UPDATE: ASCVD risk factor critique

I decided to do an extensive update of my prior ASCVD (atherosclerotic cardiovascular disease) risk analysis and critique of risk calculators, with much more information about the “non-traditional” risk factors. I felt that there should be more data/clarification, since I think we are really undertreating/not preventing the most common cause of death by relying on the current ASCVD risk calculators

 

-- Cardiovascular risk models:

    -- many of us rely on the various cardiovascular risk models to determine the appropriate approach to patients, and these are often integrated into the electronic medical records

    -- these risk models have several very important limitations:

        -- they are typically based on community data, eg from the Framingham study etc. But, in clinical practice, we are dealing with the individual and not the community:

            -- there is very clear evidence that there are many individual risk factors that are associated with increased risk of a cardiovascular event.

                  -- In terms of medical risk factors: people with rheumatoid arthritis, psoriasis, infections associated with chronic inflammation (eg HIV; even if well-controlled hepatitis), migraine, peripheral arterial disease, etc are all associated with about a two-fold increased risk of a cardiovascular event. But, even in the cases where these have been assessed on a population basis, their prevalence is low enough to not increase the community-based population risk as compared to the big four: diabetes, hypertension, smoking, and hyperlipidemia. this all means that adding them to the Framingham risk model, for example, would not add significantly to these models' predictive accuracy. ie: when tested to see if they add to the risk assessment on a population basis, they don't make the cut.

                -- there are a variety of "non -medical" risk factors that may be even more important than the "medical" ones, eg stress, depression, socioeconomic issues, environmental issues (eg air pollution), etc. And these are not include in the risk calculators that we use. See below

                -- also, the included risk factors in the calculators often have a binary assessment: people either have diabetes or not, are smoking or not….

                     -- The patients we see have very different, nonbinary situations: there is a difference in cardiovascular risk between a patient with an A1c of 7% versus one with 14%. Or between those with passive smoking or who stopped within the past 30 years vs nonsmokers

                -- and, the accuracy of the risk calculators may not hold up, with striking divergence between the AHA risk calculator and analysis of actual cohort studies: http://gmodestmedblogs.blogspot.com/2014/10/yet-another-analysis-that-lipid.html , and http://gmodestmedblogs.blogspot.com/2018/06/acc-guidelines-overestimate-cad-risk.html

 

-- standard risk factors for ASCVD risk calculators:

    -- lipids: 

        -- it is pretty clear that LDL is a complex and likely surrogate cardiovascular risk factor: http://gmodestmedblogs.blogspot.com/2023/05/cad-presumed-mechanism-and-apob-was.html )

            -- LDL is a group of particles that on electrophoresis is arbitrarily divided into those that are low-density versus those that are high density. Both groups, however, have several different particles in these two electrophoretic patterns

            -- epidemiologic studies suggest that small, dense LDL particles are three times more atherogenic than larger LDL particles

            -- these small, dense LDL particles are much more easily oxidized, and oxidized LDL is the one that binds to receptors on macrophages, creates foam cells, and these seem to ultimately develop into more lipid accumulation and atherosclerotic plaques. this distinction between oxidized and non-oxidized LDL also applies to Lp(a), another bad actor

            -- we have various mechanisms to assess LDL size, since we do not have access to direct laboratory assessment:

                -- certain conditions are associated with small, dense LDL, such as diabetes and metabolic syndrome. Therefore, it is likely that pretty much all people with diabetes should be on a statin, as well as the standard and all-important nonpharmacologic therapies regarding diet, exercise, and weight loss as appropriate. And the American Diabetes Association does in fact recommend statins at age 40, independent of LDL levels, if patients have one or more ASCVD risk factors

                -- the measurement of apolipoprotein B levels also gives a good estimation of LDL size, since there is one apolipoprotein B on the surface of each LDL particle; so for any LDL concentration measured, those with a significantly higher apolipoprotein B level are at higher risk of having small, dense LDL particles (measuring apolipoprotein B levels are recommended in the European guidelines, but not yet in the American guidelines)

                    -- there also is complexity in these apolipoproteins: some LDL and HDL particles have apolipoprotein C3 instead of their usual ones (usually apoB for LDL and apoA for HDL), and apoC3 is actually proatherogenic!!! (see https://www.ahajournals.org/doi/10.1161/CIRCULATIONAHA.111.056986 )  

