This is the second entry in a series aimed at changing the way most of us think about cardiovascular disease prevention. In my last entry, I presented four key messages about CHD prevention (http://beyondapples.org/2009/06/23/the-end-of-cardiovascular-disease-part-1/). In this entry, I will expand upon the first message and will discuss some of the primary data behind it as well as its implications.
The first key message is:
“ Heart attacks rarely occur in the absence of the major risk factors (adverse levels of blood cholesterol, blood pressure, smoking, diabetes). In at least one major health study, 90 to 100 percent of all CHD events occurred in people with at least 1 major CHD risk factor.”
Let’s begin by reviewing one of the seminal papers that first supported this idea, a study by Greenland et. al published in JAMA 2003.* In this study, the authors reviewed data from three large cohorts — the Chicago Heart Association Detection Project in Industry (Chicago study), Multiple Risk Factor Intervention Trial (MRFIT), and Framingham Heart Study (FHS). To those unfamiliar, in a cohort study a group of individuals are followed for years to study an outcome of interest; instead of randomizing enrollees to receive a drug or another type of intervention, they are simply followed and are periodically given questionnaires or blood tests in order to discern associations between the outcome of interest and these other factors. In the Chicago study, the investigators enrolled 35,000 men and women between 1967 and 1973 from 84 Chicago-area companies. Baseline data on smoking status, cholesterol levels, blood pressure, diabetes, and other demographic and cardiovascular data were collected on enrollment. Since then the cause of death in these individuals have been tracked using death certificates for over 30 years. The MRFIT study was a similar study of 360,000 men from 18 US cities enrolled between 1973 and 1975. They also had data collected only on enrollment and have also been followed for cause of death. The FHS study was a much smaller study of 5000 men and women in a single city (Framingham, Massachusetts) but had much richer data, which enabled them to not only track cause of death but also other outcome measures such as non-fatal heart attacks and onset of symptomatic coronary heart disease.
What Greenland and his colleagues did with these complex data was very simple and elegant. They looked at the number of people in these three studies that died from CHD or, in the case of the FHS study, had a CHD event, and then asked what proportion of those people had one or more major cardiovascular risk factors at enrollment. That is, they asked how many people who suffered from a major CHD event were “exposed” to a major cardiovascular risk factor earlier in life. The four risk factors they looked at were smoking status, cholesterol, blood pressure, and diabetes. They ran the analysis twice — once with standard clinical cutoffs (for example, systolic BP > 140mmHg or diastolic BP > 90 mmHg or hypertension) and once with more stringent cutoffs (for example, systolic BP > 120mmHg or diastolic > 80mmHg which includes hypertension and prehypertension). What they found was that 87 to 100 percent of people who died from CHD were exposed to one or more of these four risk factors. Using just the FHS data, over 90 percent of people who had a CHD event were exposed to one or more major risk factors. Thus the authors concluded that CHD is uncommon in the absence of one or more major risk factors for CHD; it is a myth that half of people who suffer from heart attacks have no major risk factors for the disease.
The Greenland paper is one of now many studies that have reached the same conclusion. Some of these studies were done using a similar methodology on different data sets. Others looked at clinical trial data and re-analyzed the data to figure out what proportion of people had suffered from CHD in these trials had a self-reported exposure to one or more major risk factors. While these studies have varying estimates they generally conclude that at least 80 percent of people with CHD have known risk factors for the disease.
These studies have important and relevant limitations. For one, there is a concern about “cross contamination.” That is, it is likely that some people who had risk factors at enrollment may have subsequently gotten their risk factors under control (e.g. someone who quit smoking) and that some people who had no risk factors at enrollment subsequently developed one or more risk factors (e.g. someone with normal blood pressure that became hypertensive). If significant, the latter form of contamination would lead us to under-estimate the prevalence of risk factors in people with CHD. The former would lead us to over-estimate the prevalence of risk factors. Another concern is that they do not distinguish uncontrolled from controlled risk factors.
These potential cofounders get to the fundamental significance of these studies. The studies looked at “exposure” to risk factors; if someone ever had sub-optimal risk factors for CHD they would be considered “exposed.” They do not consider people who start taking blood pressure medications and get their blood pressure under control or people who smoke 2 packs a day and then quit smoking differently from people whose risk factors remain uncontrolled. This is all well and good for the fraction of us who have optimal risk factors for CHD from the get-go but what does this mean for those of us who are already exposed to one or more risk factors. If we get our blood pressure under control or quit smoking, are we doomed to be at increased risk for CHD? The answer from this body of work is that we don’t know. However, from other medical studies (and from common sense) we know better. Studies show for example that smokers who quit reduce their risk of CHD by one-half within 1 year of quitting and within 15 years to that of people who have never smoked. We also know that risk of CHD increases with the number of risk factors present (the more risk factors the greater the risk) and with the degree of abnormality (e.g. the higher the blood pressure the greater the risk). Risk factors are not binary; they are continuous. Furthermore, risk of CHD increases with the number of risk factors but not linearly; each additional risk factor has a multiplicative effect on CHD risk. Thus even if you are in the “one or more adverse risk factors” category you can lower your risk of CHD by improving your risk factors.
Despite these limitations, the concepts presented in these studies are incredibly powerful and the message incredibly clear. CHD is preventable. Under optimal conditions – in which people are never exposed to the four major risk factors (smoking, high blood pressure, high cholesterol, diabetes) — CHD is rare. If risk factors are present we need to control them in order to minimize exposure to conditions that promote athersclerotic cardiovascular disease. Heart attacks are not random strikes of lightning; they may be terrifying but they are not arbitrary. Like thunderstorms to lightning, heart attacks brew under conditions in which adverse levels of blood cholesterol, blood pressure, smoking and diabetes are present. By controlling these risk factors, through lifestyle modifications and medications, we can stave off the storm and prevent atherosclerotic cardiovascular disease.
- Shantanu Nundy, M.D.
* Greenland P, Knoll, MD, Stamler J, et al. Major risk factors as antecedents of fatal and nonfatal coronary heart disease events JAMA 2003;290:891-897.