I read more than 50 scientific studies about yoga. Here’s what I learned.

It’s not just me.

The most recent survey suggests more than 20 million Americans practice yoga, making it one of the most popular forms of exercise. Even Vladimir Putin, a devotee of “macho sports,” added downward dog to his repertoire.

But is yoga really that great for health compared with other exercises? Does it really help improve our response to stress or correct bad posture, as often promised? Maybe our perceptions about yoga are biased. Or maybe, as some critics have pointed out, there are downsides to yoga. Who can forget the controversial New York Times story from 2012 suggesting that some people get seriously injured, or even die, on their yoga mats.

I wanted a more objective view on the health effects of yoga, so I turned to science, reading more than 50 studies and review articles and talking to seven of the world’s leading yoga researchers. Almost immediately, I was struck by how weak the research on yoga is. Most studies were small and badly designed or plagued by self-selection bias. Making matters worse, there are so many varying styles of yoga that it’s tough to say how meaningful evidence about one style is for others.

Still, what I learned is that there are a few things we can say about yoga, based on the available research. Yoga probably won’t hurt you, despite what haters claim, and it appears to be just as good for your health as other similar forms of exercise.

Even more, yoga seems to help alleviate lower back pain, improve strength and flexibility, and reduce inflammation in the body — which, in turn, can help stave off chronic disease and death. Emerging research suggests yoga can increase body awareness, or attention to the sensations and things going on inside you. That’s no small matter: Researchers think heightened body awareness can improve how well people take care of themselves.

Keep in mind, however, that other mind-body exercises — such as tai chi or meditation — can boost body awareness and reduce inflammation, too. That’s the catch with a lot of yoga research: It still hasn’t told us how much better or different yoga is for a number of health measures when compared with other forms of exercise. Finally, many of the most outlandish claims people make about yoga, like the idea that it can alleviate constipation or wring out toxins, either aren’t backed by science or haven’t been studied at all.

What is the state of yoga science?

The first randomized trial (or high-quality experiment) on yoga was published in 1975 in The Lancet. It showed that yoga was more effective than relaxation for reducing high blood pressure. But that trial only involved 34 participants, and all of them already had high blood pressure, so it is difficult to know whether the effect of the yoga would bear out in a larger trial of healthy people.

Since then, the number of yoga studies has dramatically increased, but the field is plagued by some of the same problems of that early study. Many yoga studies still involve small numbers of participants. Many lack a control group. Many don’t compare yoga to activities we’d be interested in comparing it to. (Ideally, for instance, we’d want to know how yoga measures against another form of exercise or mind-body practice — not, as one study examined, comparing whether it’s better for back pain than giving people a book on how to manage their back pain.)“For most conditions, the main problem is we don’t have enough evidence yet”

What studies do exist are often short term. There are no long-term studies on mortality or serious disease incidence. There are few long-term studies on the potential harms yoga can wreak on the body. “For most conditions,” says Holger Cramer, director of yoga research at the University of Duisburg-Essen in Essen, Germany, “the main problem is we don’t have enough evidence yet.”

Studying yoga is also tricky. Researchers generally believe blinded studies are the highest quality of research, because participants involved don’t know what intervention (such as a drug) they are receiving and their biases and perceptions don’t color the outcomes. But you can’t blind people to the fact that they’re doing yoga.

Then there’s the biggest question at the center of yoga research: How do you define yoga? “Yoga is many things to many people,” said Karen Sherman, a researcher affiliated with the Group Health Research Institute. “What you put into a yoga intervention probably impacts what you get out.”

Yoga usually involves some combination of the following: postures and poses (asanas), regulated breathing (pranayama), and meditation and relaxation (samyana). But many classes mix in other elements, from chanting to heating to music. There’s also a lot of variation in teaching quality and style. Hatha and Iyengar yoga are mostly made up of stretches and restorative poses, while ashtanga and vinyasa tend to be more vigorous and athletic. Yin yoga probably won’t make you sweat: You mostly hold postures for long periods of time for very deep stretches. In Bikram, which consists of 26 postures repeated twice in a room that’s heated at 105 degrees, you can be sure you’ll drench your yoga clothes in perspiration.

Lorenzo Cohen, chief of the integrative medicine section at MD Anderson Cancer Center, told me: “Many papers [on yoga] don’t have enough of an in-depth description of what they mean by ‘yoga.’ What was the level of training of yoga therapists? How did they choose different postures or breathing exercises?”

What’s more, there are so many components in a yoga class, it’s tough to know what might be having an affect on health: If people report feeling better after a class, was that due to the experience of being part of a larger group? Was it the teacher’s style? Was it the breathing exercises? The heat? These factors are difficult to isolate, and some of the ways yoga helps people might be hard for scientists to measure.

Still, the yoga researchers I spoke to said the quality and quantity of studies has been improving, so we may get better answers soon. “There are more researchers conducting yoga therapy studies, and when the smaller trials suggest benefit, that leads to larger, better-designed trials,” said Cohen. There is also a move to study the biological outcomes of yoga — how classes affect things like hormone levels — which will lead us to a more objective picture of what yoga can do for the body.