                    -- it was notable, for example, that a clinical trial of the CETP inhibitor torcetrapib dramatically increased HDL levels by 72% but was associated with a 25% increase in cardiovascular events (https://www.nejm.org/doi/full/10.1056/nejmoa0706628 ), likely because of the associated apolipoprotein C3 on the HDL particles (see https://www.ahajournals.org/doi/10.1161/ATVBAHA.121.317181 )

                -- there are other measurements that may be useful in differentiating LDL size, such as having higher triglyceride levels, which are associated with a greater concentration of smaller, dense LDL particles (studies suggesting that the triglyceride/HDL level is a strong predictor: 79% of patients having a ratio of >3.8  have small, dense LDLs)

            -- there are also pretty impressive data that the lower the total LDL (even as a surrogate marker), the better, with documentation in observational studies that an LDL<20mg/dL continues to add benefit with minimal risk (http://gmodestmedblogs.blogspot.com/2018/08/very-low-ldl-levels-benefit-without-harm.html , as well as the recently released further analysis of the Fourier Study: see cad high risk LDL less than 20 fourier studyCirc2023 in dropbox, or DOI: 10.1161/CIRCULATIONAHA.122.063399 

 

    -- hypertension: 

        -- It is clear that there are real differences in cardiovascular risk as well as other bad outcomes with higher blood pressures, especially if the BP is uncontrolled. 240/120 is worse than 142/70. There may well be important other aspects not in the risk calculators, such as the length of time a patient is at different blood pressure levels, or the BP effects of different medication regimens

            -- there may well be important differences in which medications are used. Hydrochlorothiazide as a single agent, for example, is probably the worst choice given its short half-life and incomplete blood pressure control during the night: http://gmodestmedblogs.blogspot.com/2023/01/hypertension-hctz-vs-chlorthalidone.html. And, amlodipine may be the best given that it is associated with the least blood pressure variability over the 24-hour period, and blood pressure variability itself seems to be an independent risk factor for cardiovascular events. Though there are no definitive head-to-head comparisons of the different medications, we know that there is much higher predictive value of 24-hour ambulatory blood pressure monitoring over clinic-based measurements for actual cardiovascular events. so it seems that this conclusion that hydrochlorothiazide is not a good solo agent for BP control is justified 

            -- and b-blockers may be less clinically effective

        -- and, as with all of these risk factors, nonpharmacologic therapies are extremely important  

 

    -- diabetes: 

        -- there is typically a binary classification: one has diabetes or doesn’t, but we know that cardiovascular events increase with higher A1C levels  (https://pubmed.ncbi.nlm.nih.gov/10938048/)

        -- from the recent literature, some medications used to treat diabetes are superior from a cardiac perspective: some seem to make cardiovascular disease worse (insulin, sulfonylureas), and some seem to decrease cardiovascular disease (GLP-1's and SGLT-2’s). this is not incorporated into any of the ASCVD risk calculators

         -- diabetes is also more complex than appears: not just the issue of binary classification, as noted above. there are several studies finding that “prediabetes”, even with A1C levels as low as 5.5% (especially in men), is associated with significant ASCVD risk, as noted in http://gmodestmedblogs.blogspot.com/2022/11/prediabetes-increases-risk-heart.html and in https://pubmed.ncbi.nlm.nih.gov/25929925/.  Of note, the “cutpoint” of the A1c of 6.5% to “define” diabetes is based on the inflection point of increasing retinopathy. And most people with diabetes (70-80%) die from cardiovascular causes!! So, it seems that “prediabetes” should be considered a risk factor??

 

    -- smoking:

        -- also a binary classification in the ASCVD risk models. Smoking 2 packs/day is actually different than 3 cigarettes/week… (and has different clinical effects)

        -- There is no incorporation of passive smoking into these risk models, though that may increase the coronary heart disease risk by about 30% (https://www.ahajournals.org/doi/10.1161/circulationaha.104.492215 )

        -- there also are suggestions that it takes about 30 years of smoking cessation to decrease the cardiovascular risk of smoking: https://www.sciencedirect.com/science/article/pii/S073510971935449X . so details on smoking history, even in those who no longer smoke, are important

        -- smoking seems to have potential adverse effects, even smoking only one cigarette a day seems to have about half of the attributable risk of smoking a pack a day: http://gmodestmedblogs.blogspot.com/2018/01/smoking-just-one-cigarette-one-too-many.html. This supports educating patients who smoke that the goal really needs to be zero cigarettes

