Skip main navigation

The Expected 30-Year Benefits of Early Versus Delayed Primary Prevention of Cardiovascular Disease by Lipid Lowering

Originally publishedhttps://doi.org/10.1161/CIRCULATIONAHA.120.045851Circulation. 2020;142:827–837

Abstract

Background:

Lipid-lowering recommendations for prevention of atherosclerotic cardiovascular disease rely principally on estimated 10-year risk. We sought to determine the optimal time for initiation of lipid lowering in younger adults as a function of expected 30-year benefit.

Methods:

Data from 3148 National Health and Nutrition Examination Survey (2009–2016) participants, age 30 to 59 years, not eligible for lipid-lowering treatment recommendation under the most recent US guidelines, were analyzed. We estimated the absolute and relative impact of lipid lowering as a function of age, age at initiation, and non–high-density lipoprotein cholesterol (HDL-C) level on the expected rates of atherosclerotic cardiovascular disease over the succeeding 30 years. We modeled expected risk reductions based on shorter-term effects observed in statin trials (model A) and longer-term benefits based on Mendelian randomization studies (model B).

Results:

In both models, potential reductions in predicted 30-year atherosclerotic cardiovascular disease risk were greater with older age and higher non–HDL-C level. Immediate initiation of lipid lowering (ie, treatment for 30 years) in 40- to 49-year-old patients with non–HDL-C ≥160 mg/dL would be expected to reduce their average predicted 30-year risk of 17.1% to 11.6% (model A; absolute risk reduction [ARR], 5.5%) or 6.5% (model B; ARR 10.6%). Delaying lipid lowering by 10 years (treatment for 20 years) would result in residual 30-year risk of 12.7% (A; ARR 4.4) or 9.9% (B; ARR 7.2%) and delaying by 20 years (treatment for 10 years) would lead to expected mean residual risk of 14.6% (A; ARR 2.6%) or 13.9% (B; ARR 3.2%). The slope of the achieved ARR as a function of delay in treatment was also higher with older age and higher non–HDL-C level.

Conclusions:

Substantial reduction in expected atherosclerotic cardiovascular disease risk in the next 30 years is achievable by intensive lipid lowering in individuals in their 40s and 50s with non–HDL-C ≥160 mg/dL. For many, the question of when to start lipid lowering might be more relevant than whether to start lipid lowering.

Clinical Perspective

What Is New?

  • This study examines the benefits of early versus delayed lipid lowering for prevention of atherosclerotic cardiovascular disease as predicted by 2 models.

  • Model A assumes benefit accrues principally from lesion stabilization and can be estimated based on statin randomized, controlled trials with relative risk reduction constant over time.

  • Model B assumes benefit is based on both lesion stabilization and prevention of new lesion formation.

What Are the Clinical Implications?

  • Both models demonstrate that greater benefit is achieved with earlier lipid lowering, particularly in those with elevated levels of non–high-density lipoprotein cholesterol.

  • As expected, benefit is greater with model B than model A.

Introduction

Editorial, see p 838

Primary prevention of arteriosclerotic cardiovascular disease prescribes a healthy lifestyle for everyone plus pharmacologic therapy for those with extremely high levels of low-density lipoprotein cholesterol (LDL-C), diabetes mellitus, or an elevated 10-year risk of a cardiovascular event. However, almost half of all cardiovascular events occur before age 65 years, the age at which the risk of a cardiovascular event begins to increase rapidly.1,2 Based on the work of Stary and colleagues,3 the life course of atherosclerosis may be divided into 3 phases, which may progress at different rates in different individuals. During the incubation phase, which covers approximately the first 3 decades of life, simple, infiltrative benign lesions, initiated by trapping of apoB (apolipoprotein B) particles within the arterial wall, are gradually converted to complex lesions, capable of triggering a clinical event. Over the next 3 decades, more lesions form and extend within the arterial wall. The 10-year risk of a clinical event during this midperiod of life depends on the extent of disease that is present, generally very low at the beginning, but then gradually increasing as disease multiplies and matures. During the third phase, which covers the next 3 decades, new lesion formation may continue in some, but progression of disease will be the principal change in most, with risk increasing ever more rapidly as the normal architecture of the arterial wall is replaced by disease. It is the relentless progression of disease over the life span that produces the progressive increase in cardiovascular risk with age.

Most lipid-lowering strategies that reduce cardiovascular clinical events work by reducing the number of apoB particles in plasma, which in turn decreases the rate at which apoB particles enter and are trapped within the arterial wall.4 Two different benefits are possible. First, lesions may not form. Second, existing lesions may be stabilized and undergo partial healing. If a lesion is not formed, protection is complete and prevention perfect. If an advanced lesion is stabilized or partially healed, protection is partial, and prevention imperfect. Accordingly, the protection provided by lipid-lowering strategies would be expected to vary depending on when during the disease process the intervention was begun. The earlier the intervention, the greater the benefit should be. But how early, and in whom?

