Adherence to medication is a significant factor when dealing with the prevention of cardiovascular diseases. Medications can reduce the risk of stroke or heart attack, but they only work to the extent that patients adhere to the medications. The first line of treatment is often statin medication. When patients have low adherence to low statins, past studies suggest that to increase the adherence, the physicians should increase the surveillance and remind the patient of the benefits of taking the medication. For cardiovascular diseases, the literature has studied adherence to statins widely, considering demographics and medical factors and their relationship to adherence. The American College of Cardiology suggests constant surveillance over the first year that the physician prescribes the medication, checking medication effects and adherence. Afterwards surveillance is typically reduced and patient adherence is checked less frequently. In this study, we focus on long-term adherence with an emphasis on patients after the first year of medication. We study factors associated with patient adherence using longitudinal data for statins in a large cohort of patients seen in the national Veterans Affairs health system. We further develop predictive models and assess their performance as a means to help physicians identify patients with poor adherence.