        -- so, it is pretty clear that smoking is a potent cardiovascular risk factor (the most important in several studies), and that we should continue helping patients stop smoking completely whenever possible

 

---------------------------------------------------

-- Other medical risk factors not in the current risk factor calculators, many of them likely because they are not prevalent enough in the community to make their presence statistically valid.  of note, the recurrent theme here is the association of these conditions with chronic systemic inflammation 

 

    -- visceral (central) obesity:

        -- the cardiometabolic effects of visceral obesity are significant, presumably related to the associated increased systemic inflammation. Prior studies have not found similar cardiovascular associations with non-visceral obesity. Which means that:

            -- BMI is not necessarily an accurate cardiac risk factor, though BMI is a pretty good surrogate marker

            -- abdominal obesity does correlate strongly with visceral obesity. So, the waist circumference is more indicative of abdominal obesity than BMI

            -- hard to assess the actual specific risk from visceral obesity: there are overall strong associations between visceral obesity with pretty much all of the traditional risk factors (hypertension, prediabetes/diabetes, hypercholesterolemia, hypertriglyceridemia, metabolic syndrome, with odds ratios from 2.5-fold to 15-fold (see https://www.nature.com/articles/s41598-021-88587-9 ))

            -- the 3rd NHANES study with 15K adults: normal weight people (BMI 22) with central obesity: twice the mortality risk vs normal weight people without, more so in men (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4995595/pdf/nihms811228.pdf ). of note, they did not control for obesity-related cardiovascular risk factors since "epidemiologic obesity research has shown that it might be inappropriate for controlling for factors in the causal pathway between obesity and death, such as diabetes, atherogenic dyslipidemia and hypertension": the point here is that obesity is highly associated with almost all of the standard cardiovascular risk factors, so trying to disentagle the association mathematically is fraught

 

    -- other medical risk factors are important to consider on an individual basis: chronic infections (eg hepatitis, HIV), migraine, peripheral arterial disease, rheumatologic diseases (lupus, rheumatoid arthritis, psoriasis, etc), depression, air pollution. The common denominator for most of these (and likely other conditions) is chronic systemic inflammation 

 

    -- HIV:

        -- HIV disease is associated with increased ASCVD risk, on the order of 1.5 to 2-fold increase in acute myocardial infarction or coronary heart disease (see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3964878/pdf/nihms-571007.pdf , for example. however, this association is complex: could be that much of the actual association might be from some of the meds used (eg abacavir, most protease inhibitors), the standard risk factors, other risk factors (stress etc as noted below). Those with well-controlled HIV seem to have a lower risk, though it remains above that of the general population.  A notable recent study on HIV treatment found that those with a quite “low” 10-year risk (4.5%) and mean LDL of 108, when given pitavastatin 4mg, had a pretty striking 35% reduced risk of major adverse cardiovascular effects over just 5 years (see http://gmodestmedblogs.blogspot.com/2023/09/hiv-infection-pitavastatin-dec-cardiac.html). This somewhat obliquely supports the contention that persons with HIV should be treated more aggressively to prevent cardiovasc events than their non-HIV counterparts

 

    -- Chronic Kidney Disease:

      -- both reduced GFR and proteinuria are independently associated with increased cardiovascular events in observational studies, and the combo of these renal problems increases the risk further

            -- for example, a large general population cohort (see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3993088/pdf/nihms-544166.pdf ) found that:

                --  an eGFR of 60 confers an 18% increased all-cause mortality and also cardiovascular mortality, and an eGFR of 45 confers a 57% increase

                -- albuminuria with an albumin-to-creatinine ratio of just 10 is associated with a 20% increase, and a microalbumn of 30 has a 63% increase

                -- and the combo of increased eGFR and albuminuria has a significantly higher cardiovasc risk, with increases as either or both the eGFR or albuminuria levels increased

            -- several other studies have come to similar conclusions

 

    -- Psoriasis:

        -- several  studies have noted increased cardiovascular risk with psoriasis (eg see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8168628/ )

            -- a large UK study found that MI incidence  increased stepwise with psoriasis severity: MI per 1000 people went from 3.58 in controls to 4.04 if mild psoriasis to 5.13 with severe psoriasis (JAMA. 2006;296(14):1735)

            -- a US study of >3000 patients vs 2500 controls found about a 2-fold risk for ischemic vascular disease, cerebrovascular disease, and peripheral vascular disease, as will as overall-mortality of about 2-fold

