A rule of thumb from the 1800s to predict the height of children based on their parents鈥� stature is still far more accurate than the best genomic predictions.
In 1886, , a Victorian scientist who also pioneered the field of eugenics, published a technique to predict the height of children. It averages the height of both parents and makes adjustments for age and sex.
鈥淚t鈥檚 really not rocket science,鈥� says , a geneticist at Erasmus University Medical Center in Rotterdam in the Netherlands.
Advertisement
Aulchenko鈥檚 team compared this approach to a more complicated calculation based on gene variations linked to height. Researchers previously identified these mutations by scanning the genomes of tens of thousands of people, and then hunting for single letter changes shared by people of similar stature.
Minor correlation
Aulchenko鈥檚 team analysed 54 of these genetic variants across 5748 Dutch people and calculated a simple score for each person. The higher a person鈥檚 score, the more gene variants linked to tall stature he or she possessed.
But when these numbers were plotted against each person鈥檚 height, and adjusted for age and sex, Aulchenko鈥檚 team found only a minor correlation between a person鈥檚 genetic score and their actual height. Galton鈥檚 method, on the other hand, proved about 10 times better at guessing the height of another 550 people.
, a geneticist at the Broad Institute of Massachusetts Institute of Technology and Harvard University in Cambridge, Massachusetts, isn鈥檛 surprised by the poor prognostic ability of genes linked to height 鈥� including a handful identified by his lab. The team noted just as much when it published its findings in 2008.
This is because most genetic variations have a miniscule affect on height and many more have yet to be identified, Aulchenko says. A few traits, such as eye colour, are easier to predict with genes, and new research will certainly improve genetic prognostication. But, for now, most of our features remain a genetic black box.
Journal reference:
