John Tsagaris, a traditional acupuncturist, is the darling of the glamour glossies and the well-heeled. If you phone for an appointment at his EnergyBodies clinic in Chelsea (£140 for a one-hour treatment) you’ll go on a waiting list and see him in a month if you’re lucky. Tsagaris admits that it’s hard for him to keep up with the demand.

The reason, he says, is that acupuncture works: he treats skin disorders, signs of skin ageing, stress, infertility, menopausal symptoms, back problems, facial paralysis . . . and had “amazing” results, he claims.

Six months ago, the scientific rationalists among us would have had grounds for pooh-poohing all this as unproved, pricey pampering. But that’s all changed. Science has suddenly — and very publicly — taken a turn in favour of the needle.

Last month, new official guidance said that patients with persistent lower back pain should be offered acupuncture on the NHS. The recommendation, from the NHS rationing body NICE, said that the scientific evidence now showed that acupuncture was not only effective at treating back pain, but also cost effective when other treatments, such as painkillers, were found not to be working.

This came hot on the heels of a report from the influential medical research review body the Cochrane Collaboration, saying that headache and migraine sufferers clearly fared better with acupuncture than painkillers.

Even more astoundingly, a major study from the same respected body last year indicated that acupuncture could make women more fertile. A large scientific review found that women undergoing fertility treatment were more likely to give birth successfully if they had acupuncture.

So is Tsagaris right? Acupuncture works? The short answer is, yes — for head and back pain. But for everything else, it depends on whom you talk to. Organisations such as the British Acupuncture Council, which represent traditional acupuncturists, claim there is evidence that acupuncture successfully treats a huge range of conditions including hay fever, seasonal depression, arthritis, irritable bowel syndrome — even diabetes.

At the other end of the scale is Professor Edzard Ernst, chair of complementary therapy at Exeter and Plymouth universities. He, like many medical researchers, is aware of the dangers of being swayed by the evidence of one or two small studies — you need to study the effects on thousands of people before you can say “this works” or recommend that people should pay for a treatment.

He doesn’t buy the supposed evidence on fertility, he says, because subsequent good reviews of the evidence have indicated that acupuncture has no effect. In most other fields, there has been too little good research, he says.But even Professor Ernst, a known sceptic of complementary medicines, acknowledges pretty strong evidence that acupuncture kills pain and helps nausea sufferers. But that doesn’t mean that it should be provided on the NHS.

“I think NICE has over-emphasised the benefits of acupuncture,” he says. “The effect is pretty small and it remains to be seen whether this will have a significant impact on how effectively we treat people.”

One of his worries is whether acupuncture actually has any effect on the body, or whether it’s simply a glorified “placebo” — a treatment that makes us feel better simply because we expect to be made better. If it is just a trick of the light, should the NHS — or anyone — be paying for it?

Between the scepticism and the evangelism, however, lies the middle ground occupied by the medical acupuncturists — doctors who want to ground this Eastern approach in Western science. Some of the recent research they’ve been involved in provides insight into how acupuncture works, and why it’s so difficult to prove that it does.

Traditional acupuncture revolves around a theory of energy flow (qi) around the body. Acupuncture needles inserted into points along hypothetical channels called meridians can encourage the energy flow, thus promoting health. Medical acupuncturists use many of the acupuncture points, but they don’t hold to the theory.

They believe instead that inserting needles affects the central nervous system and modulates pain pathways — which is why it works so well on headaches and backaches. According to Dr Mike Cummings, medical director of the British Medical Acupuncture Society, this has a knock-on effect on a range of other conditions — arthritis or hay fever for example — because pain can encourage tissue inflammation. Recent MRI brain-scanning research has backed them up, showing that the insertion of acupuncture needles activates an area of the brain associated with the production of natural painkilling opiates.

The interesting thing about acupuncture studies, however, is that administering “sham” acupuncture — where needles are inserted into the wrong place, or aren’t pushed in properly, or toothpicks are applied to the skin instead — often seem to have the same effect. To the sceptics, this suggests that acupuncture doesn’t actually have a curative effect: it just cons people into feeling well.

But Dr Cummings believes the research shows that both acupuncture and sham acupuncture are “super placebos”: they are significantly better than other treatments like painkillers, or no treatment at all, because touch and pressure bring a physical response in the body. In a recent editorial in the British Medical Journal he pointed out that trials don’t show much difference between “real” and “sham” acupuncture because they both work.

“Acupuncture brings a real neurophysiological response, which is more than placebo. Touching, pressure, pointing to where the pain is all have an effect, and you can measure their effects in the nervous system.”

Many traditional acupuncturists are arriving at similar conclusions. Tsagaris is adamant that while the language differs, Eastern and Western approaches are not so far apart. “Five thousand years ago in China, they didn’t have MRI scans, so they developed their own means of describing it,” he says.

“The idea of energy in acupuncture is a huge concept. But basically inserting a needle provokes an electrical and biochemical response which promotes communication between cells. It’s been shown that it ignites the brain with electrical lights.”

Proof is a hard thing to come by for any treatment, and acupuncture — whether it be traditional or Western — is going to take longer than most to crack. In the meantime, spending your painkiller cash on a relaxing acupuncture session doesn’t seem such a stupid move.