Background <p>Low physical activity and poor sleep quality increase the risk of diabetes. This study investigates how physical activity and sleep patterns influence both diabetes status onset and long-term survival using cross-sectional methods and causal analysis.</p> Methods <p>A total of 11,128 adults (≥ 20&#xa0;years) from 2011–2018 are included in this study. Physical activity and sleep quality are classified according to indicators <i>M</i> and <i>S</i>, respectively. The relationships between physical activity, sleep quality and diabetes status are explored using binary and multinomial logistic regression, COX proportional hazards regression, and subgroup analyses. Two-sample Mendelian randomization is performed for physical activity, four sleep patterns, and diabetes.</p> Results <p>Physical activity and sleep quality are significantly associated with a reduced risk of diabetes (<i>P</i> &lt; 0.05). Inactivity significantly increases the mortality risk in individuals with diabetes (HR = 1.83, 95%CI = 1.42–2.35, <i>P</i> &lt; 0.001). Age, biological sex and BMI differ significantly across diabetes status (<i>P</i> &lt; 0.001). We describe causal relationships between physical activity and four sleep patterns with diabetes and a strong negative correlation between sleep apnea and reduced diabetes risk (OR &gt; 1, <i>P</i> = 0.018).</p> Conclusions <p>Improving physical activity and sleep quality, particularly interventions for sleep apnea, may help to reduce the risk of developing diabetes, and the results provide scientific support for the development of public health strategies.</p>