 

    -- Rheumatoid Arthritis/Lupus

        -- rheumatoid arthritis has been found in several studies to lead to a 1.5-2-fold increased risk of coronary artery disease (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3890244/ )

        -- cardiovascular disease is the leading cause of death in people with lupus, with 2- to 3-fold cardiovascular risk (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10012401/)

 

    -- Migraine

        -- migraine is associated with a 2- to 3-fold increased risk of angina and stroke: https://www.frontiersin.org/articles/10.3389/fcvm.2022.1044465/full#:~:text=Association%20of%20migraine%20with%20cardiovascular%20disease&text=From%20the%20perspective%20of%20specific,9.38)%20(Table%203, or doi.org/10.3389/fcvm.2022.1044465).  migraine is associated with neuroinflammation: https://thejournalofheadacheandpain.biomedcentral.com/articles/10.1186/s10194-021-01271-1 

 

    -- Air pollution:

        -- many studies have found that air pollution is associated with an increased risk of cardiovascular disease, likely related to the associated increased systemic inflammation. The MESA study (Multi-Ethnic Study of Atherosclerosis), for example, found a progressive increase in atherosclerosis over 10 years related to the intensity of pollution exposure http://gmodestmedblogs.blogspot.com/2016/06/air-pollution-and-heart-disease.html

 

    -- Stress:

        -- many articles over the years have found that chronic stress is associated with increased cardiovascular risk

            -- patients with stable CAD in a recent prospective study found that a stressful public speaking task led to significant ischemia, and after 5 years follow-up was associated with 2.5-fold risk of composite of cardiovascular death or MI, that mental stress was a stronger predictor of cardiovasc events than ischemia found on exercise testing (also found in earlier studies), and that the combo of mental stress-induced ischemia and conventional stress-testing induced-ischemia was much more predicative of a cardiovasc event:  http://gmodestmedblogs.blogspot.com/2021/11/mental-stress-and-heart-disease.html

            -- another study documented that endothelial dysfunction associated with mental stress predicted a 78% increase in major cardiovascular outcomes:  http://gmodestmedblogs.blogspot.com/2022/01/stress-induced-cardiovascular-disease.html 

            -- mental stress is associated with systemic inflammation, endothelial dysfunction, coronary artery vasospasm, and hypertension

             -- there are some modulators of perceived stress, such as social supports, which modify the attributable risk of stress (and social supports should be incorporated into the risk models…). Or exercise, which decreases chronic inflammation and might decrease the physiological effects of stress on the body

     -- Racial disparities:

            -- there have been analyses noting marked Black/White differences in cardiovascular risk. it is clear overall that racial differences reflect differing social environments and stressors (and are not genetically determined). in this light, a recent analysis from the National Health and Nutrition Examination Survey (NHANES) 1999-2018, with  9.4 years of follow-up, found that the 54% increase in Black vs White CVD mortality became insignificant when controlled for the social factors of unemployment, low family income, food insecurity lack of homeownership and unpartnered status (see Ann Intern Med 2023; 176:1200-1208, doi:10.7326/M23-0507)

 

    -- Depression:

        -- a large pooled analysis of >500K participants in 22 cohorts (https://pubmed.ncbi.nlm.nih.gov/33320224/ ) found a step-wise increase in coronary heart disease, stroke and cardiovascular disease as depression scores increase

        -- a large observational study (https://www.ahajournals.org/doi/epub/10.1161/JAHA.122.028332) found that self-reported depression had a greater than 2-fold odds of having the composite of MI, angina and stroke in young adult

        -- also, atypical antipsychotic meds (may also be used to help treat depression) may be associated with increased cardiovasc deaths, cardiomyopathy, and arrhythmias: https://www.frontiersin.org/articles/10.3389/fphar.2021.815151/full

        -- many studies have documented that depression is associated with chronic systemic inflammation

 

    -- Peripheral Arterial Disease:

        -- many studies have found that PAD is a strong predictor of future MI and stroke: https://www.ahajournals.org/doi/10.1161/CIR.0000000000001005 , with at least a doubling of the risk

      -- there is clearly an overlap between risk factors for ASCVD and PAD, especially low HDL, as well as high apoB, lipoprotein(a), and small LDL levels (rather than total LDL); elevated systolic blood pressure; and diabetes/smoking (these 2 latter are especially powerful risk factors for PAD)

 

