Two words decide whether a bottle of fish oil is harmless or dangerous. Not the words on the label. The words “together with” or “in place of.”
The National Center for Complementary and Integrative Health, the NIH center that studies these approaches, draws the line right there. A non-mainstream approach used alongside conventional medical care is complementary. The same approach used instead of conventional care is alternative. NCCIH says genuine alternative use is rare.
That sounds like vocabulary. It isn’t. A Yale research team found that patients with early-stage breast, prostate, lung or colorectal cancer who used alternative medicine as their only anticancer treatment carried about two and a half times the risk of death compared with matched patients who got standard care.
So the distinction carries real weight. What it does not carry is any information about whether the thing works. That’s the part almost everyone gets backwards.
The definition is about how you use it, not what it is
NCCIH’s three terms are easy to mix up, so here they are with the difference made obvious.
Complementary means you’re using a non-mainstream approach on top of your regular medical care. You keep the blood pressure medication and add tai chi.
Alternative means you’re using it instead. You drop the blood pressure medication and keep only the tai chi.
Integrative means conventional and complementary care brought together in a coordinated way, with one plan and providers who talk to each other. Whether a given clinic does that is a separate question, handled in its own article here.
Notice what changes between the first two. Not the product. Not the dose. Not the practitioner. Only what else you are or aren’t doing. The identical bottle of fish oil is complementary when it sits next to a statin and alternative when it replaces one. The risk profile of those two situations is not remotely similar.
This also means nobody can tell you from the outside which category you’re in. Your physician can’t see the gap between what was prescribed and what you’re taking. Only you know that.
The four families, and the ones that don’t fit the grid
NCCIH organizes these approaches by what they mainly act through. In practice, four groupings cover most of what people use.
Nutritional is by far the largest. Dietary supplements, herbal products, probiotics and special diets all live here. This is where the money is, where the interactions are, and where the regulatory gaps are.
Psychological approaches work through the mind. Mindfulness practice is the common example.
Physical approaches work through the body from the outside. Massage therapy and spinal manipulation are the ones Americans use most.
Mixed approaches refuse to sit in one box. Yoga, tai chi and acupuncture combine psychological and physical elements. Mindful eating combines psychological and nutritional ones. Older NCCIH material used two broader labels you’ll still see in print: “natural products” for the nutritional group and “mind and body practices” for the rest.
Then there’s a category the grid can’t handle. Traditional healing practices, Ayurvedic medicine, traditional Chinese medicine, homeopathy, naturopathy and functional medicine aren’t single interventions. Each arrives with its own theory of what causes disease and its own way of deciding what’s wrong with you.
That matters more than it sounds. A massage is a massage. But a whole system may give you a diagnosis your physician does not recognize, and a plan built on it. When two frameworks disagree about what your lab results mean, you’re the one standing in the middle.
How many people do this
The numbers are larger than most readers expect, and they’re worth stating with their dates attached.
A 2012 national survey found that more than 30% of US adults and about 12% of children had used an approach outside typical conventional care. NCCIH also reports roughly 36% of adults in 2002 and about 38% in 2007. In 2012, 17.7% of adults had taken a dietary supplement other than a vitamin or mineral in the past year, with fish oil the most common of those.
Two practices grew fast. Yoga use among adults went from 9.5% in 2012 to 14.3% in 2017. Meditation went from 4.1% to 14.2% over the same five years, more than tripling.
Those figures come from NCCIH’s overview page, which was last updated in April 2021, so the freshest numbers on it are from 2017. More recent work exists: a 2024 research letter in JAMA by Richard Nahin and colleagues at NCCIH compared survey years 2002, 2012 and 2022. NCCIH’s summary of that analysis says overall use climbed across the two decades, that yoga, meditation and massage increased the most, and that more adults were using these approaches for pain in 2022 than in 2002. The published percentages weren’t available to verify here, so treat the direction as solid and the exact figures as something to look up.
