Cardiac risk factors in children predict adult outcomes

EVIDENCE is building that cardiovascular disease (CVD) risk factors seen in children are a predictor of CVD events in middle age. The question for GPs and other physicians is, what should be the next step?

A global team of investigators from the International Childhood Cardiovascular Cohort (i3C) Consortium reported this month in the New England Journal of Medicine that CVD risk factors, when measured in childhood, strongly predict occurrence of CVD events in middle age. The findings were obtained from a 40-year follow-up study of 38 589 children from the US, Finland and Australia.

The risk factors included those that are conventionally used in CVD risk screening for adults – blood pressure, serum lipids, body mass index, and smoking – factors shown repeatedly in prospective cohort studies in adulthood to be important predictors of later CVD.

However, similar evidence has not been available for these same risk factors when they are seen in children. The reason for this is that to obtain that evidence requires a study of sufficiently large sample size and very long duration, spanning the period from childhood to the age at which CVD events start to become common – middle age. The practical difficulties in mounting such a study seemed insurmountable until 2002, when the Australian investigators, who had embarked on a follow-up of Australian school children surveyed in 1985, suggested to their counterparts in the US and Finland that they might pool data from their cohorts and follow them for occurrence of CVD events by mid-life. All cohorts had commenced enrolling participants in the 1970s and 1980s and had measured CVD risk factors at least once in childhood at ages 3–19 years.

The five risk factors in children found to be associated with risk of an adult CVD event — the majority of which were fatal or non-fatal myocardial infarctions or strokes — were systolic blood pressure, serum cholesterol and triglycerides, body mass index, and smoking.

All risk factors in children were significantly associated with increased risk for adult CVD events, both fatal and non-fatal, with the hazard ratios for those whose values were above a clinically defined cut-point – eg. high BMI or high cholesterol – ranging from 3.39 to 2.13.  The risk among those who had more than one risk factor, identified through a risk “score”, was considerably higher than for a single risk factor. Children whose combined risk scores were in the top 5% of the distribution were estimated to have a risk for a CVD event more than six times that of children at the lower end of the risk factor distribution.

These estimated risks are reasonably similar to those seen in adults and they are clinically meaningful.

The question that must arise in an observational study of this kind is whether we should assume that the associations observed are causal. The fact that early atherosclerosis has been found to be frequently present in children under the age of 10 years makes the findings highly plausible. The design of the study also makes it unlikely that commonly encountered biases could explain the findings. While the individual risk factors have not been adjusted for each other, the set of risk factors has been adjusted for location of the cohort, for socio-economic status, and race. Loss to follow-up, which is a major potential source of bias in prospective cohort studies is not a concern here because the elevated risks were found in the analysis that involved well ascertained fatal events identified through death registries.

Finally, in all observational studies we are concerned that reverse causation, whereby the disease causes changes in risk factors, might explain the results. In this study, the risk factors were measured long before any disease-related symptoms were likely to have been experienced.

We believe these findings must be taken very seriously, particularly in the context of the childhood obesity epidemic, which has largely occurred since these cohorts were initiated. The New England Journal of Medicine editor wrote in an accompanying editorial to the main article:

“If [cardiovascular] risk factors can be identified early in life, we as clinicians have opportunities to address health issues early and might uncouple risk from an inexorable march toward CVD and death. We have been waiting for hard data showing that risk factors seen in childhood forecast future disease, and now we have a good start concerning CVD.”

The question, though, is what should be the next step?

Recommendations concerning policy and implementation strategies for CVD prevention commencing in childhood have been developed over recent years but in a setting where there was no certainty about how important reducing those risk factors might be. A concerted effort to review effectiveness of various strategies to address the problem of CVD risk factors in children is now imperative. This effort must involve policymakers within government and health care to ensure that optimal approaches are identified and implemented.

Professor Terence Dwyer is Emeritus Professor of Epidemiology at the University of Oxford. He is a Professorial Fellow at the Murdoch Children’s Research Institute in Melbourne, and Honorary Professorial Fellow in the Department of Paediatrics at the University of Melbourne. He was the senior author of the New England Journal of Medicine paper cited in this article.

Professor Alison Venn is the Director of the Menzies Institute for Medical Research at the University of Tasmania.

The statements or opinions expressed in this article reflect the views of the authors and do not necessarily represent the official policy of the AMA, the MJA or InSight+ unless so stated.

Subscribe to the free InSight+ weekly newsletter here. It is available to all readers, not just registered medical practitioners.

If you would like to submit an article for consideration, send a Word version to mjainsight-editor@ampco.com.au.