    -- Risk factors associated with other cardiac diseases:

       -- aortic sclerosis: associated with 50% increased risk of cardiovascular death and MI: https://pubmed.ncbi.nlm.nih.gov/15628107/

     -- atrial fibrillation: associated with 2-fold increase in cardiovascular mortality and major cardiovascular events: https://www.bmj.com/content/354/bmj.i4482

 

    -- Hyperuricemia/gout:

        --  many, but not all studies, have found a strong relationship between serum uric acid levels and cardiovascular events, with a dose-response curve: http://gmodestmedblogs.blogspot.com/2020/01/uric-acid-threshold-for-increased-cad.html ; this blog also refers to studies finding that allopurinol decreases cardiovascular outcomes

  

    -- Alcohol:

            -- though there has long been a finding and rationale that moderate alcohol consumption is cardioprotective (ie, observational studies finding cardioprotection, felt to be attributed to alcohol raising HDL levels). more recent studies have found the opposite http://gmodestmedblogs.blogspot.com/2023/04/mendelian-randomization-alcohol-does.html

                -- nondrinkers in many observational studies have included former drinkers, many of whom stopped drinking because of significant health problems (perhaps some related  directly to alcohol use disorders). Controlling for prior drinking has decreased the observed increased risk for “nondrinkers”. Also, likely related to this selection bias, the constellation of medical risk factors in several studies have found that the baseline cardiovascular risk is significantly higher in “non-drinkers” than “moderate drinkers” (ie, moderate drinkers tend to exercise more and eat better than nondrinkers, who may well be sicker)

                -- mendelian randomization studies (an evaluation that simulates an actual randomized controlled trial: see the associated URL above for more information on this) have found that there is no safe lower alcohol limit for ASCVD risk

 

    -- Unhealthy or healthy diet:

        -- there are more and more data on the role of diet in cardiovascular disease:

            -- non-sugar sweeteners: increased risk of cardiovasc mortality and nonfatal events, stroke, all-cause mortality, and hypertension: http://gmodestmedblogs.blogspot.com/2023/06/health-effects-of-non-sugar-sweeteners.html .

            -- ultra-processed foods: increased cardiovasc events, likely associated with less consumption of healthy foods, but also have adverse effects on satiety, glycemic response, gut microbiome, lipid changes, obesity, inflammation, oxidative stress, insulin resistance, hypertension…. (see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8483964/ )

            -- red meats: these change the gut microbiome to produce TMAO (trimethylamine-N-oxide), which is a profound atherogen (see http://gmodestmedblogs.blogspot.com/2014/10/heart-failure-microbiome.html and http://gmodestmedblogs.blogspot.com/2013/04/red-meat-and-heart-dz.html

        -- but, there are profoundly cardiovascularly-helpful diets:

            -- the most studied one is the Mediterranean diet, which even when compared to a low fat diet, is associated with major decreases in adverse cardiovascular outcomes http://gmodestmedblogs.blogspot.com/2018/06/mediterranean-diet-with-olive-oil-or.html . There is also quite impressive evidence that a low glycemic index diet is also cardioprotective http://gmodestmedblogs.blogspot.com/2023/02/diet-and-cardiovasc-or-diabetes-outcomes.html , and this diet is best for improving the HDL/LDL ratio

        -- so, diet should be included in the ASCVD risk calculator, both as potentially a high level bad thing (ultra-processed foods, red meat, non-sugar sweeteners…) and as a potentially strong factor in lowering risk (Mediterranean diet, etc)

 

    -- Lack of exercise or lots of exercise:

        -- many, many studies have found that sedentary lifestyle increases cardiovascular risk:

            -- a large systematic review found 34% increase in cardiovascular disease in those with sedentary behavior, though a 29% decrease if performing long-term physical activity: https://www.frontiersin.org/articles/10.3389/fpubh.2022.1018460/full

            -- animal and adult studies (as well as a relevant evolutionary model) finding that, for example, adults watching more than 2 hours of TV/d had a 125% increase in cardiac events: http://gmodestmedblogs.blogspot.com/2019/01/exercise-new-evolutionary-model.html

        -- many, many studies on overall benefit of exercise:

            -- exercise is associated with decreased all-cause as well as cardiovascular (and cancer) mortality, even for weekend warriors: http://gmodestmedblogs.blogspot.com/2022/08/exercise-good-for-you-even-if-weekend.html

            -- the American Heart Association issued a scientific statement stressing cardiorespiratory fitness as part of the risk assessment for cardiovascular disease, as well as all-cause mortality (http://gmodestmedblogs.blogspot.com/2017/01/cardiovascular-fitness-new-vital-sign.html )

                -- one MET (metabolic equivalent) of higher cardiorespiratory fitness is associated with a 10-25% survival improvement as well as decreased cardiovascular mortality, hospitalization (including for heart failure). [and, by the way, lower levels of dementia, metabolic syndrome, prediabetes, diabetes, hypertension….]

