Research supporting various uses of cannabis or cannabis-based medications provides low-quality evidence and is often beset with significant potential for bias, MedPage Today reports. Publishing their findings in JAMA, the Journal of the American Medical Association, researchers conducted a meta-analysis of 79 cannabis trials including 6,462 participants.

Only four of the studies were judged to be at low risk for bias, with 70 percent deemed at high risk. In some cases, this potential for bias resulted from a lack of double-blinding or placebo controls.

Four studies of medical pot or a cannabinoid compound’s effect on low appetite among people with HIV showed that the compounds had little effect. Three of the trials used a placebo; all four had a high risk of bias.

Various compounds tested in the trials did not show a benefit for nausea, vomiting or appetite among people with various health problems. Cannabinoids showed a slight benefit in the realms of chronic pain and spasticity (stiffness or muscle spasms resulting from multiple sclerosis).

The compounds were linked to an increased risk of various side effects, including dizziness, dry mouth, nausea, fatigue, drowsiness, euphoria, vomiting, disorientation, drowsiness, loss of balance, and hallucination.