Background: Emerging evidence has shown a higher rate of loneliness and social isolation among patients with diabetes, and loneliness and isolation has been associated with high risk of cardiovascular disease (CVD). Hypothesis: We investigated the extent to which the risks of CVDs could be lowered by improvement of loneliness and social isolation, individually or in conjunction with controlling traditional risk factors in diabetic patients. Methods: A total of 18,783 participants diagnosed with diabetes from the UK Biobank were included. A two-item scale and a three-item scale were used to assess the improvement of loneliness and isolation status, respectively. The degree of traditional risk-factor control was defined as the number of glycated hemoglobin (HbA1c), blood pressure (BP), low-density-lipoprotein cholesterol (LDL-C), smoking and microalbumin reaching the target range. Results: During a mean follow-up of 10.7 years, 3,129 incident total CVD events were documented, including 2,627 coronary heart disease (CHD) and 721 stroke. We found that each point increase in loneliness improvement scale was significantly associated with an 11% lower risk of CVD (HR, 95%CI, 0.89, 0.84-0.93) and a 12% lower risk of CHD (HR, 95%CI, 0.88, 0.83-0.93), but not with stroke. The improvement scale of social isolation was not associated with CVDs. Loneliness showed greater relative importance than lifestyle risk factors in preventing CVD among diabetic patients. We found a significant additive interaction between loneliness and the degree of traditional risk-factor control on the risk of CVDs ( P for additive interaction=0.002). Moreover, compared with non-diabetes controls, the diabetes-related excess CVD risk might be eliminated when three or more risk factors were controlled in non-lonely patients, but more than four are required in lonely patients. Conclusion: Among diabetes patients, improvement of loneliness is associated with a lower risk of CVD and shows an additive interaction with traditional risk factor control.