One method used to correct children's short sight may not help them - and could actually do harm - research has suggested.

The practice of "undercorrecting" short-sightedness, or myopia, is not recommended by optometry bodies as a way to deal with the problem, and leading members have hotly disputed it is carried out at all.

But experts have told the BBC it has been used for decades when prescribing glasses or contact lenses.

New Scientist magazine reports a study by a British optometrist who intended to confirm the practice works.

But rather than prove the technique had merit, he was forced to cut short his work when it became apparent that children's eyes were actually getting worse.

The findings have raised concerns that millions of people worldwide may have worse eyesight and may even be more likely to go blind because they were given a treatment that does not work.

Optometrist Professor Daniel O'Leary, of Anglia Polytechnic University in Cambridge, studied 94 children in Malaysia.

The study was meant to run for three years, but was cut short after two years when it became clear that the children's eyesight was getting worse.

Strong evidence

Paul Adler, of the College of Optometrists, said: "It is the strongest evidence I have seen in this field.

"It could change prescribing practice worldwide."

Short-sightedness is caused by the inability of the muscles in the eye to flatten the lens enough to focus light from distant objects directly on to the retina.

Instead, the point of focus is in front of the retina, leading to a blurred image.

Glasses correct the problem by moving the focal point back onto the retina.

It has never been accepted practice to undercorrect children or adults in this country

College of Optometrists

But when people wearing normal glasses look at close objects, the focus point is usually behind the retina.

The theory is that to try to "refind" this focal point for near objects, their eyeballs actually elongate.

Not only does this make distance vision even worse, it also increases the risk of serious eye diseases such as retinal detachment, glaucoma and retinopathy, all of which can lead to blindness.

According to this theory, undercorrection should help stop the eyeball elongating.

When they undercorrect, optometrists prescribe a lens that focuses light from distant objects just in front of the retina, rather than exactly on it.

Theory

No glasses is the worst option of all

Professor Daniel O'Leary

However, the theory is based on the scantiest of evidence - a study of just 33 Japanese children in 1965, and animal studies involving chicks.

Professor O'Leary's team undercorrected the sight of half the children, and fully corrected the rest.

Then they measured the length of the eyeball with ultrasound every six months.

They found that the eyeball elongates faster when vision is undercorrected.

As a result, the vision of the children with undercorrected short-sightedness deteriorated more rapidly.

Professor O'Leary believes the eye cannot tell whether the focal point is in front of the retina or behind it. Instead, it just grows backwards if the image is out of focus.

"No glasses is the worst option of all," he said.

"But don't undercorrect. Go for full correction."

Frank Munro, president of the College of Optometrists, told BBC News Online that the practice of undercorrection was no longer used.

However, he said it had been employed by optometrists in the past.

In an official statement, the College of Optometrists said: "The suggestion that optometrists have been routinely undercorrecting short-sightedness when prescribing glasses or contact lenses in the UK is incorrect.

"It has never been accepted practice to undercorrect children or adults in this country."