The content of this website is intended for healthcare professionals only

Height may be risk factor for varicose veins

Biobank analysis confirms expected risk factors and machine learning techniques highlight new ones

Ingrid Torjesen

Tuesday, 25 September 2018

The taller you are, the more likely you are to develop varicose veins, according to a study* published in Circulation.

The research led by Stanford University School of Medicine used data from the UK Biobank to look for varicose vein risk factors using machine learning combined with epidemiological methods in 413,519 participants. They also screened for genetic markers using genomewide association studies in 337,536 of the participants, 9,577 of whom had varicose vein disease.

The study confirmed that established risk factors, including being older, female, overweight or pregnant, or having a history of deep vein thrombosis, were all associated with varicose veins.

It also showed that surgery on the legs, family history, lack of movement, smoking and hormone therapy are risk factors associated with the condition. However, the researchers were surprised when machine learning also showed an unexpected correlation between the condition and height.

Typically, researchers use genomewide association studies to examine DNA variation that may be associated with an increased risk for a particular illness. Using this method, the researchers identified the 30 regions on the genome associated with varicose veins. But the researchers also used another method involving machine learning, a type of artificial intelligence, to cast a giant net to discover any previously unknown risk factors.

"These methods represent new ways of thinking about research," said Erik Ingelsson, professor of cardiovascular medicine at Stanford University School of Medicine said.

"You go in without a hypothesis about a specific biological mechanism and scan for something new. You could say that you turn the machine loose on it. In this case, we included 2,716 predictors of varicose veins in this machine-learning algorithm. Then we let the algorithms find the strongest predictors of varicose veins."

When height emerged from the machine-learning analysis as a possible risk factor, the researchers conducted further tests to see if it was an actual cause for the disease using mendelian randomisation analyses, a statistical technique to determine causal effects.

"Our results strongly suggest height is a cause, not just a correlated factor, but an underlying mechanism leading to varicose veins," Ingelsson said.

In addition to height, the machine-learning algorithm showed that bioimpedance, a measure of how well the body impedes electric current flow, is a strong predictive marker for varicose veins. This measurement could potentially be used as a diagnostic tool to predict for varicose veins, Leeper said.

*Fukaya E, Flores AM, Lindholm D, et al. Clinical and Genetic Determinants of Varicose Veins Prospective, Community-Based Study of ≈500000 Individuals. Circulation. 2018; DOI:10.1161/CIRCULATIONAHA.118.035584

Registered in England and Wales. Reg No. 2530185. c/o Wilmington plc, 5th Floor, 10 Whitechapel High Street, London E1 8QS. Reg No. 30158470