Abstract Body (Do not enter title and authors here): Background: We recently found poor glycemic control (Hgb A1c >8%) to be strongly associated with increased risk of cardiovascular (CV) and limb events in diabetic patients with peripheral artery disease (PAD).

Hypothesis: The association between glycemic control and risk of CV and limb events is modified by age.

Methods: Using data from Peripheral Artery Disease: Long-term Survival Study (PEARLS), we identified diabetic PAD patients with at least one Hgb A1c in the two years before the index (date of PAD diagnosis), and at least one Hgb A1c within 2 years after the index but before first CV (myocardial infarction, stroke) or limb (chronic limb threatening ischemia, major amputation) event. Longitudinal glycemic control was assessed using HgbA1c and categorized as tight (≤7%; reference), moderate (>7% to 8%), and poor (>8%). Multivariable Cox proportional hazards models with longitudinal HgbA1c as a time-varying exposure and death as a competing risk examined the association between glycemic control and risk of clinical events. Analyses were stratified by median age (<70 years vs. >70 years).

Results: Among 45,934 patients, 97.8% were men, and 19.8% were Black. Compared to the <70-year group, those >70 years had lower baseline Hgb A1c (7.3% vs. 7.8%). Older patients were also less likely to be Black, be active smokers but had a higher prevalence of most co-morbidities (Table 1). In risk-adjusted analyses using longitudinal HgbA1c as a time-varying exposure, poor glycemic control was associated with increased risk of CV events (Figure 1), with Hgb A1c >8% associated with a 44% increased hazard compared to HgbA1c <7% category (sub-distribution hazard ratio [sHR]: 1.44; 95% CI: 1.37-1.52). The association between glycemic control and risk of limb events was even stronger, with 74% higher risk in those with HgbA1c>8%, compared to HgbA1c <7% (sHR: 1.74; 95% CI: 1.65-1.83; Figure 2). In both age groups, there was no heterogeneity in the association between glycemic control with CV (P-value for interaction term: 0.40) and limb events (P-value for interaction term: 0.09).

Conclusions: We found a graded, positive association between poor glycemic control and risk of CV and limb events. This association was consistent in both younger and older patients. Our findings indicate that optimal glycemic control may be beneficial in reducing the risk of clinical events, regardless of age.