To answer these questions, it is essential to know the potential benefits of lipid lowering earlier in life versus the costs of delaying lipid lowering to later in life. We estimated the absolute and relative impact of interventions of different durations, starting at different ages and in individuals with different levels of non–high-density lipoprotein cholesterol (non–HDL-C), on the risk of cardiovascular disease over the succeeding 30 years. We use 2 different conceptual biological models: in model A, the benefit from lipid lowering occurs only by stabilization of preexisting advanced plaques; in model B, the benefit of prevention of lesion formation plaques is incorporated as well as the benefit of plaque stabilization. Model A uses the benefit observed in the randomized, controlled trials (RCTs) of statins in individuals at risk for atherosclerotic cardiovascular disease (ASCVD), assuming it was derived largely by stabilization of preexisting lesions. Model B adds evidence from Mendelian randomization studies, which have demonstrated that the benefit per unit lower of the atherogenic apoB lipoproteins over a lifetime is 2- to 3-fold greater than the benefit produced by similar decreases later in life.5

Methods

Study Population

Our study population included 3148 US civilian noninstitutionalized National Health and Nutrition Examination Survey (NHANES) participants, age 30 to 59 years (data available at https://wwwn.cdc.gov/nchs/nhanes/Default.aspx). Their data were extrapolated using NHANES weights to represent the US population. The 4 most recent cycles were included: 2009 to 2010, 2011 to 2012, 2013 to 2014, and 2015 to 2016. From 4858 eligible individuals, we excluded those with self-reported history of ASCVD (n=255) or currently on lipid-lowering treatment (n=472). We excluded individuals with a likely recommendation for lipid-lowering therapy under the most recent American College of Cardiology/American Heart Association guidelines6: individuals with diabetes mellitus (n=425), LDL-C ≥190 mg/dL (4.94 mmol/L; n=102), or calculated 10-year ASCVD risk 7.5% or higher (n=262). Furthermore, we excluded participants with missing records for LDL-C levels or inputs to the 30-year risk calculator or triglycerides ≥400 mg/dL (4.52 mmol/L; n=194), consistent with LDL-C estimation using the Friedewald equation.7 In a sensitivity analysis, we further excluded only individuals with ASCVD risk ≥20%, consistent with the more stringent treatment threshold. Diabetes mellitus was defined as hemoglobin A1c concentration ≥6.5% or fasting glucose level ≥126 mg/dL or a self-report of diagnosis. Lipid-lowering and antihypertensive treatments were ascertained based on self-reported questionnaires. All individuals in the NHANES cohort provided informed consent. Data used in this study are publicly available, and the survey has been approved by the National Center for Health Statistics ethics review board. The analyses reported in this article were considered exempt by McGill University Institutional Review Board.

Statistical Analyses

All analyses followed NHANES analytic guidelines, taking into account the complex, multistage probability-sampling design of using sample weights to account for nonresponse rates and oversampling of certain segments of the population. Analyses were stratified by age group (30 to 39, 40 to 49, 50 to 59 years) as a main driver of risk and cholesterol level (non–HDL-C <130 mg/dL [3.38 mmol/L], 130 mg/dL [3.38 mmol/L] to 160 mg/dL [4.16 mmol/L], ≥160 mg/dL [4.16 mmol/L]) as a main driver of relative risk reduction (RRR). Baseline characteristics of participants were expressed as weighted means with corresponding 95% confidence limits.

We adopted a 30-year risk perspective from “today.” Thus expected risk and risk reductions for 30- to 39-year-old patients were estimated until age 60 to 69, for 40- to 49-year-old patients until age 70 to 79, and for 50- to 59-year-old patients until age 80 to 89. ASCVD risk (cardiovascular death, myocardial infarction, and stroke) was estimated using the additive 30-year Framingham equation.8 This approach calculates the total risk in a given time horizon as a sum of annual risks and accounts for the competing risk of death from other causes. This additive feature enables estimation of risk for any given time subset within the time horizon, including residual risk after a given time period. For example, remaining risk after 10 years is estimated as the sum of annual risks from years 11 through 30.

RRR attributable to lipid lowering was estimated according to the 2 models postulated in the introduction. Model A estimates the benefit from stabilization of preexisting advanced plaques, based on a constant 22% RRR per 1 mmol/L reduction in LDL-C observed in RCTs of lipid-lowering agents.4 Model B estimates the benefit from stabilization of advanced plaques as well as prevention of lesion formation. It is based on the formula proposed by Ference et al,5 which interpolates the 22% RRR per 1 mmol/L reduction in LDL-C over 5 years from RCTs and the 54% RRR over 40 years from Mendelian randomization studies. The resulting formula is a function of anticipated decrease in LDL-C as well as treatment duration. In the primary analysis, we assumed lipid-lowering strategy consistent with 40% reduction in LDL-C, which corresponds to a blended effect of high- and moderate-intensity statin (50% and 30% reductions, respectively), and in sensitivity analyses we evaluated each effect separately. Expected absolute risk reduction (ARR or benefit attributable to lipid lowering) in estimated 30-year risk was calculated by multiplying the risk that could be reduced by lipid-lowering therapy and the corresponding RRR because the benefit is a function of baseline risk and baseline level of LDL-C.9,10 Because the risk model allows estimation for any given subset of time and the RRR formula is based on treatment duration, when combined, they allow estimation of the benefit attributable to lipid lowering as a function of delay in treatment initiation. For example, the expected ARR in estimated 30-year risk for someone with treatment delayed by 10 years is calculated as the risk between years 11 and 30 multiplied by RRR with 20 years of treatment (30 – 10).

The fraction of risk avoided was calculated as the ratio of ARR and 30-year risk. This is different from the RRR, because it is influenced by treatment duration. Numbers of events expected and potentially avoided were estimated by multiplying the number of US individuals in each age–cholesterol group combination by the estimated 30-year risk and ARR, respectively. All analyses were performed using SAS version 9.4 (SAS Institute Inc).

Results

Table 1 presents the distribution of risk factors according to age group (30 to 39, 40 to 49, 50 to 59 years) and non–HDL-C category (<130, 130 to 159, ≥160 mg/dL), adjusting for population sampling weights. The mean 10-year risk for those >40 years is well below the treatment threshold level. Cholesterol levels are fairly stable across age categories, whereas systolic blood pressure is higher in older age. The predicted 30-year risk of ASCVD is higher in older age and higher non–HDL-C level, from a mean of 4.2% (95 CI, 3.9%–4.5%) among 30- to 39-year-old patients with non–HDL-C <130 mg/dL (3.38 mmol/L) mg/dL to a mean of 21.5% (95% CI, 20.4%–22.6%) among 50- to 59-year-old patients with non–HDL-C ≥160 mg/dL (4.16 mmol/L).