1) Is yoga likely to hurt you?

No, probably not.

This question first came up in 2012, when the New York Times published a splashy article suggesting that yoga can wreck your body. The piece, adapted from the book The Science of Yoga: The Risks and Rewards, suggested yoga caused widespread harm to its practitioners — from ruptured disks and stroke to brain injury.

But that piece was largely based on cherry-picked anecdotes, exaggerating these horrible cases to suggest they were representative of the broader yoga experience when they simply aren’t.

Cramer has studied published reports of injuries and other harms from yoga for several review and told me this: “We found yoga is as safe as any other activity. It’s not more dangerous than any other form of exercise.” He added: “Yoga is not 100 percent safe, but nothing is 100 percent safe.”

In a 2013 review of case studies, Cramer identified 76 unique incidents of adverse events from yoga. “Most adverse events affected the musculoskeletal, nervous, or visual system,” he concluded. “More than half of the cases for which clinical outcomes were reported reached full recovery, 1 case did not recover at all, and 1 case died.”

Most often, people got into trouble with the headstand pose, followed by shoulder stand and postures that required putting one or both feet behind the head

Most often, people got into trouble with the headstand pose, followed by shoulder stand, postures that required putting one or both feet behind the head, the lotus position, and forceful breathing. Pranayama-, hatha-, and Bikram-style yoga practices had the most adverse events associated with them.

Based on these cases, Cramer and his co-author had this practical advice for how to stay safe in yoga: Beginners should avoid advanced postures (such as headstands), and people with chronic health conditions (such as glaucoma) should consult their doctors before diving in. “Yoga,” they added, “should not be practiced while under the influence of psychoactive drugs.”

As for long-term yoga harms, Cramer pointed to two studies on joint and cervical disc degeneration in people who have been doing yoga for a while. But the studies had contradictory results, “so long-term health consequences cannot be clearly derived from the available evidence,” Cramer said.

I asked MD Anderson’s Cohen for his take. “There can, of course, be negative consequences if done incorrectly, like any body manipulation,” he said, “but if you have the right teacher this will not happen.” Even if a lot of yoga over a lifetime leads to injury, it’s not clear those harms outweigh the benefits, or that people would have been better off running or weightlifting all the time.

2) How does yoga compare with other forms of exercise?

The short answer is: It depends on the type of yoga. Yoga classes that involve nothing more than lying around on piles of blankets and breathing aren’t likely to build your muscles. But more strenuous types of yoga like ashtanga can be surprisingly similar to other forms of vigorous exercise.

“Some studies show, depending on yoga style, that it doesn’t improve fitness as much as other forms of exercise,” Cramer says. “But for rigorous and intense styles [of yoga] — that can be a form of exercise. It strongly depends on what you do when you do yoga.”

For example, a few high-quality studies have shown that certain types of yoga can indeed make people stronger. One small, randomized trial in the Journal of Strength and Conditioning Research — which compared eight weeks of Bikram yoga with no intervention in 32 young, healthy, adults — found that the yoga classes improved people’s deadlift strength.

The Bikram classes didn’t, however, improve the participants’ aerobic capacity. Another before-and-after study, published in the Journal of Clinical Nursing, found that hatha yoga could improve aerobic fitness in older people. Still, it’s not clear that yoga is ideal here compared with, say, running or swimming.

Considering the mixed findings, a bit of common sense is helpful: If you go to a class mainly focuses on relaxation and doesn’t elevate your heart rate, you’re probably not getting a good cardio workout or building muscles.

If you go to a more athletic yoga class that tires out your muscles and makes you pant, you can probably count on it helping to make you stronger. If you are panting in a yoga class, you’re probably improving aerobic capacity to some extent. That said, if your main goal is building aerobic capacity, you might want to try running or swimming instead of or in addition to yoga.

3) Does yoga really reduce stress and anxiety?

For all the talk of yoga lifting moods and calming people, the studies on this question are still surprisingly inconclusive.

Karen Pilkington, a research fellow at the University of Westminster, knows this research better than anybody: She sifted through the science for systematic reviews on yoga for both anxiety and depression. (Here’s another, more recent systematic review of the evidence on yoga for depression by other authors.)

“There are indications that it might be helpful,” Pilkington says. “But lots of the studies are quite small, and we definitely need more and bigger studies.” In other words, yoga might help with mood disorders, but we don’t yet know for sure because the studies to date have generally been so badly designed. “We can’t say there’s conclusive evidence that yoga works for depression and anxiety,” she explained. “We’re still really exploring yoga as an idea for testing and the best ways of testing it.”

One complication here: When it comes to conditions like anxiety and depression, it can be difficult to untangle whether it’s the yoga that’s helping or simply the act of going out, joining a group on a regular basis, and so on. Even if yoga turns out to be unequivocally beneficial for depression, other research has shown that exercise in general can alleviate depression. So, again, it’s not clear that yoga is the only way to get this benefit.

