Age at type 2 diabetes onset and glycaemic control: results from the National Health and Nutrition Examination Survey (NHANES) 2005–2010

Seth A. Berkowitz 0 James B. Meigs 0 Deborah J. Wexler 0 0 D. J. Wexler Diabetes Center, Department of Medicine, Massachusetts General Hospital/Harvard Medical School , Boston, MA, USA Aims/hypothesis We tested the hypothesis that age younger than 65 years at type 2 diabetes diagnosis is associated with worse subsequent glycaemic control. Methods A cross-sectional analysis of data from participants in the 2005-2010 National Health and Nutrition Examination Survey was performed. For adults with self-reported diabetes, we dichotomised age at diabetes diagnosis as younger (<65 years) vs older (65 years). The primary outcome of interest was HbA1c >9.0% (75 mmol/mol). Secondary outcomes were HbA1c >8.0% (64 mmol/mol) and >7.0% (53 mmol/mol). We used multivariable logistic regression for analysis. Results Among 1,438 adults with diabetes, a higher proportion of those <65 years at diagnosis compared with those 65 at diagnosis had an HbA1c >9.0% (14.4% vs 2.5%, p <0.001). After adjustment for sex, race/ethnicity, education, income, insurance, usual source of care, hyperglycaemia medication, duration of diabetes, family history, BMI and waist circumference, age <65 years at diagnosis remained significantly associated with greater odds of HbA1c >9.0% (OR 3.22, 95% CI 1.54, 6.72), HbA1c >8.0% (OR 2.72, 95% CI 1.43, 5.16) and HbA1c >7.0% (OR 1.92, 95% CI 1.18, 3.11). The younger group reported fewer comorbidities, but were less likely to report good health (OR 0.54, 95% CI 0.36, 0.83). Conclusions/interpretation Younger age at type 2 diabetes diagnosis is significantly associated with worse subsequent glycaemic control. Because patients who are younger at diagnosis have fewer competing comorbidities and complications, safe, aggressive, individualised treatment could benefit this higher-risk group. - Abbreviations CDC Centers for Disease Control and Prevention CHF Congestive heart failure COPD Chronic obstructive pulmonary disease CVA Cerebrovascular accident ESRD End stage renal disease NHANES National Health and Nutrition Examination Survey Patient-centredness is a priority in type 2 diabetes care [1]. With increasing recognition that the benefits and burdens of treatment differ by patient population, identifying subgroups at high risk of poor outcomes is an important goal, which may facilitate population management for diabetes. In this sense, different phenotypes of type 2 diabetes may identify populations that may need and benefit from more intensive interventions. Mounting evidence suggests that those with onset of type 2 diabetes in early or mid-adult life, compared with those with onset at an older age (65 or older), may have a more severe disease course, with increased risk of microvascular complications and worse glycaemic control [25]. While these differences in severity of dysglycaemia may be due to known risk factors such as longer duration of diabetes and higher BMI, they may also reflect more significant insulin deficiency in those diagnosed at younger ages. In this study we tested the hypothesis that those diagnosed at a younger age would have worse glycaemic control, even after adjustment for duration of diabetes, higher BMI and other known risk factors for worse glycaemic control. Data source and study sample We conducted a cross-sectional analysis combining three cycles of the National Health And Nutrition Examination Survey (NHANES). NHANES is a series of epidemiological surveillance surveys conducted by the Centers for Disease Control and Prevention (CDC) in community-dwelling participants [6]. Trained interviewers meet participants in their homes and administer a structured questionnaire in English, Spanish or with an interpreter. Participants then travel to a mobile examination centre (MEC) for physical examination and blood draws for laboratory analysis. Data are collected in 2-year cycles, which can be pooled to allow ascertainment of a larger number of cases. For this analysis, we pooled cycles in order to have enough participants diagnosed with diabetes at an older age and to allow robust adjustment for potential confounders. Full methodological details of NHANES have been previously published [6]. Our study included all adult (age >20 years) NHANES participants from 20052010 with diabetes [7]. We excluded participants who were pregnant at the time of examination. To minimise inclusion of type 1 diabetes patients, we excluded patients who were diagnosed before the age of 30 and started on insulin around the time of diagnosis. This approach is in accordance with previous evaluations of NHANES data [8, 9]. A participant was considered to have diabetes if he or she answered Yes to the question Other than during pregnancy, have you ever been told by a doctor or healthcare professional that you have diabetes or sugar diabetes? This method has been used in previous studies [1012] and CDC publications [13]. Sensitivity of this measure has been reported to be >95% in previous NHANES analyses [9], and specificity has been reported to be as high, at 97% [11]. We did not include participants with biochemical but not self-reported diabetes because age at diagnosis could not be determined in these cases. The Partners HealthCare Human Research Committee exempted this study from institutional review board review. Age at diabetes onset For this study, we dichotomised age at type 2 diabetes diagnosis as occurring in younger adult life (age <65 years) compared with older adult life (age 65 years) based on patient self-report. This dichotomisation was based on previous observations of differential effects of diabetes on health status by age in this range [3, 14]. In exploratory analyses, we also treated age at diagnosis as a continuous variable. Outcomes We used HbA1c percentage as a measure of glycaemic control. Our primary outcome of interest was whether the HbA1c value was >9.0% (75 mmol/mol), which represents out-of-control hyperglycaemia for all patients [15]. In order determine whether glycaemic control was worse at other commonly used thresholds [15], we conducted secondary analyses using outcomes of HbA1c above or below 8.0% (64 mmol/mol) and 7.0% (53 mmol/mol). To determine whether age at diabetes diagnosis was associated with comorbidity, we used responses on previously validated self-report items [6] for end stage renal disease (ESRD), current asthma diagnosis, congestive heart failure (CHF), CHD, chronic obstructive pulmonary disease (COPD), cerebrovascular accident (CVA) and history of malignancy of any kind. We also considered the health status of patients with the question Would you say your health in general is excellent, very good, good, fair, or poor?. We dichotomised this question into good (excellent, very good or good) vs poor (fair or poor) health [14]. Response to this item correlates highly with a range of health outcomes [1620]. Demographic and socioeconomic variables We considered several demographic and socioeconomic variables that (...truncated)