        -- and, exercise is an effective way to decrease systemic inflammation

 

-- multiple risk factors:

    -- observational studies for hypertension and smoking (those are the ones I have seen) have found that those people with multiple medical risk factors have more than additive cardiovascular risk. These studies have both found that the patient’s attributable cardiovascular risk from hypertension or smoking alone is twice as high with an additional risk factor, and four times as high with 2 other risk factors  (ie, the risk conferred by one risk factor seems to have a geometric increase in the final cardiovascular risk if another risk factor is present and not simply an additive one)

 

-- QRISK3 risk calculator:

    -- there actually is a risk calculator developed in the UK that includes many of the above factors, though it is definitely cumbersome, and does rely on some binary assessments: http://gmodestmedblogs.blogspot.com/2017/08/a-new-atherosclerosis-risk-calculator.html

        -- in particular, they include the basic stuff (smoking, diabetes, blood pressure) but also CKD stage 3,4, or 5; atrial fibrillation, angina or heart attack in 1^st degree relative, blood pressure measured on treatment, migraines, rheumatoid arthritis, lupus, atypical psychotic meds, corticosteroids, erectile dysfunction, calculated cholesterol/HDL ratio, BMI: https://qrisk.org/

        -- and, not surprisingly, QRISK3 development used a database that did not include all of the above inputs (pretty few of the individuals had complete data on all of these criteria)

 

So, ….. 

   -- A really big concern with the cardiovascular risk models is their very short-term horizon, typically 5 to 10 years (the 2013 AHA guidelines do emphasize the utility of a 30-year predictive model, though the data to support this is pretty limited)

        -- it is clear from studies that ASCVD is a progressive disease, with lipid streaks in pretty much everyone aged 15 to 34: 20% of men and 8% of women have advanced atherosclerotic lesions by age 30-34

        -- this is the reason (i assume) that the American Association of Pediatrics recommends that all children have lipid screening between ages 9 to 11 and again between 17 to 21 see https://publications.aap.org/pediatrics/article/146/1_MeetingAbstract/636/4732/Pediatric-dyslipidemia-screening-by-pediatricians?autologincheck=redirected

        -- it is also clear that ASCVD is still the most common cause of death in older people, and about a quarter of older people who develop clinical cardiac disease do not have evident cardiovascular disease before the age of 70

        -- studies confirm that statins as well as vigorous dietary changes do lead to significant changes in the composition of atherosclerotic plaques within two months (probably by decreasing the level of inflammation in those plaques, which is largely responsible for the rupture of these plaques; and, it is these newer, smaller, lipid-rich plaques that have more inflammation and a more fragile, thinner fibrous cap that ruptures more easily because of both shear stress by the passing blood as well as inflammation-related elaboration of metalloproteinases that weakens the fibrous cap, leading to thromboses and acute coronary syndromes ( >50% of the culprit lesions in patients with acute coronary syndromes have <50% arterial stenosis). this is why the primary prevention of acute coronary syndromes is aggressive medical management and not stenting)

            -- several studies have confirmed the utility of assessing coronary artery calcium scores as a surrogate marker of atherosclerotic disease. But calcifications reflect older and more established atherosclerotic lesions, and the CAC score is not so useful in younger patients who have earlier lesions (see cad calcium score less useful in younger pts JAMACardiol2022 in dropbox, or doi:10.1001/jamacardio.2021.4406)

            -- so, the older folks with scores >0 likely have more diffuse coronary artery disease with more advanced lesions (and calcium), though it still seems more likely that the younger, fresher, less calcified lesions are the ones more likely to lead to acute coronary syndromes

            -- which means that CAC scoring is in fact beneficial in cardiovasc risk assessment, but especially so in those >60yo

        -- clinical studies with statins have found that there are decreased cardiovascular events within about six months or so after starting the medication (i.e., supporting their use in those with a life expectancy greater than six months)

