Primary Prevention Statins in Older Adults: Personalized Care for a Heterogeneous Population

Deirdre O'Neill, MD, MSc; Neil J. Stone, MD; Daniel E. Forman, MD


J Am Geriatr Soc. 2020;68(3):467-473. 

In This Article

Case Study 1

Patient A.G., an 80-year-old woman with a past medical history of breast cancer (in remission postmastectomy and radiation), hypertension, iron deficiency anemia, osteoarthritis, urge incontinence, and chronic obstructive pulmonary disease, presents to clinic. She has no history of coronary artery disease, stroke, or peripheral vascular disease. She is currently taking losartan-HCTZ 100/25 mg daily, ferrous sulfate 325 mg twice/day, acetaminophen 1000 mg 3 times/day, albuterol two puffs inhaled every 4 hours as needed, tiotropium bromide 18 μg inhaled daily, and fluticasone propionate 250 μg inhaled daily.

A recent wellness visit showed blood pressure of 130/82, heart rate 68 bpm and regular, and a benign cardiopulmonary examination. Lipid panel showed total cholesterol 240 mg/dL, low density lipoprotein-cholesterol (LDL-C) 189 mg/dL, high-density lipoprotein-cholesterol (HDL-C) 55 mg/dL, and triglycerides 199 mg/dL and hemoglobin (Hb) A1c 5.9%.

A.G. lives independently with her husband in a community for older adults. She walks with a cane. She is independent in all activities of daily living (ADLs) but requires assistance to get in and out of the bathtub due to osteoarthritis. She and her husband do their finances and medications together to catch mistakes, but she remains fully independent in cooking and cleaning. She goes for 10- to 15-minute walks 2 to 3 days per week and participates in an aquatic exercise program and yoga class for older adults once per week. She is an ex-smoker (quit 5 years ago; 50 pack-year smoking history) and drinks two to three glasses of wine per week.

A.G. says she wants to remain living independently with her husband until "their final days." She values going to church on Sunday, participating in a weekly bridge league, and visiting with her great-grandchildren. She says she is concerned about CV risk reduction, but when asked specifically about cholesterol-lowering medication, she has not been on one previously and has concerns, having heard "dreadful things" about that class of medications from her friends, family, and television. As a result of the patient's reluctance to use statin medications, a coronary artery calcium (CAC) score is performed to resolve uncertainty regarding her risk status. This showed a score of 390, or 75th percentile for age, sex, and ethnicity.

Case 1 Summary & Discussion

A.G. is an 80-year-old woman with multimorbidity but who still maintains a satisfying quality of life. She continues to be active in her older years and has a goal of being functionally independent and able to enjoy activities that are meaningful to her. She does not have clinical ASCVD, and her ASCVD risk factors include medically controlled hypertension and she recently quit smoking. She is concerned about her cholesterol and has thought about statins, but she is hesitant given the bad press associated with them.

The current cholesterol guidelines suggest a IIb level of evidence for initiating a moderate intensity statin in a primary prevention patient age 75 years or older, who has an LDL of 70 to 189 mg/dL.[1] A IIb recommendation is considered "weak," implying that although some data indicate benefit, they are not conclusive.

Shared decision making, in contrast to paternalistic care, engages patients in the decision-making process.[9] It also implies the need to inform patients with risk and benefit data that is specific to them and in a way they understand. Health literacy, or the ability to understand basic health information and be involved in making healthcare decisions, is essential to the shared decision-making process.[10] Yet it is estimated that only 12% of adults in the United States have proficient health literacy.[11] Most read at an 8th-grade level, and 20% read at a grade 5 level or below.[12] The US Department of Health and Human Services recommends public health information be published at a grade-7 reading level; however, health-based information often exceeds this level.[13] Challenges to comprehension among older adults are compounded by predictable age-related cognitive declines. Loss of executive cognition is especially common, with impairments in decision making even in those without significant memory loss.[14,15]

Numeracy, or the ability to use, interpret, and communicate numbers, is also limited among many older adults. Two-thirds of the general population of Americans score in the lowest two levels of numeracy[16] making decision making even more difficult. Pictographs are pictorial representations of statistical risk, most often representing the denominator as a number of dots and the numerator as a percentage of those dots prominently highlighted. Pictographs were shown to help convey numerical information more effectively than tables, graphs, or percentages, and they are increasingly being used to facilitate shared decisions when numerical perspectives are necessary.[17]