Table 1. Baseline Characteristics by Age Group and Non–High-Density Lipoprotein Cholesterol (Non–HDL-C) Category

Age Group, y, and VariablesNon–HDL-C Category, mg/dL
Non–HDL-C <130130 ≤ Non–HDL-C < 160Non–HDL-C ≥160
30 to 39
 Age, y34.1 (33.8, 34.4)34.4 (34.0, 34.7)34.8 (34.4, 35.2)
 Women, %63.4 (59.7, 67.1)44.1 (38.6, 49.7)28.9 (22.4, 35.3)
 LDL-C, mg/dL89.6 (88.2, 91.0)121.9 (120.4, 123.4)151.7 (149.4, 154.0)
 HDL-C, mg/dL56.3 (54.8, 57.8)50.5 (48.9, 52.1)47.5 (45.6, 49.4)
 Non–HDL-C, mg/dL105.0 (103.4, 106.5)144.8 (143.8, 145.8)183.6 (181.0, 186.3)
 Triglycerides, mg/dL77.0 (73.5, 80.4)114.7 (108.5, 120.9)159.7 (148.9, 170.5)
 Systolic blood pressure, mm Hg113.2 (112.0, 114.3)115.5 (113.9, 117.2)119.3 (117.4, 121.2)
 Hypertension TRT, %4.6 (2.4, 6.9)4.8 (2.3, 7.3)4.5 (1.5, 7.6)
 Smoking, %17.9 (14.2, 21.7)20.8 (16.0, 25.6)24.4 (18.7, 30.0)
 30-year cardiovascular risk, %4.2 (3.9, 4.5)6.8 (6.4, 7.3)11.0 (10.2, 11.8)
 10-year ASCVD risk, %NANANA
40 to 49
 Age, y43.9 (43.5, 44.3)44.7 (44.3, 45.1)44.3 (43.9, 44.6)
 Women, %64.5 (59.4, 69.6)46.8 (39.9, 53.6)45.0 (38.1, 52.0)
 LDL-C, mg/dL90.9 (88.9, 93.0)122.1 (120.4, 123.7)149.5 (147.2, 151.9)
 HDL-C, mg/dL60.9 (58.9, 63.0)53.6 (51.8, 55.4)49.6 (48.0, 51.1)
 Non–HDL-C, mg/dL106.6 (104.6, 108.5)143.1 (142.0, 144.3)181.2 (179.2, 183.1)
 Triglycerides, mg/dL78.2 (73.9, 82.5)105.3 (98.8, 111.8)158.3 (149.2, 167.5)
 Systolic blood pressure, mm Hg115.5 (113.4, 117.5)116.1 (114.5, 117.6)119.1 (117.3, 120.9)
 Hypertension TRT, %11.3 (7.4, 15.2)10.4 (6.0, 14.7)13.0 (8.2, 17.8)
 Smoking, %13.9 (9.4, 18.3)19.9 (14.1, 25.7)18.9 (13.5, 24.3)
 30-year cardiovascular risk, %7.9 (7.3, 8.5)12.6 (11.7, 13.5)17.1 (16.2, 18.1)
 10-year ASCVD risk, %1.0 (0.9, 1.1)1.8 (1.6, 2.0)2.6 (2.3, 2.8)
50 to 59
 Age, y53.6 (53.0, 54.3)54.0 (53.5, 54.5)54.3 (53.7, 54.9)
 Women, %62.6 (54.8, 70.4)68.2 (60.0, 76.5)74.6 (67.3, 81.9)
 LDL-C, mg/dL92.2 (88.7, 95.8)124.1 (122.0, 126.3)153.2 (150.5, 155.9)
 HDL-C, mg/dL65.0 (60.5, 69.6)60.5 (57.4, 63.5)57.2 (54.5, 59.8)
 Non–HDL-C, mg/dL108.0 (104.5, 111.5)144.0 (142.3, 145.7)183.0 (180.0, 186.0)
 Triglycerides, mg/dL79.1 (69.8, 88.4)99.1 (92.5, 105.6)149.0 (138.1, 160.0)
 Systolic blood pressure, mm Hg120.6 (118.1, 123.1)120.5 (118.2, 122.7)120.3 (118.5, 122.1)
 Hypertension TRT, %14.8 (8.6, 21.0)12.7 (7.4, 17.9)14.9 (9.9, 20.0)
 Smoking, %17.2 (11.2, 23.3)6.5 (1.5, 11.5)16.2 (10.0, 22.4)
 30-year cardiovascular risk, %13.9 (12.6, 15.2)16.3 (15.3, 17.3)21.5 (20.4, 22.6)
 10-year ASCVD risk, %2.8 (2.4, 3.2)2.9 (2.6, 3.2)3.6 (3.3, 3.9)

Results shown as mean ± 95% confidence limits. Participants with 10-year atherosclerotic cardiovascular disease (ASCVD) risk ≥7.5%, cardiovascular disease, diabetes mellitus, low-density lipoprotein cholesterol (LDL-C) ≥190 mg/dL, triglycerides ≥400 mg/dL, or on lipid-lowering therapy were excluded. TRT indicates anti-hypertensive treatment.

