Background: Statins are associated with cardiovascular disease risk reduction due to their direct effect on low-density lipoprotein cholesterol (LDL-C) levels. However, effective LDL-C control requires adherence to statin treatment. The association between adherence to statins and long-term LDL-C control has barely been studied in real-life contexts (e.g., primary care).
Objectives: To describe the impact of different levels of adherence to statin treatment on LDL-C control up to 5 years from the first statins prescription.
Methods: Retrospective cohort study using UK primary care data (The Health Improvement Network). We evaluated individuals aged 40 to 99 between 2006 and 2018, observed from their first statin prescription date up to 5 years. For all patients, historical data of statins prescriptions were organised, using the NICE criteria to define the dosage level of each prescription record (1 = low intensity, 2 = medium intensity, 3 = high intensity). For each three months of individual follow-up (quarter), we selected the highest dosage recorded to define the dosage of that period. Then, we performed a latent class growth analysis to identify classes or clusters of patients with different adherence patterns (i.e., each class was characterised by a different type of adherence). For each class identified, we described the non-linear LDL-C trajectory over time (up to 5 years of follow-up) by fitting fractional polynomial models. This allowed visualising longitudinal LDL-cholesterol change/control related to each class/type of adherence reached.
Results: We observed 60,257 patients, 32,298 men (53.6%) and 27,959 women (46.4%). From latent class growth analysis, we detected six classes of adherence that were labelled as 1) extremely poor (16.9%), 2) very poor (16.4%), 3) poor (11.9%), 4) moderate (12.8%), 5) good adherence with medium dose (35.7%), 6) good adherence with higher dose (6.3%). LDL-C trajectories differed across adherence classes. For example, the group with “extremely poor adherence” controlled their cholesterol levels for a very short period ( < 1 year) and then got back to baseline levels of LDL-Cholesterol. Conversely, people with good adherence showed a stable control of their LDL-Cholesterol levels over time (up to 5 years). The trend was consistent across adherence classes, which means the better the adherence, the better the LDL-C control.
Conclusions: We identified six classes of adherence to statin treatment, from extremely poor to good adherence, up to 5 years from treatment initiation. More than 1/3 of patients showed good adherence with medium dosage. People with better adherence reached better LDL-C control.