Standardized approaches to shared decision making have been established. It begins with a healthcare professional inviting the patient to participate in the decision-making process, making it known that they have options and pointing out that their individual values and goals are important.[9,18,19] All options are presented to a patient, specifying risks and benefits. Next the patient is assisted in evaluating options, basing on literature as well as their own goals. This step requires understanding each patient's preferences and values. The healthcare professional facilitates decision making while allowing the patient time to consider the decision, ask questions, talk to family members, and address concerns. Shared decision making was shown to improve patient's knowledge regarding their own health, to improve communication regarding options and possible risks of treatment/care, to decrease decisional conflict in patients and physicians, and to increase patient participation in care and increase value congruence between patients and physicians.[9,10,20–22]

To engage in optimal shared decision making, additional information can be helpful. A CAC score may be useful.[23,24] A CAC is calculated using noncontrast cardiac computed tomography. It takes 10 to 15 minutes to complete and exposes the patient to a very low radiation dose, an average of less than 1.5 mSv.[25] A CAC score is particularly useful in older adults when it is very low because it can help "de-risk" an individual who might otherwise seem to be an intermediate or high risk for ASCVD based on their age or other CV risks that are common with age (eg, hypertension). In these older patients, a CAC score of 0 is particularly useful because emerging evidence shows its utility in reclassifying risk and supporting decisions not to initiate statin therapy in adults aged up to 80 years.[26,27]

The BioImage trial used CAC scores to reclassify ASCVD risk including more than 5800 patients up to age 80 years without cardiovascular disease (CVD) at baseline.[26] The CAC score was found to improve the specificity for coronary events by 22% (P < .0001), without a loss in sensitivity. The CAC score was 0 in about one-third of patients, with the de-risking of such patients driving the study results. Recently, a substudy of the BioImage trial was published, evaluating 13 markers of cardiovascular (CV) risk to determine which were the greatest negative risk markers and best able to predict older adults at low CVD risk. It showed that a CAC of 0 and 10 or lower were the strongest negative risk markers for CVD, associated with remarkably low CV risk and net reclassification index of .23 to .28.[27]

Although the positive predictive value of a high CAC score is established in younger adults, its value remains uncertain in older adults. The proportion of more stable fibrocalcific and healed plaques increases in older adults, such that CAC usually increases with age,[28] reducing the diagnostic and prognostic specificity for ASCVD in older adults.[28,29] Still, clinical implications of high CAC scores are being investigated. Tota-Maharaj et al studied the use of the CAC score to predict mortality in more than 44 000 patients, comparing those younger than 45 years to those 75 years or older. After multivariable adjustment, a higher CAC score was independently predictive of increased mortality (compared with CAC = 0) in both the younger than 45 and the 75-year or older age groups.[30]

Data pertaining to elevated CAC in older adults are also available in the MESA cohort, a population-based sample of more than 6000 participants including various ethnicities, sexes, and ages (45–84 years).[31] CAC scores were collected in MESA to document the prevalence and progression of subclinical CV disease. Through the use of this population database, a MESA risk calculator was created, allowing a patient's traditional CV risk factors and CAC score to be entered, and providing an estimated 10-year coronary heart disease (CHD) risk score ( This has value in illustrating CHD risk for both patients and clinicians, allowing the selection of statin therapy for those who are presumably most likely to benefit.

In older adults, not only is specificity of elevated CAC less certain, but longevity may also be insufficient for statin benefits to manifest. Notably, goals to better clarify the utility of positive CAC scores are being prioritized by PREVENTABLE investigators in a proposed ancillary study proposal.

Despite such persistent CAC uncertainties in regard to older age, the ACC/AHA expert consensus document on CAC scoring developed fixed cutoff points and their associated relative risk (RR) of CVD, using a random effects model on univariable and summary RR ratios from six publications on CAC scores, as seen in Table 1.[32] This consensus document was derived from six investigations, with a cumulative population of more than 27 000 patients, ranging from 45 to 96 years of age.[33,34] Using these fixed cutpoints, a CAC score of 100 to 400 is consistent with moderate CVD risk, suggesting increased risk of future coronary events, such that this individual should possibly be reclassified as high risk.[25] Older adults remain significantly underrepresented in this consensus statement, potentially limiting the generalizability of cutpoints to older adults and making future trials to corroborate or modify these recommendations likely.

Overall, A.G. is at increased risk of ASCVD, as evidenced by both her typical CV risk factors as well as her CAC score. However, consideration of her life expectancy, goals of treatment, predominant fears and concerns regarding treatment, as well as clinician perception of anticipated risks and benefits of treatment, all need to be considered as parts of shared decision making. The CAC score helps inform this patient's decision by better appreciating her residual risk even after controlling blood pressure, stopping cigarettes, and walking regularly.