Doctors should not blindly extrapolate data from research studies

It is common practice in health care today for physicians to read articles published in leading journals within their fields and extract what they think is important for their patients. Though in most cases it is to the patient’s benefit that their physician stays up to date with the latest scientific advances, it is nonetheless important that doctors pay particular attention to research methodology and how generalizable these studies are to their patients, and not simply extrapolate without proper due diligence. Physicians must remain cautious in terms of what they deduce from such studies, for history has shown that even the most popular and widely accepted research methodologies on which these studies are based can sometimes be flawed. Let us take the classical twin method as an example.

Twins have captured public attention for centuries, and the scientific study of the twin phenomenon is richly documented in the literature. Sir Francis Galton is often credited with initiating the field of twin study as a means of investigating questions regarding the influence of “genetics” vs. “environment” on the human condition.  However, he did not account for zygosity in his studies, and it was not until Hermann Werner Siemens published “Zwillingspathologie” in 1924 that the true value of studying twins to answer basic scientific questions was first realized.  Siemens thereby launched the “classical twin method” by utilizing the study of differences between monozygotic twins (MZ) who share 100 percent of the same genes and dizygotic twins (DZ) who share an average of 50 percent of the same genetic code. This realization has allowed twin studies to bring understanding to many areas of life, from common traits to variation in IQ.  Many questions, however, still remain unanswered. Consequently, it has been presumed that such answers remain illusive due to the complexity of environmental influences.

If MZ twins share 100 percent identical DNA, then members of that pair should theoretically be indistinguishable from one another. However, this is clearly not the case.  While on casual inspection close friends or even parents commonly confuse twins, careful inspection often allows for visual distinction. For example, anatomic evaluation suggests MZ twins rarely demonstrate 100 percent anatomic concordance, pointing to the role environment plays in anatomic expression. Studies have estimated a 50 percent genetic and 50 percent environmental influence on the human condition, and efforts have been made to distinguish their roles.

Nonetheless, a critical issue can be raised regarding what constitutes as “environment.” For example, identical twins not only share the same genetics, but also share the same chorion, which is the outermost membrane surrounding the fetus during prenatal development, and this may make them more similar than non-identical twins who never share the same chorion. Thus, is a 500-fold increase in risk for autism in identical twins due to the same genetics, or due to another form of environment, namely, the “prenatal environment”? This problem of definition thereby presents a major challenge to twin research. Even if the “equal environments assumption” were true, any study that demonstrates more concordance with identical twins cannot claim genetics as the cause since these identical twins share the same “prenatal environment” as well.

Although the simplicity of the twin method has contributed to its popularity through the years, more recently, other doubts have been raised regarding its validity and the analysis of results derived from it. Important questions have been raised, such as whether twins can be considered as representative of the general population. For example, some researchers claim that adverse intra-uterine environments caused by sharing a womb suggests twins cannot be representative of the non-twin population. More importantly, the fundamental “equal environments assumption” has come under question, and there exists the possibility that identical twins share more similar experiences and environments after they are born than due non-identical twins.

As such, it can be argued that the entire twin method, which serves as the basis for many heredity vs. environment studies, is flawed, and monozygotic twins cannot be viewed as truly identical. Further supporting this claim, identical twins have been found to be discordant for hand preference, and important differences have been found to exist in dentition among identical twins. Until future research provides a more accurate description of what constitutes “environment” as well as accounts for the issue of prenatal factors and chorionicity, the classical twin method remains imperfect.

As William Osler suggests, “medicine is a science of uncertainty and an art of probability.” History has demonstrated time and time again that even the most widely accepted ideas and research methodologies can have flaws (the classical twin method being just one example). It is likely, if not probable, that in 100 years time, much of what we hold true from research studies today will be proven wrong. New research methodologies with improved validity will almost certainly be developed. It is, therefore, imperative that doctors not blindly extrapolate data from research studies published in scientific journals. It is the duty of the physician to think critically, paying close attention to the methods section and ensuring the patient cohort included in the study is also representative of the physician’s own patient population.

Robert Dorfman is a medical student.

Image credit:

Leave a Comment

Most Popular

✓ Join 150,000+ subscribers
✓ Get KevinMD's most popular stories