Figures 1 through 3 and Tables 2 through 4 illustrate how much the estimated 30-year ASCVD risk across age and non–HDL-C groups would be reduced if lipid-lowering therapy was initiated at different time points between today’s age and the age at the end of the 30-year period. This presentation adopts a 30-year perspective but could be recalculated for other time horizons. Estimates are based on the 2 main postulated models: A (benefit only from stabilization of preexisting plaque based on RCT data) and B (benefit interpolating effects observed in RCTs and Mendelian randomization studies). For example, the red horizontal line in panel 1 of Figure 1 shows the average predicted 30-year ASCVD risk in 30- to 39-year-old patients with non–HDL-C <130 mg/dL as 4.2%. This predicted risk could be reduced to 3.3% (model A; darker pink area) or 2.3% (model B; red area) if lipid lowering started immediately and would remain at 4.2% if treatment was delayed by the full 30 years. The residual risk that remains after lipid-lowering intervention is depicted by the red area for model B and red plus darker pink area for model A estimated as a function of delay in treatment. The light pink area below the red estimated 30-year risk line (model A) and light pink plus darker pink areas below the red line are the expected ARRs in the 30-year risk achieved with lipid lowering as a function of delay in initiation of lipid lowering. In our example, it is 0.9% (model A) or 1.9% (model B) if treatment starts immediately and lasts 30 years and decreases to 0 as treatment delay approaches 30 years. Accordingly, Table 2 presents the potential gains in reducing the estimated 30-year risk that result from initiating therapy at any point during the 30 years as well as the losses from delaying the initiation of lipid lowering according to the 2 postulated benefit models.

Table 2. Thirty-Year Risk Reduction as a Function of Delay in Initiation of Lipid Lowering by 40% (Age Range 30 to 39 Years)

Delay in TreatmentNon–HDL-C <130130 ≤ Non–HDL-C < 160Non–HDL-C ≥160
010203001020300102030
30-year risk4.24.24.24.26.86.86.86.811.011.011.011.0
 Model A
  Residual risk3.33.53.74.25.05.35.96.87.48.19.211.0
  ARR0.90.70.401.81.50.903.62.91.70
  % Risk avoided20.517.010.5026.922.213.6032.326.516.10
 Model B
  Residual risk2.32.93.64.23.14.35.76.84.16.28.811.0
  ARR1.91.20.603.72.51.106.94.72.20
  % Risk avoided43.929.413.2054.637.217.0062.543.119.90

Thirty-year cardiovascular disease risk = absolute risk reduction (ARR) + residual risk; ARR = 30-year cardiovascular disease risk × relative risk reduction. Participants with 10-year atherosclerotic cardiovascular disease risk ≥7.5%, cardiovascular disease, diabetes mellitus, low-density lipoprotein cholesterol ≥190 mg/dL, triglycerides ≥400 mg/dL, or on lipid-lowering therapy were excluded. Values are percentages. Non-HDL-C indicates non-high-density lipoprotein cholesterol.

Figure 1.

Figure 1. Thirty-year risk reduction as a function of delay in initiation of lipid lowering by 40% (baseline age range 30 to 39 years, 10-year atherosclerotic cardiovascular disease risk <7.5%). The x axis represents the delay in onset of lipid-lowering therapy to prevent atherosclerotic cardiovascular disease: 0 indicates treatment for the full 30 years; 30 represents no treatment over 30 years. The y axis depicts atherosclerotic cardiovascular disease risk over 30 years. The horizontal red line represents the 30-year risk (%) estimated at the initial age for the age group assuming no treatment during the subsequent 30 years. Model A assumes constant relative risk reduction over 30 years, whereas model B assumes increasing risk reduction over the same time interval. The light pink area under the horizontal red line represents the benefit (absolute risk reduction) of lipid lowering for model A for any given delay in the onset of treatment. The light pink + darker pink areas represent the benefit for model B after lipid lowering for any given delay in the onset of treatment. The red area shows the residual risk according to model B that remains as a function of delay in lipid lowering and the red + darker pink areas show the residual risk according to model A. Non-HDL-C indicates non-high-density lipoprotein cholesterol.

The potential reductions are more pronounced for higher age and higher non–HDL-C categories (Tables 3 and 4). For example, immediate initiation of lipid lowering in individuals 50 to 59 years of age with non–HDL-C ≥160 mg/dL (4.16 mmol/L; Table 4) would be expected to reduce their average predicted 30-year risk of 21.5% to 14.5% (model A) or 8.0% (model B), for an ARR of 7.0% (model A) or 13.5% (model B). Delaying lipid lowering by 10 years (20 years of treatment) would result in residual risk of 16.0% (ARR of 5.5%) according to model A or 12.5% (ARR of 9.0%) according to model B. Delaying by 20 years (10 years of treatment) leads to expected mean residual risk of 18.4% (ARR of 2.1%) based on model A and 17.6% (ARR of 3.9%) based on model B. We observe that the slope of the achieved ARR in predicted 30-year risk as a function of delay in treatment is also higher in older age groups (higher 30-year risk) and for higher non–HDL-C categories (more potential for risk reduction attributable to lipid lowering; Figures 1 through 3). Risks and benefits for 40-year-old patients with non–HDL-C >160 mg/dL are similar to those expected for 50-year-old patients with non–HDL-C between 130 and 160 mg/dL (30-year risk 17.1% versus 16.3%, respectively) and expected ARRs with immediate treatment of 5.5% versus 4.4%, respectively (model A) or 10.6% versus 9.0%, respectively (model B). As expected, the effect is lower when using model A (which assumes constant RRR of 22% across the entire follow-up) versus model B (RRR increases with treatment duration). The fraction of risk avoided is reduced by about 1/5th when modeling a 20-year delay in treatment initiation and cut in half with no delay.