        -- studies of statins for primary prevention as well as for people who have clear clinical cardiovascular disease (secondary prevention) both find that there is a 30% decreased subsequent cardiovascular event. But the absolute risk of an event is significantly higher in those who have had a prior cardiovascular event. so, primary prevention is really important 

 

    -- so, if one considers atherosclerosis to be a long-term progressive disease, that the first cardiac event not only has a significant mortality associated with it (on the order of 30 to 40%, much of it occurring prior to reaching the hospital: Law MR. Arch Intern Med. 2002; 162: 2405-2410) and also leads to increased risk of another cardiac event despite statins, then:

        -- it makes sense to check lipid levels of all adult patients (and teenagers if you happen to see them). And when the LDL level is high:

            -- for young and basically healthy people,  I believe it is very important to really stress the nonpharmacologic interventions of maintaining an appropriate weight, eating healthy foods (e.g. Mediterranean diet, low glycemic index diet, etc.), and doing regular exercise.

            -- for premenopausal women who do not have diabetes, their ASCVD risk in large cardiovascular registries is quite small, so the overwhelming focus (I think) is helping with nonpharmacologic interventions, especially motivational interviewing

                -- statins are contraindicated for women who might become pregnant (not so good to decrease lipids in babes in utero since low cholesterol could affect their brain development)

                -- menopause is associated with a dramatic increase in atherosclerosis, and ASCVD reduction should be pursued aggressively (nonpharmacologically and meds)

        -- it also makes sense to continue aggressive lipid treatments into old age, assuming a 6-12 month life expectancy (and, anecdotally, i do continue to treat people into their 90s, many with “high intensity” statins, without adverse issues and (i assumed, since few die from cardiovascular events) continued benefit

        -- And, a 10-year cardiovascular risk factor analysis in a person who is 20-50 years old is likely to not be very useful in preventing long-term cardiovascular events, unless the calculated risk is very high (ie, a low calculated 10-year risk horizon may mean nothing for a 20 to 50 year old person)

 

Gestalt:

-- so, how does one get through this morass of information. I really am not a nihilist who only critiques the information in the medical literature (though I do think that it is important to highlight the rather frequent inadequacies). And I probably am much more aggressive in preventive cardiology than most. And I mostly use the method of gestalt:

    -- since we in primary care are treating the actual patients in front of us, I do think that these limitations of the risk factor models, based on large population data, are often not very useful when applying them to an individual patient

    -- we do know that several of the risk factors that do not make it into the risk calculators are really important in developing cardiovascular disease. Some of these, as above, are medical risk factors (PAD, HIV, depression, etc), and some of these are not (stress levels, lack of exercise, less healthy diet, lack of support systems, environmental exposures such as air pollution)

    -- so, pretty much all patients in the US have multiple risk factors, if one were to include all of the above in predicting the risk of ASCVD; and we know from several studies that combinations of risk factors can be more than additive in the cardiovascular risk

    -- my approach has been to try to do a qualitative assessment of the aggregate of the individual patient’s risk factors. eg,  if a patient has diabetes along with depression, or a stressful life, or PAD, or HIV, I will be more aggressive in discussing nonpharmacologic as well as pharmacologic interventions, perhaps with a very aggressive lipid goal of <40 in those who seem to be at very high risk, which may well require more than one med and more intensive motivational interviewing to help improve diet/exercise/etc

        -- I should mention that I also disagree with the AHA classifications of statin intensity, since I do have some patients on the "moderate-intensity" atorvastatin 20 mg who have an LDL of under 40, and many patients on the "high-intensity" atorvastatin 40 having a much less significant LDL response than with rosuvastatin 40mg , another "high-intensity" med

 

i do realize that this "gestalt" approach is not easily quantifiable, which makes it hard to assess in RCTs (a limitation of “evidence-based medicine”)

And, I’m sure that the above list of cardiovascular risk factors is undoubtedly not complete...

 

all of this suggests strongly that we should be screening earlier (like, everyone) and treating people pretty aggressively (nonpharmacologically and with drugs as needed). and that this should continue perhaps until patients no longer have life expectancy of at least 6-12 months

 

geoff

 

 

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

  

to get access to all of the blogs:

 

 go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

  -- click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category

  -- or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them

 

if you would like to see the articles in this blog, please email me. 

 

please feel free to circulate this to others. also, if you send me their emails (gmodest@bidmc.harvard.edu), i can add them to the list