Supplements have their own tracking, and it’s more current. The CDC’s National Center for Health Statistics reported in February 2021 that 57.6% of adults aged 20 and over had used a dietary supplement in the past 30 days during 2017 to 2018. Women used them more than men, 63.8% versus 50.8%. Among women 60 and older, use hit 80.2%. Across all adults, 13.8% used four or more products; among adults 60 and over, 24.9% did. Back in 2007 to 2008 the age-adjusted figure was 48.4%, and by 2017 to 2018 it had reached 56.1%. The most common products were multivitamin-mineral combinations, then vitamin D, then omega-3 fatty acids.
Read that first number again. More than half of American adults took a supplement in the past month. This is not fringe behavior. It’s median behavior.
The label tells you nothing about whether it works
Below, we cover the correction that does the most work in this article. “Complementary” describes a pattern of use. It says nothing about quality, safety or effectiveness. Inside the category, the evidence runs from genuinely strong to genuinely absent, and you have to sort practice by practice and condition by condition.
Established for specific uses. Acupuncture for chronic pain is the best-documented example. Andrew Vickers and colleagues pooled individual patient data from 39 randomized trials covering 20,827 patients with musculoskeletal pain, osteoarthritis, chronic headache or shoulder pain, published in The Journal of Pain in 2017. Acupuncture outperformed both sham acupuncture and no-acupuncture control for every one of those conditions. The gap was close to half a standard deviation against no acupuncture and roughly 0.2 standard deviations against sham, and about 15% of the effect faded by one year. The authors concluded that the pain reduction cannot be explained by placebo effects alone.
Notice the honesty inside that finding. The margin over a fake needle is small. The margin over nothing is substantial. Both facts are true at once, and anyone quoting only one of them is selling something.
Preliminary. NCCIH describes evidence for massage for cancer pain, yoga for fatigue after cancer treatment, and yoga and meditation for quitting smoking as preliminary. Preliminary means promising and unfinished. It does not mean proven, and it does not mean worthless.
Insufficient. This is most of the field. NCCIH notes that the amount of evidence varies widely and that acupuncture, yoga, spinal manipulation and meditation are the most-studied approaches, which is another way of saying everything else has been studied less.
Tested and failed. NCCIH states directly that large, rigorous studies of several dietary supplements found they did not work for the conditions being tested. A negative trial is a real result, and it belongs in the same category as a positive one.
The practical consequence: “acupuncture works” is not a claim anyone can evaluate. “Acupuncture modestly reduced chronic musculoskeletal pain compared with sham needling in pooled trial data” is. Every useful sentence in this field names a specific practice, a specific condition and a specific comparison. Sentences that skip those three things are advertising.
A related point about history. Many of these approaches have been used for hundreds or thousands of years. That’s a fact about culture, not about biology. Long use tells you a practice was tolerable and meaningful to people. It does not tell you whether it changes the course of a disease, because nobody was running controlled comparisons.
The safety problems that belong to this category specifically
Three risks show up here that don’t show up the same way in conventional care.
The regulatory gap. The NIH Office of Dietary Supplements lays it out: medicines need FDA approval before they can be sold, and supplements do not. The burden sits with the company, which has to keep proof that what it sells is safe and that its label tells the truth. That evidence only has to reach FDA first when the product carries a new dietary ingredient, the term for anything introduced after October 15, 1994. Everything else goes on the shelf first. Manufacturing has to meet good manufacturing practice rules, a product shown to be unsafe can be pulled or recalled, and claims in the advertising fall to the Federal Trade Commission. All of that is after-market enforcement, not pre-market proof.
Interactions. The Office of Dietary Supplements names concrete ones. Vitamin K can reduce how well warfarin works. St. John’s wort speeds up the breakdown of many medicines, which can drop a drug below the level that was doing the job. A few supplements can make you bleed more easily, and a few change the way your body handles anesthesia, which is why a surgical team wants your full list well before the date. Risk goes up with high doses, with combining many products, and with substituting a supplement for something prescribed.