Table 3. Thirty-Year Risk Reduction as a Function of Delay in Initiation of Lipid Lowering by 40% (Age Range 40 to 49 Years, 10-Year Atherosclerotic Cardiovascular Disease [CVD] Risk <7.5%)

Delay in TreatmentNon–HDL-C < 130130 ≤ Non–HDL-C < 160Non–HDL-C ≥ 160
010203001020300102030
30-year risk7.97.97.97.912.612.612.612.617.117.117.117.1
 Model A
  Residual risk6.26.67.17.99.29.911.012.611.612.714.617.1
  ARR1.71.40.803.42.71.605.54.42.60
  % Risk avoided20.817.010.3026.921.913.0031.925.815.20
 Model B
  Residual risk4.45.66.97.95.78.010.612.66.59.913.917.1
  ARR3.62.31.006.94.62.0010.67.23.20
  % Risk avoided44.429.513.0054.736.716.3062.042.118.80

Thirty-year CVD risk = absolute risk reduction (ARR) + residual risk; ARR = 30-year CVD risk × relative risk reduction. Participants with 10-year atherosclerotic CVD risk ≥7.5%, CVD, diabetes mellitus, low-density lipoprotein cholesterol ≥190 mg/dL, triglycerides ≥400 mg/dL, or on lipid-lowering therapy were excluded. Values are percentages. Non-HDL-C indicates non-high-density lipoprotein cholesterol.

Table 4. Thirty-Year Risk Reduction as a Function of Delay in Initiation of Lipid Lowering by 40% (Age Range 50 to 59 Years, 10-Year Atherosclerotic Cardiovascular Disease [CVD] Risk <7.5%)

Delay in TreatmentNon–HDL-C < 130130 ≤ Non–HDL-C < 160Non–HDL-C ≥ 160
010203001020300102030
30-year risk13.913.913.913.916.316.316.316.321.521.521.521.5
 Model A
  Residual risk10.911.512.613.911.912.814.316.314.516.018.421.5
  ARR3.02.31.304.43.52.007.05.53.10
  % Risk avoided21.116.89.7027.321.912.7032.525.914.80
 Model B
  Residual risk7.69.812.213.97.310.413.816.38.012.517.621.5
  ARR6.34.11.709.05.92.5013.59.03.90
  % Risk avoided44.829.112.3055.336.715.9062.942.118.30

Thirty-year CVD risk = absolute risk reduction (ARR) + residual risk; ARR = 30-year CVD risk × relative risk reduction. Participants with 10-year atherosclerotic CVD risk ≥7.5%, CVD, diabetes mellitus, low-density lipoprotein cholesterol ≥190 mg/dL, triglycerides ≥400 mg/dL, or on lipid-lowering therapy were excluded. Values are percentages. Non-HDL-C indicates non-high-density lipoprotein cholesterol.

Table 5 translates the risks and reductions into expected population counts. It presents the expected number of ASCVD events in the next 30 years as well as the number of events potentially avoided as a function of lipid-lowering delay. As expected, the largest numbers of events potentially avoided occur in the oldest age groups, for those with the highest non–HDL-C levels, and with the least delay in lipid lowering.

Table 5. Expected Number of 30-Year Events Avoided as a Function of Delay in Initiation of Lipid Lowering by 40%

Treatment delayNon–HDL-C < 130130 ≤ Non–HDL-C < 160Non–HDL-C ≥ 160
010203001020300102030
Age 30 to 39 years
 Population size16 115 0259 991 9877 476 465
 Expected events672 749683 454820 265
 Events avoided
  Model A141 121116 20671 0910182 052149 54591 0700266 037216 986130 3790
  Model B300 863200 93089 7450370 674251 300113 8360514 776353 158161 1620
Age 40 to 49 years
 Population size11 805 1559 937 4028 978 368
 Expected events933 6431 252 1131 536 473
 Events avoided
  Model A197 128160 25695 5230336 477271 763159 9860491 530394 645229 8520
  Model B419 837276 925120 5620683 790456 141199 8980954 678643 849284 3640
Age 50 to 59 years
 Population size5 643 9746 851 0896 441 861
 Expected events784 0571 118 3311 385 199
 Events avoided
  Model A167 717132 62274 7710304 488243 059139 6600450 343356 084201 0070
  Model B356 205228 77894 3140617 190407 307174 4030871 337579 518248 4560

Participants with 10-year atherosclerotic cardiovascular disease risk ≥7.5%, cardiovascular disease, diabetes mellitus, low-density lipoprotein cholesterol ≥190 mg/dL, triglycerides ≥400 mg/dL, or on lipid-lowering therapy were excluded. HDL-C indicates high-density lipoprotein cholesterol.

The results remain directionally similar (with even larger magnitude of effects) in sensitivity analyses where we exclude only individuals with estimated 10-year ASCVD risk ≥20% (Table IA and IB in the Data Supplement). Similarly, the effects of modeling high- and moderate-intensity statin–induced RRR lead to directionally similar and proportionally larger or smaller effects (Table IIA and IIB and Table IIIA and IIIB in the Data Supplement, respectively).

Discussion

The current paradigm for lipid-lowering recommendations is based on the decision whether lipid-lowering drug therapy should be initiated. This work provides tools to understand how to change the paradigm to “when,” rather than “if,” intensive lipid lowering should be initiated. Our analysis indicates that the absolute benefit from treatment for 30 years with lipid-lowering therapy of a 30-year-old patient with no current indications for lipid-lowering drug therapy and low levels of non–HDL-C is small. By contrast, the expected benefit from 30 years of treatment starting at age 50 years in the high non–HDL-C subgroup is substantial, even in the face of a low predicted 10-year risk. However, the later the starting age, the greater the number of events that will have occurred before therapy has begun. The expected 30-year benefit of treating 40-year-old patients with non–HDL-C ≥160 mg/dL (4.16 mmol/L) appears substantial. The consequences of delay in lipid-lowering therapy initiation to age 50 or 60 years rather than age 40 years are also substantial, as is evident from Table 4.