As for stress reduction, there are a few randomized trials — the highest-quality evidence — that have suggested yoga is about as effective as simple relaxation programs, more helpful than no intervention, but not as effective as stretching. Pretty mixed results.

4) Does yoga have long-term health benefits?

Possibly. To be clear, there’s currently no direct evidence on yoga’s long-term benefits. Researchers simply haven’t tracked yogis over a span of 20 years or more and followed up to see whether they get diseases at a lower rate than non-yogis.

But that’s not the whole story. There are also some randomized controlled trials suggesting that yoga may improve quality of life for diabetes patients, reduce cardiovascular disease risk factors, and even help people manage high blood pressure

How can this be? One possibility is that yoga can help reduce inflammation in the body — which turns out to be surprisingly beneficial.

You can think about inflammation in two ways. There’s helpful inflammation, as when your body’s immune system mounts a response to bacteria in a cut. There’s also harmful inflammation. When you’re stressed, your body’s inflammatory response can go into overdrive, hampering its ability to fight off viruses and disease. People who are inactive, obese, or eat an unhealthy diet have higher levels of harmful inflammation. And researchers have found associations between inflammation and various chronic diseases, including cancer, cardiovascular disease, and diabetes.

Yoga — like other mind-body exercises such as tai chi and meditation — seems to be particularly helpful at reducing harmful inflammation. A 2014 meta-analysis on the effects of mind-body therapies on the immune system found that yoga reduces inflammation-based blood markers. So did this 2014 randomized control trial looking at women with breast cancer and breast cancer survivors.

Michael Irwin at UCLA’s medical school, one of the authors of a 2015 descriptive review on inflammation and mind-body exercises, said, “When you look at the aerobic exercise necessary to decrease inflammation, people have to maintain very vigorous levels.” But not with yoga, he continued. “Even practices with minimum levels of physical activity [like Iyengar stretches] can have large effect sizes.” Researchers don’t yet know why, though they think the meditative components of yoga, tai chi, and meditation may have something to do with it.

“When you compare long-term yoga practitioners to people doing other forms of exercise, you have better body awareness in yogis”

In the past few years, some academics have also begun to study “body awareness” and its relationship to a number of health outcomes. Body awareness, according to this article in PLOS One, “involves an attentional focus on and awareness of internal body sensations.” The idea is that in becoming more aware of your body, you might be more likely to notice when you’re stressed, in pain, or tired, and perhaps less likely to abuse your body.

This 2013 study, involving 18 people with chronic neck pain who practiced Iyengar yoga once a week for nine weeks, found that participants reported increased body awareness after practicing. In this 2005 study, which compared three groups of women — 43 doing Iyengar and ashtanga yoga, 45 doing aerobic exercise, and 51 who did neither activity — the yogis reported more body awareness and body satisfaction than the other groups. These were both qualitative studies, not experimental trials, so it’s difficult to know whether the yogic types are more “body aware” to begin with or whether the yoga made the difference.

“When you compare long-term yoga practitioners to people doing other forms of exercise,” said Cramer, “you have better body awareness in yogis.”

5) Does yoga help with lower back pain?

Yes. Yoga does seem to help alleviate lower back pain, in both the short and long term. The evidence around this is some of the best yoga research we have. This meta-analysis of the evidence on yoga for back pain, from 2013, sums it up.

[There’s] found strong evidence for short-term effectiveness and moderate evidence for long-term effectiveness of yoga for chronic low-back pain in the most important patient-centered outcomes. Given the low number of adverse events, yoga can be recommended as an additional therapy to patients who do not improve with education on self-care options.

That said, there are some caveats here, too. The leading researcher on this question, Karen Sherman, notes that it’s still not clear whether yoga is any better than any other form of exercise for back pain. “It’s certainly a viable form of exercise, but is it better than other exercises [for back pain]? Probably not,” she says.

Some of the best research on lower back pain has only looked at one specific type of yoga — viniyoga.

Researchers haven’t figured out why yoga helps with back pain. This Cochrane Review protocol offered a few guesses, suggesting that improved flexibility and muscle strength, as well as relaxation and body awareness, may help.

6) Does yoga improve flexibility and balance?

The research here is limited. But there are a few high-quality controlled trials that suggest yoga can help increase flexibility in young people, healthy seniors, and computer users. One small study on stroke patients suggested that yoga may improve balance, as did this study in healthy young adults.

7) Can yoga really stimulate digestion or wring out toxins?

Anyone who has taken a yoga class has probably been exposed to wild claims by the teacher that certain poses will do anything from wring out toxins to stimulate a particular part of the colon and alleviate constipation. But the science backing up these kinds of very specific claims was scant, so I asked one of the researchers, Cohen, about where they come from.

“[Teachers are] basing it on personal experience, on anecdote, on the lineage of practice that’s been handed down,” he said. “They are probably not basing it on Western-style analytic techniques that followed a control trial design. We just haven’t gotten there yet with yoga research, testing particular poses or breathing techniques for particular outcomes.”