Contamination. NCCIH reports that supplements have been found containing hidden prescription drugs and other undeclared compounds, and that this shows up especially in products sold for weight loss, sexual performance, and athletic or bodybuilding use. If a product is marketed around those three goals, the odds that it contains something not on the label go up.
The harm is measurable. A 2015 study in the New England Journal of Medicine by Andrew Geller and colleagues used surveillance from 63 US emergency departments between 2004 and 2013 and estimated 23,005 emergency department visits a year from adverse events related to dietary supplements, leading to about 2,154 hospitalizations. Adults aged 20 to 34 accounted for 28.0% of visits and unsupervised children for 21.2%. Weight-loss and energy products were behind 71.8% of the supplement-related events involving palpitations, chest pain or a racing heart. Among adults 65 and older, choking or a pill stuck in the throat caused 37.6% of these visits, with vitamins and minerals involved in 83.1% of those.
Keep that in proportion. There were 155.4 million emergency department visits in the US in 2022. Twenty-three thousand is a small slice. It is also not zero, and the pattern inside it is specific enough to be useful: the trouble clusters in weight-loss and energy products among young adults, and in swallowing problems among older ones.
One more layer applies to anything delivered by a practitioner rather than swallowed. NCCIH points out there is no standardized national credentialing system for complementary health practitioners. What counts as qualified is set state by state, and it is not the same from one discipline to the next. Checking a practitioner is a subject of its own in this library.
The small group that takes the real risk
Two 2018 studies from the same Yale research team, both using the National Cancer Database, do more to clarify this topic than anything else published.
The first, in the Journal of the National Cancer Institute, identified 281 patients with nonmetastatic breast, prostate, lung or colorectal cancer who used alternative medicine as their sole anticancer treatment. Matched against patients who received conventional treatment, the alternative-only group had a hazard ratio for death of 2.50, with a 95% confidence interval of 1.88 to 3.27. By cancer type, the hazard ratio was 5.68 for breast, 4.57 for colorectal and 2.17 for lung. The authors described this choice as rare.
The second, in JAMA Oncology, looked at complementary use instead of alternative use. Out of 1,901,815 patients, 258 were coded as receiving unproven treatments from nonmedical personnel in addition to at least one conventional treatment. Those patients did not delay starting conventional treatment. But they refused parts of it far more often: radiotherapy 53.0% versus 2.3%, chemotherapy 34.1% versus 3.2%, hormone therapy 33.7% versus 2.8%, surgery 7.0% versus 0.1%. Five-year overall survival was 82.2% compared with 86.6%.
Then comes the finding that explains everything. In a model that left out treatment refusal and delay, complementary use was associated with more than double the risk of death, a hazard ratio of 2.08. Once refusal and delay were added to the model, the association vanished. The hazard ratio dropped to 1.39 with a confidence interval of 0.83 to 2.33, which crosses 1 and so is not statistically meaningful.
Read that plainly: the complementary treatment itself did not appear to be what shortened lives. Refusing conventional treatment did.
The limits deserve stating. Both studies are observational, not randomized. The groups of interest were tiny, 281 and 258 patients. The registry code used to identify complementary medicine is a blunt instrument, not an inventory of what people took. You cannot stretch findings about several hundred cancer patients into a claim about everyone who takes a probiotic.
But the mechanism travels even when the numbers don’t. The hazard in this category is almost never the herb. It’s what the herb replaced, and the months that went by while it did.
Four questions before you add anything
This is a short screen you can run on yourself, and it takes about two minutes.
-
What exactly, how much, how often, and for what? Write down the actual product name, the dose printed on the label, and the problem you’re trying to solve. “Some turmeric for inflammation” is not something anyone can check for interactions.
-
What, if anything, is this replacing? Be literal. If the honest answer includes a prescription you’ve stopped taking, a procedure you’ve postponed or an appointment you didn’t make, you have crossed from complementary into alternative, and the risk picture changes completely.
-
Who else knows? Your prescriber, your pharmacist, your surgeon and your dentist all need the list. The Office of Dietary Supplements suggests keeping a running record of each product’s name, dose, how often you take it and why, and it offers a printable form for exactly that.