We modeled the potential benefits of lipid lowering using 2 different assumptions. In model A, benefit was estimated based on the 22% RRR per 1 mml/L reduction in LDL-C observed in the RCTs of statins (based on ≤5 years of treatment), whereas in model B, benefit was estimated based on interpolation of the 54% RRR over 40 years from Mendelian randomization with the estimate of benefit from the statin trials. In both models, in all scenarios, benefit is greater with earlier intervention and with higher levels of non–HDL-C. Furthermore, as expected, the benefit predicted by model B exceeds the benefit predicted by model A; for example, immediate initiation of lipid lowering can reduce 30-year risk of ASCVD by about a quarter according to model A and by about half according to model B.

Trapping of apoB particles within the arterial wall leads to the formation of foam cells, to the release of bioactive proinflammatory molecules such as oxidized phospholipids, which provoke a complex cellular and biological response, to the induction of immune responses, as well as a variety of other biological responses and outcomes including plaque formation, calcification, neovascularization, and the deposition of extracellular crystalline cholesterol. Trapping of apoB particles also plays a critical role in provoking the changes within the arterial wall or the endothelium that are the immediate progenitors of a clinical event, such as plaque rupture and endothelial erosion.5,11,12

Reducing apoB particle number in plasma will reduce the rate at which all these events in the progression of atherothrombosis occur. Reducing trapping of apoB particles will also tend to restore normal endothelial function and shift the cellular balance within the arterial wall from inflammatory cells toward smooth muscle cells, promoting enhancement of fibrous tissue around plaques and, therefore, reducing the chance of plaque rupture.5,11,12 The more advanced the disease already present within any arterial segment, the fewer new lesions can be created. In the statin randomized trials, the average age of the participants was generally in the 6th and 7th decades or older.4 Accordingly, the positive effects of statins on disease progression and lesion stabilization within the arterial wall are presumably the principal bases for their favorable clinical effects. Thus, our model A, which intends to capture the benefit of plaque stabilization, uses RRR from statin RCTs.

By contrast, model B, which also intends to capture the benefit of prevention of new lesion formation, interpolates between statin RCTs4 and Mendelian randomization results.5 The complex lesions that begin to appear in the later 3rd and early 4th decade of life are the outcomes of the intricate and multifaceted sequences of biological reactions to the trapping of apoB particles within the arterial wall that have taken place over this period. However, the lower the number of apoB particles in plasma, the fewer apoB particles that will enter and be trapped within the arterial wall, and the lower should be the rate at which new lesions will be formed.13 This presumption is implicit based on the potent relations between the concentrations in plasma of the markers of the apoB lipoproteins—LDL-C, non–HDL-C, and apoB—with cardiovascular risk.13

RCTs have demonstrated that statins can reduce, but not eliminate, clinical events in patients already at higher risk for ASCVD events. Limited benefit later in the disease process should be anticipated because statins at most can modify the abnormalities produced by atherosclerosis but cannot entirely reverse them. By contrast, to the extent formation of complex lesions is reduced, prevention of later clinical events by statins should be complete. The results of Mendelian randomization studies are consistent with this hypothesis.5 So are the results of a meta-analysis of nearly 900 000 healthy adults, which demonstrated a much steeper relation between total cholesterol and risk in subjects 40 to 49 years of age compared with those 70 to 89 years of age.14 Similar findings have been reported for LDL-C, non–HDL-C, and apoB.15 Moreover, a meta-analysis by Law et al16 demonstrated an inverse relationship between increasing age and expected reduction in event rate on statin treatment. Finally, an extensive recent review of trials of LDL-C lowering demonstrated that the benefit per mmol/L lowering of LDL-C is greater if initiated at an earlier age,17 a finding consistent with the degree of benefit being influenced by the stage of disease. Therefore, assigning a greater RRR to lipid lowering earlier in life, as assumed by model B does, is not unreasonable.

Atherosclerosis proceeds at different rates in different individuals. Nevertheless, in general, those who are younger will have less established disease than those who are older. The formula to estimate benefit in model B uses treatment duration, not age at initiation. In reality, these 2 factors are intertwined: earlier initiation allows longer duration and longer duration usually implies earlier initiation. Accordingly, model B may be most appropriate for the 30 to 39 years age group but overestimate benefit in the 50 to 59 years age group.

In our results, we present data aggregated across categories of age and cholesterol. However, personalized treatment decisions should be made based on individual estimates of risk and potential treatment benefit, which are easily obtained using our approach and can be plotted over the next 30 years. Figures 1 through 3 can then serve as benchmarks with which to contrast the personalized estimates of risk and benefit. The time horizon can also be changed from 30 years to whatever the intended goal might be. We have shown previously that the levels of non–HDL-C between ages 35 and 40 years will identify with considerable precision a subgroup of individuals whose levels of non–HDL-C will remain high over the next 30 years and who, as a group, are at substantial risk of cardiovascular disease.18 This group, which represents about 20% of the population, represents a potential target for earlier intervention. For any patient to make an informed decision requires that the potential benefit of early intervention be quantitated. Age and the level of the atherogenic lipoproteins are essential elements to an informed decision.

Our work has several limitations. First, this is a modeling study, based on the estimates of 30-year risk calculated using an established tool7 and estimates of benefit from lipid lowering calculated using a tool proposed by the European Society of Cardiology/European Atherosclerosis Society, which incorporates results of statin RCTs and Mendelian randomization studies. Second, the analysis necessarily involves extrapolation of the results of RCTs to longer periods of time because no trials of lipid lowering over extended periods of time have been carried out. Third, whereas our results apply to any intensive lipid-lowering modality, including dietary and other lifestyle interventions, we realize that pharmacologic intervention with statins will be the most common approach. Statins are associated with side effects that vary in frequency and severity and these must be taken into account in formulating any pharmacologic intervention strategy.19 Fourth, we estimate benefit for a fixed duration of lipid lowering. An alternative approach would estimate benefit until fixed age, leading to even higher expected benefit for the younger age groups.

