The first global review of complementary medicines for weight loss in 16 years suggests their use cannot be justified based on current evidence.
Researchers found that while some herbal and dietary supplements resulted in marginal weight loss compared to a placebo, they did not benefit health.
They called for more research into their long-term safety.
The sale of diet pills, powders and liquids containing plant or animal products has grown in popularity.
The global industry was estimated to be worth $41bn (£29.3m) last year.
- Appetite drug could mark ‘new era’ in obesity
- ‘I’ll never go back to the gym again’
- Instagram fixes appetite suppressant mistake
Just 20% of new products are audited annually to make sure they provide evidence backing their claims.
In some countries, the only requirement is that supplements contain acceptable levels of non-medicinal products.
Unlike pharmaceutical drugs, clinical evidence for their safety and effectiveness is not required before they hit the market, says the review’s lead author Erica Bessell from the University of Sydney.
“Our rigorous assessment of the best available evidence finds that there is insufficient evidence to recommend these supplements for weight loss,” she said.
“Even though most supplements appear safe for short-term consumption, they are not going to provide weight loss that is clinically meaningful.”
The Australian researchers carried out a systematic review of all randomised trials comparing the effect of herbal supplements to placebos (or dummy treatment) on weight loss, up to August 2018.
The data was analysed from 54 studies involving 4,331 healthy overweight or obese adults aged 16 years or older.
Weight loss of at least 2.5kg (5.5lbs) was considered clinically meaningful.
Herbal supplements included in the analysis were green tea, garcinia cambogia, mangosteen, white kidney bean, ephedra, African mango, yerba mate, veld grape, licorice root and East Indian globe thistle.
The analysis found that only one of these – white kidney bean – resulted in a statistically, but not clinically, greater weight loss of 1.61kg (3.5Ibs) than the placebo.
Some combination preparations containing African mango, veld grape, East Indian globe thistle and mangosteen showed promising results, but were investigated in three or fewer trials, often with poor research methodology or reporting.
These findings should be interpreted with caution, researchers say.
The researchers also carried out a systematic review, up to December 2019, of 67 randomised trials comparing the effect of dietary supplements to placebos for weight loss in 5,194 healthy overweight or obese adults aged 16 years or older.
Dietary supplements included in the analysis were: chitosan, glucomannan, fructans and conjugated linoleic acid.
The analysis found that chitosan (-1.84 kg), glucomannan (-1.27 kg), and conjugated linoleic acid (-1.08 kg) resulted in statistically, but not clinically, significant weight loss compared to placebos.
“The tremendous growth in the industry and popularity of these products underscores the urgency for conducting larger more rigorous studies to have reasonable assurance of their safety and effectiveness for weight loss.”
‘Quick-fix solution’
Some dietary supplements, including modified cellulose and blood orange juice extract, showed promising results but were only investigated in one trial and need more evidence before recommending them for weight loss, researchers say.
“Herbal and dietary supplements might seem like a quick-fix solution to weight problems, but people need to be aware of how little we actually know about them,” says Ms Bessell.
She said very few high-quality studies had been performed on the supplements, with little data on long-term effectiveness.
“The tremendous growth in the industry and popularity of these products underscores the urgency for conducting larger more rigorous studies to have reasonable assurance of their safety and effectiveness for weight loss,” she added.
The findings are being presented at The European Congress on Obesity held online this year.