-
What’s my stopping rule? Decide in advance what improvement you expect and by when you’d expect to see it. Open-ended use with no endpoint is how people wind up on seven products and can’t tell which, if any, is doing anything.
A fifth consideration isn’t a question so much as a flag. If the person recommending a product is also the person selling it, that doesn’t make the advice wrong, but it is information you should price in.
When to stop improvising and get medical help
Some situations are not candidates for a self-directed experiment, and it’s worth naming them specifically.
Get a medical evaluation, not a supplement, for a new lump, a mole that’s changing shape or color, bleeding you can’t account for, or weight loss you didn’t intend. These symptoms need a diagnosis first. What you do after the diagnosis is a separate conversation.
Get help now, by calling 911, for chest pressure, sudden weakness or numbness on one side, slurred speech, a drooping face or trouble breathing.
Go back to a clinician if you’ve been treating something at home for more than a few weeks with no change. The experiment has given you its answer.
Tell the surgical team everything you take, well before any operation or procedure, including products you consider too minor to mention. Bleeding risk and anesthesia response are the reasons, and both are real.
And if any practitioner, in any field, tells you to stop a prescribed medication, do not stop. Call the clinician who prescribed it and repeat exactly what you were told. That single instruction is the most dangerous thing you are likely to encounter in this space, and it is the one worth rehearsing in advance.
The thing to carry out of all this is simple. The word “complementary” describes your behavior, not the product’s quality. It tells your doctor how you’re using something. It tells you nothing about whether it helps. Those are two different questions, and you need an answer to both.
Sources
- National Center for Complementary and Integrative Health. “Complementary, Alternative, or Integrative Health: What's In a Name?.” nccih.nih.gov
- National Center for Complementary and Integrative Health. “Are You Considering a Complementary Health Approach?.” nccih.nih.gov
- National Center for Complementary and Integrative Health. “Safety.” nccih.nih.gov
- National Center for Complementary and Integrative Health. “About NCCIH.” nccih.nih.gov
- National Center for Complementary and Integrative Health. “National Health Interview Survey (NHIS).” nccih.nih.gov
- National Center for Complementary and Integrative Health. “Use of Complementary Health Approaches Overall and for Pain Management by U.S. Adults in 2002, 2012, and 2022.” nccih.nih.gov
- National Center for Complementary and Integrative Health. “Credentialing, Licensing, and Education.” nccih.nih.gov
- NIH Office of Dietary Supplements. “Dietary Supplements: What You Need to Know.” ods.od.nih.gov
- Centers for Disease Control and Prevention, National Center for Health Statistics. “Dietary Supplement Use Among Adults: United States, 2017-2018 (Data Brief No. 399).” cdc.gov
- Centers for Disease Control and Prevention, National Center for Health Statistics. “FastStats: Emergency Department Visits.” cdc.gov
- PubMed. “Vickers AJ et al. "Acupuncture for Chronic Pain: Update of an Individual Patient Data Meta-Analysis." The Journal of Pain, 2017. DOI: 10.1016/j.jpain.2017.11.005.” doi.org
- PubMed. “Johnson SB et al. "Use of Alternative Medicine for Cancer and Its Impact on Survival." Journal of the National Cancer Institute, 2018. DOI: 10.1093/jnci/djx145.” doi.org
- PubMed. “Johnson SB et al. "Complementary Medicine, Refusal of Conventional Cancer Therapy, and Survival Among Patients With Curable Cancers." JAMA Oncology, 2018. DOI: 10.1001/jamaoncol.2018.2487.” doi.org
- PubMed. “Geller AI et al. "Emergency Department Visits for Adverse Events Related to Dietary Supplements." New England Journal of Medicine, 2015. DOI: 10.1056/NEJMsa1504267.” doi.org
- PubMed. “Nahin RL, Rhee A, Stussman B. "Use of Complementary Health Approaches Overall and for Pain Management by US Adults." JAMA, 2024. DOI: 10.1001/jama.2023.26775.” doi.org