Author + information
- Themistocles L. Assimes, MD, PhD⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Themistocles L. Assimes, Stanford Population Health Sciences Building, 1070 Arastradero Road, Suite 300, Palo Alto, California 94304
The familial aggregation of coronary artery disease (CAD) was first noted in several relatively small epidemiological studies performed in the 1950s and early 1960s (1). Some of these early reports actually predated the identification of the major “traditional” risk factors of CAD by Framingham investigators. Thus, it was initially not clear what proportion of the familial aggregation of CAD observed was simply a consequence of the familial aggregation of traditional risk factors.
A number of larger case-control and cohort studies followed in the 1970s and 1980s that unequivocally established family history as an independent risk factor for CAD (1). Through multivariate analyses, these studies also suggested that only a fraction of the familial aggregation observed could be accounted for by the familial aggregation of traditional risk factors of CAD. Curiously, family history was not included in the original Framingham risk profile published in 1976 and subsequently adopted by the influential Adult Treatment Panels (2,3). The main reasons cited for omitting this variable were its limited incremental predictive value over other major risk factors and the challenge in measuring it reliably. Consequently, as the Framingham risk score became widely implemented in the 1990s, research focused on further defining the clinical utility of documenting a family history in the primary prevention of CAD probably received less attention than it deserved. Instead, scientists focused on estimating what proportion of the unidentified factors responsible for familial aggregation was genetic in nature (vs. cultural or environmental). Landmark adoption and twin studies concluded that this proportion was substantial, especially among families with early manifestation of disease (4,5).
Over the past 10 years, several other groups have developed cardiovascular disease risk scores for clinical use (6). In Europe, the SCORE (Systematic Coronary Risk Evaluation) project has produced multiple country-specific risk scores for fatal CAD, whereas the ASSIGN (Assessing Cardiovascular Risk to Scottish Intercollegiate Guidelines Network/SIGN to Assign Preventative Treatment) and QRISK (QRESEARCH Cardiovascular Risk algorithm) projects have produced risk scores for cardiovascular disease (CVD) among Scottish and English populations, respectively. In the United States, the Reynolds risk score was recently developed in women and was subsequently also demonstrated to be effective in a cohort of men. In contrast to the Framingham cohort, family history was found to have adequate incremental predictive value over other major traditional risk factors in all 3 of the new scores where this information was available.
In this context, the report by Chow et al. (7) in this issue of the Journal is a timely one. This important INTERHEART substudy documents, for the first time, the ability of a measure of family history of CVD, in this case, parental history of myocardial infarction (MI), to predict the risk of an important adverse cardiovascular outcome, in this case nonfatal MI, across all major geographic regions and race/ethnic groups of the world. Consistent with previous studies involving more homogeneous populations, the excess risk conferred by a parental history of MI was independent of all other traditional risk factors, and its magnitude was related to the age at onset of disease and the number of family members affected (1,5,8).
One finding of this study that may initially appear inconsistent with previous studies was the limited attenuation observed of the overall age-, sex-, and region-adjusted odds ratio (OR) for parental history of MI after further adjustment for all INTERHEART risk factors (1,5,8). The degree of attenuation is important because it indirectly reflects the contribution of familial aggregation of traditional risk factors to the association between parental history of MI and risk of MI. In this study, the point estimate of the OR after this full adjustment decreased from 1.81 to 1.74, suggesting that this contribution is small. However, 3 consequences of the case-control design of this study may have contributed to this seemingly muted degree of attenuation. First, no incident fatal cases could be enrolled. Previous prospective studies incorporating validated CVD events in first-degree relatives have demonstrated a generally higher degree of attenuation of risk ratios among subjects with a family history of premature CVD and among subjects with a more severe outcome including early age of onset and death from CVD compared with other cases (5,8). In this context, the larger degree of attenuation observed in INTERHEART among cases with early onset MI, parental history of premature MI, parental history of MI in both parents, or a combination thereof is reassuring. Second, adjusting for hypertension status instead of a more accurate pre-event blood pressure may have also led to a lesser degree of attenuation of the OR. Last, the study produced somewhat unexpected relationships between parental history of MI and several behavioral risk factors including smoking, physical activity, and diet. These unexpected relationships may have negatively affected the attenuation forces on the OR by the biological risk factors.
Among a subset of cases and controls, the investigators assessed the influence of a small number of genetic variants previously found to be associated with CAD. These variants collectively would be expected to attenuate the OR, yet no attenuation was observed. The authors appropriately emphasize the power limitations of their genetic analyses given the small number of variants used and the fact that most of them influenced risk factors already included in the multivariate analysis. These genetic analyses are worth repeating after expanding the genotyping effort in this sample set to all variants that have been robustly associated with CAD and its major risk factors through more recent genome-wide association studies (9).
It will likely take several years and possibly decades before cardiovascular epidemiologists are able to identify the numerous genetic and nongenetic factors responsible for familial aggregation of heart disease. In the meantime, the study by Chow et al. (7) combined with recent efforts to develop new risk scores around the world have undoubtedly reignited interest in understanding the role of this traditional risk factor in the primary prevention of CAD. In developed countries, the predictive ability of a family history probably improved over the past few decades in parallel with an enhanced ability to diagnose heart disease and the public's increasing awareness of its causes and consequences. In the future, the predictive ability of a family history may further improve as linked electronic medical records facilitate the reliable identification of family members with cardiovascular disease as well as traditional risk factors. Finally, in developing countries around the world undergoing the epidemiological transition to chronic diseases, a crucial need to identify and treat people at risk of CAD using cost-effective nonlaboratory-based risk scores has emerged (10). Clearly, measures of familial risk as implemented in this study have the potential to play an important role in meeting this need and consequently substantially reducing the global burden of cardiovascular disease.
Dr. Assimes has reported that he has no relationships to disclose.
↵⁎ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
- American College of Cardiology Foundation
- Goldbourt U.,
- De Faire U.,
- Berg K.
- National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III)
- Berger J.S.,
- Jordan C.O.,
- Lloyd-Jones D.,
- Blumenthal R.S.
- Chow C.K.,
- Islam S.,
- Leonelo Bautista L.,
- et al.
- Hindorff L.A.,
- Junkins H.A.,
- Mehta J.P.,
- Manolio T.A.