These data suggest that substantial intermediate- and long-term ASCVD risk reduction can be achieved by aggressive lipid lowering in early to middle adulthood, when short-term risks are low, especially in those with non–HDL-C levels >160 mg/dL. We created an instrument that allows estimation of losses associated with delay of lipid-lowering therapy, according to 2 different models, which can be used for the selection of the preferred time for initiation of lipid lowering in the next 30 years. Such information is essential if the patient is to make an informed decision about preventive lipid-lowering therapy, as recommended by the current American College of Cardiology/American Heart Association guidelines.

Figure 2.

Figure 2. Thirty-year risk reduction as a function of delay in initiation of lipid lowering by 40% (baseline age range 40 to 49 years, 10-year atherosclerotic cardiovascular disease risk <7.5%). The x axis represents the delay in onset of lipid-lowering therapy to prevent atherosclerotic cardiovascular disease: 0 indicates treatment for the full 30 years; 30 represents no treatment over 30 years. The y axis depicts atherosclerotic cardiovascular disease risk over 30 years. The horizontal red line represents the 30-year risk (%) estimated at the initial age for the age group assuming no treatment during the subsequent 30 years. Model A assumes constant relative risk reduction over 30 years, whereas model B assumes increasing risk reduction over the same time interval. The light pink area under the horizontal red line represents the benefit (absolute risk reduction) of lipid lowering for model A for any given delay in the onset of treatment. The light pink + darker pink areas represent the benefit for model B after lipid lowering for any given delay in the onset of treatment. The red area shows the residual risk according to model B that remains as a function of delay in lipid lowering and red + darker pink areas show the residual risk according to model A. Non-HDL-C indicates non-high-density lipoprotein cholesterol.

Figure 3.

Figure 3. Thirty-year risk reduction as a function of delay in initiation of lipid lowering by 40% (baseline age range 50 to 59 years, 10-year atherosclerotic cardiovascular disease risk <7.5%). The x axis represents the delay in onset of lipid-lowering therapy to prevent atherosclerotic cardiovascular disease: 0 indicates treatment for the full 30 years; 30 represents no treatment over 30 years. The y axis depicts atherosclerotic cardiovascular disease risk over 30 years. The horizontal red line represents the 30-year risk (%) estimated at the initial age for the age group assuming no treatment during the subsequent 30 years. Model A assumes constant relative risk reduction over 30 years, whereas model B assumes increasing risk reduction over the same time interval. The light pink area under the horizontal red line represents the benefit (absolute risk reduction) of lipid lowering for model A for any given delay in the onset of treatment. The light pink + darker pink areas represent the benefit for model B after lipid lowering for any given delay in the onset of treatment. The red area shows the residual risk according to model B that remains as a function of delay in lipid lowering and red + darker pink areas show the residual risk according to model A. Non-HDL-C indicates non-high-density lipoprotein cholesterol.

Supplemental Materials

Data Supplement Tables I–III

Footnotes

Sources of Funding, see page 836

https://www.ahajournals.org/journal/circ

The Data Supplement, podcast, and transcript are available with this article at https://www.ahajournals.org/doi/suppl/10.1161/CIRCULATIONAHA.120.045851.

Continuing medical education (CME) credit is available for this article. Go to http://cme.ahajournals.org to take the quiz.

Allan D. Sniderman, MD, Mike and Valeria Rosenbloom Centre for Cardiovascular Prevention, McGill University Health Centre, Royal Victoria Hospital, Glen Site C04.4180, 1001 Boulevard Décarie, Montréal, Québec H4A 3J1, Canada. Email

