analysis. Over-representation analysis (ORA) was then performed by mapping significant proteins to the Gene Ontology database to identify biological pathways associated with PN. Results: Among 2,514 participants (median age 74.0 years; 43% male; 16% Black adults; 31% with diabetes), 39% had PN based on monofilament testing. After adjustment for demographic and clinical covariates, PN was associated with 301 proteins at the nominal threshold, of which 11 (including Carbonic anhydrase 3, Seizure 6-like protein, Myomesin-2, Alpha-actinin-2, Myosin light chain 6B, and Myosin-binding protein C) remained significant after Bonferroni correction ( Figure ). ORA of the Bonferroni-significant proteins revealed enrichment in biologic processes related to muscle cell development, differentiation, and assembly of structural units (sarcomeres and myofibrils) ( Figure ). Conclusions: Older adults with PN, as detected by monofilament testing, show altered protein expression in pathways related to muscle development and organization. These pathways may explain previous observations in aging population linking PN to reduced functional decline and an increased risk of falls and related morbidities.

Objective: To examine the effect of the DASH4D diet (a DASH-style diet tailored for diabetes) on biomarkers of glycemia. Research Design and Methods: In this controlled feeding trial, adults with type 2 diabetes were fed four diets in a random order: DASH4D diet or comparison diet (representative of a typical American diet), each with higher or lower sodium. Each feeding period lasted 5 weeks. Using an intention-to-treat approach, we estimated the effect of the DASH4D (versus comparison) diet on fructosamine, fasting glucose, and HbA1c with linear mixed effect models. Results: Among 101 participants (mean age 67 years, 65% female, 87% Black adults), compared to the comparison diet, the DASH4D diet significantly reduced end-of-period fructosamine (adjusted difference: -5.6 μmol/L, P=0.002), fasting glucose (adjusted difference: -4.5 mg/dL; P=0.02) and HbA1c (adjusted difference: -0.09%-points; P=0.04). Conclusion: Our results support recommending the DASH4D diet for glycemic management in type 2 diabetes.

There is growing concern about increases in prediabetes and type 2 diabetes in young adults. Young-onset type 2 diabetes is associated with a more aggressive disease course than diabetes diagnosed later in life.1

Introduction: In a recent randomized trial, we showed that a DASH-style diet optimized for adults with type 2 diabetes (DASH4D) reduced mean glucose assessed by continuous glucose monitoring (CGM). However, the impact of DASH4D on postprandial glycemic response (PPGR), or glucose dynamics following meal taking, is unclear. Objective: Quantify the effect of the DASH4D diet on the PPGR time series and evaluate the proportion of the overall glycemic benefit of the DASH4D diet attributable to PPGR. Methods: The DASH4D trial had a 4-period crossover design. Adults with type 2 diabetes were randomized to an order of four diets: DASH4D or a typical American dietary pattern (comparison), each with lower or higher sodium. Calories were adjusted to maintain a stable weight. As sodium was not expected to impact PPGR, we combined the lower and higher sodium arms within each diet. Feeding periods were 5 weeks, with ≥ 1 week break between periods. CGM devices were worn from the 3 rd to 5 th weeks, recording up to 14 days of data. In a subset of participants, staff recorded meal timing during CGM wear. We fit a functional model regressing the PPGR time series (CGM glucose 1 hour before to 4 hours after meal start time) on diet type, including participant-specific random effects and adjustment for age, sex, body mass index (BMI), and time of day. We also applied functional regression-based mediation analyses to estimate the proportion of the diet effect on mean glucose that was mediated by differences in PPGR. Results: We collected PPGR data from 768 meals across the 65 participants who consented to meal monitoring (median age 68 years, 66% female). The DASH4D diet reduced PPGR, ranging from a difference -4.5 mg/dL at meal onset to -14.7 mg/dL from 1-2 hours after the start of the meal ( Fig. 1a ). There was a significant difference in PPGR between the DASH4D and comparison diets over the entire observation period ( Fig. 1b ). Differences in PPGR mediated 88% of the overall effect of the DASH4D diet on CGM mean glucose ( Fig. 2 ). Conclusion: Among adults with type 2 diabetes, the DASH4D diet improved glycemic control primarily by reducing PPGR.

OBJECTIVE:To examine the effect of the Dietary Approaches to Stop Hypertension for Diabetes (DASH4D) diet (a DASH-style diet tailored for diabetes) on biomarkers of glycemia. RESEARCH DESIGN AND METHODS:In this controlled feeding trial, adults with type 2 diabetes were fed four diets in a random order: DASH4D diet and comparison diet (representative of a typical American diet), each with higher and lower sodium. Each feeding period lasted 5 weeks. Using a modified intention-to-treat approach, we estimated the effect of the DASH4D (versus comparison) diet on fructosamine, fasting glucose, and HbA1c with linear mixed-effects models. RESULTS:Among 101 participants (mean age 67 years, 65% female, 87% Black adults), compared with the comparison diet, the DASH4D diet significantly reduced end-of-period fructosamine (adjusted difference: -5.6 μmol/L, P = 0.002), fasting glucose (adjusted difference: -4.5 mg/dL; P = 0.02), and HbA1c (adjusted difference: -0.09 percentage points; P = 0.04). CONCLUSIONS:Our results support recommending the DASH4D diet for glycemic management in type 2 diabetes.

The American Diabetes Association (ADA) recommends continuous glucose monitoring (CGM) metrics for monitoring glucose control, including time in range (TIR 70–180 mg/dL), time below range (TBR <70 and <54 mg/dL), time above range (TAR >180 and >250 mg/dL), mean glucose, glucose coefficient of variation (CV) (1). Cluster analysis of these clinical CGM metrics may help us uncover distinct glycemic subgroups.