References

  • 1. Sniderman AD, Thanassoulis G, Williams K, Pencina M. Risk of premature cardiovascular disease vs the number of premature cardiovascular events.JAMA Cardiol. 2016; 1:492–494. doi: 10.1001/jamacardio.2016.0991CrossrefMedlineGoogle Scholar
  • 2. Mortensen MB, Nordestgaard BG. 2019 vs. 2016 ESC/EAS statin guidelines for primary prevention of atherosclerotic cardiovascular disease [published online March 30, 2020].Eur Heart J. doi: 10.1093/eurheartj/ehaa150. https://europepmc.org/article/med/32227172CrossrefGoogle Scholar
  • 3. Stary HC, Chandler AB, Dinsmore RE, Fuster V, Glagov S, Insull W, Rosenfeld ME, Schwartz CJ, Wagner WD, Wissler RW. A definition of advanced types of atherosclerotic lesions and a histological classification of atherosclerosis: a report from the Committee on Vascular Lesions of the Council on Arteriosclerosis, American Heart Association.Circulation. 1995; 92:1355–1374. doi: 10.1161/01.cir.92.5.1355LinkGoogle Scholar
  • 4. Collins R, Reith C, Emberson J, Armitage J, Baigent C, Blackwell L, Blumenthal R, Danesh J, Smith GD, DeMets D, et al. Interpretation of the evidence for the efficacy and safety of statin therapy.Lancet. 2016; 388:2532–2561. doi: 10.1016/S0140-6736(16)31357-5CrossrefMedlineGoogle Scholar
  • 5. Ference BA, Ginsberg HN, Graham I, Ray KK, Packard CJ, Bruckert E, Hegele RA, Krauss RM, Raal FJ, Schunkert H, et al. Low-density lipoproteins cause atherosclerotic cardiovascular disease: 1: evidence from genetic, epidemiologic, and clinical studies: a consensus statement from the European Atherosclerosis Society Consensus Panel.Eur Heart J. 2017; 38:2459–2472. doi: 10.1093/eurheartj/ehx144CrossrefMedlineGoogle Scholar
  • 6. Grundy SM, Stone NJ, Bailey AL, Beam C, Birtcher KK, Blumenthal RS, Braun LT, de Ferranti S, Faiella-Tommasino J, Forman DE, et al. 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA guideline on the management of blood cholesterol: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines.Circulation. 2019; 139:e1082–e1143. doi: 10.1161/CIR.0000000000000625LinkGoogle Scholar
  • 7. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge.Clin Chem. 1972; 18:499–502.CrossrefMedlineGoogle Scholar
  • 8. Pencina MJ, D’Agostino RB, Larson MG, Massaro JM, Vasan RS. Predicting the 30-year risk of cardiovascular disease: the Framingham Heart Study.Circulation. 2009; 119:3078–3084. doi: 10.1161/CIRCULATIONAHA.108.816694LinkGoogle Scholar
  • 9. Soran H, Schofield JD, Durrington PN. Cholesterol, not just cardiovascular risk, is important in deciding who should receive statin treatment.Eur Heart J. 2015; 36:2975–2983. doi: 10.1093/eurheartj/ehv340CrossrefMedlineGoogle Scholar
  • 10. Thanassoulis G, Williams K, Altobelli KK, Pencina MJ, Cannon CP, Sniderman AD. Individualized statin benefit for determining statin eligibility in the primary prevention of cardiovascular disease.Circulation. 2016; 133:1574–1581. doi: 10.1161/CIRCULATIONAHA.115.018383LinkGoogle Scholar
  • 11. Borén J, Williams KJ. The central role of arterial retention of cholesterol-rich apolipoprotein-B-containing lipoproteins in the pathogenesis of atherosclerosis: a triumph of simplicity.Curr Opin Lipidol. 2016; 27:473–483. doi: 10.1097/MOL.0000000000000330CrossrefMedlineGoogle Scholar
  • 12. Borén J, Chapman MJ, Krauss RM, Packard CJ, Bentzon JF, Binder CJ, Daemen MJ, Demer LL, Hegele RA, Nicholls SJ, et al. Low-density lipoproteins cause atherosclerotic cardiovascular disease: pathophysiological, genetic, and therapeutic insights: a consensus statement from the European Atherosclerosis Society Consensus Panel.Eur Heart J. 2020; 338:1650. doi: 10.1093/eurheartj/ehz962CrossrefGoogle Scholar
  • 13. Sniderman AD, Thanassoulis G, Glavinovic T, Navar AM, Pencina M, Catapano A, Ference BA. Apolipoprotein B particles and cardiovascular disease: a narrative review.JAMA Cardiol. 2019; 4:1287–1295. doi: 10.1001/jamacardio.2019.3780CrossrefMedlineGoogle Scholar
  • 14. Prospective Studies Collaboration, Lewington S, Whitlock G, Clarke R, Sherliker P, Emberson J, Halsey J, Qizilbash N, Peto R, Collins R. Blood cholesterol and vascular mortality by age, sex, and blood pressure: a meta-analysis of individual data from 61 prospective studies with 55,000 vascular deaths.Lancet. 2007; 370:1829–1839. doi: 10.1016/S0140-6736(07)61778-4CrossrefMedlineGoogle Scholar
  • 15. Sniderman AD, Islam S, McQueen M, Pencina M, Furberg CD, Thanassoulis G, Yusuf S. Age and cardiovascular risk attributable to apolipoprotein B, low-density lipoprotein cholesterol or non-high-density lipoprotein cholesterol.J Am Heart Assoc. 2016; 5:e003665. doi: 10.1161/JAHA.116.003665LinkGoogle Scholar
  • 16. Law MR, Wald NJ, Rudnicka AR. Quantifying effect of statins on low density lipoprotein cholesterol, ischaemic heart disease, and stroke: systematic review and meta-analysis.BMJ. 2003; 326:1423. doi: 10.1136/bmj.326.7404.1423CrossrefMedlineGoogle Scholar
  • 17. Wang N, Fulcher J, Abeysuriya N, Park L, Kumar S, Di Tanna GL, Wilcox I, Keech A, Rodgers A, Lal S. Intensive LDL cholesterol-lowering treatment beyond current recommendations for the prevention of major vascular events: a systematic review and meta-analysis of randomised trials including 327 037 participants.Lancet Diabetes Endocrinol. 2020; 8:36–49. doi: 10.1016/S2213-8587(19)30388-2CrossrefMedlineGoogle Scholar
  • 18. Pencina KM, Thanassoulis G, Wilkins JT, Vasan RS, Navar AM, Peterson ED, Pencina MJ, Sniderman AD. Trajectories of non-HDL cholesterol across midlife: implications for cardiovascular prevention.J Am Coll Cardiol. 2019; 74:70–79. doi: 10.1016/j.jacc.2019.04.047CrossrefMedlineGoogle Scholar
  • 19. Newman CB, Preiss D, Tobert JA, Jacobson TA, Page RL, Goldstein LB, Chin C, Tannock LR, Miller M, Raghuveer G, et al; American Heart Association Clinical Lipidology, Lipoprotein, Metabolism and Thrombosis Committee, a Joint Committee of the Council on Atherosclerosis, Thrombosis and Vascular Biology and Council on Lifestyle and Cardiometabolic Health; Council on Cardiovascular Disease in the Young; Council on Clinical Cardiology; and Stroke Council. Statin safety and associated adverse events: a scientific statement from the American Heart Association.Arterioscler Thromb Vasc Biol. 2019; 39:e38–e81. doi: 10.1161/ATV.0000000000000073LinkGoogle Scholar

eLetters(0)

eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.