Benefits
Salivary cortisol and mood scores in stressed adults on a two-herb blend
In a maker-funded four-week trial in 56 moderately stressed adults, Relora, a magnolia plus phellodendron blend standardized to honokiol and berberine, lowered salivary cortisol exposure by 18 percent versus placebo and improved mood-state scores, including stress, depression, fatigue and vigor. Honokiol itself was not tested alone.
Short-term anxiety scores in women taking a two-herb blend
In a maker-funded six-week trial of Relora in overweight premenopausal women, 40 enrolled and 26 finished. State (short-term) anxiety fell more than with placebo, while the trait questionnaire for long-standing anxiety or depression and salivary cortisol did not differ. Phellodendron adds berberine, so the result cannot be credited to honokiol.
Sleep quality and time to fall asleep in a two-herb blend trial
The six-week Relora trial in women measured sleep quality and how long it took to fall asleep, and neither differed from placebo. In mice, injected honokiol shortened the time to fall asleep and increased non-REM sleep through the benzodiazepine site of the GABA-A receptor, which has not been shown in people.
Body weight in women who eat more under stress (two-herb blend)
In a six-week maker-funded Relora trial of overweight women who eat more under stress, a separate weight report found an average 1.5 kg gain within the placebo group and none with Relora, and fewer Relora users gained 1 kg or more. The trial's main report found no difference in appetite or body measures versus placebo. Neither tested honokiol alone.
GABA-A receptor activity in mouse and cell studies
In brain neurons and cells carrying recombinant GABA-A receptors, honokiol increased receptor responses, most strongly at delta-containing subtypes. In mice, flumazenil, which blocks the benzodiazepine site, cancelled its sleep effect. This is animal and cell work and has not been measured in people.
Mechanism of action
GABA-A receptor activity (mouse and cell studies)
Honokiol enhanced both synaptic and extra-synaptic GABA-A currents in hippocampal neurons and modulated every recombinant receptor subunit combination tested, most strongly delta-containing receptors. In mice, its sleep effect and the activation of sleep-promoting neurons disappeared when the benzodiazepine site was blocked with flumazenil.
Cortisol and the HPA axis
The cortisol finding belongs to Relora, the magnolia plus phellodendron blend, not to honokiol on its own. A four-week trial in 56 moderately stressed adults saw salivary cortisol exposure 18 percent lower than placebo, while a six-week trial in women found no difference; its separate weight report described only a nonsignificant trend toward lower evening cortisol.
Antibacterial activity against mouth bacteria (lab data)
In test-tube work, honokiol, magnolol and magnolia bark extract stopped the growth of Porphyromonas gingivalis, Fusobacterium nucleatum and Streptococcus mutans at 8 to 31 micrograms per milliliter. No human study has tested honokiol on its own for this.
Clinical trials
Randomized double-blind placebo-controlled trial of Relora 250 mg at breakfast and dinner (500 mg a day) for 4 weeks, funded by the maker Next Pharmaceuticals (Talbott et al. 2013, J Int Soc Sports Nutr)
56 adults (35 men, 21 women) screened for moderate stress.
Salivary cortisol exposure was 18 percent lower than placebo (p<0.05), and mood-state scores were better, including overall stress (11 percent lower), depression (20 percent lower), fatigue (31 percent lower) and vigor (18 percent higher). No adverse events were reported. The product is a magnolia plus phellodendron blend standardized to honokiol and berberine.
Randomized double-blind placebo-controlled pilot of Relora 250 mg three times daily for 6 weeks; the blend is standardized to not less than 1.5 percent honokiol and 0.1 percent berberine, and the study was funded by the maker Next Pharmaceuticals (Kalman et al. 2008, Nutr J)
40 overweight premenopausal women aged 20 to 50 who eat more under stress and scored above average for anxiety; 26 completed (16 on Relora, 10 on placebo).
Relora lowered state (short-term) anxiety versus placebo. Trait anxiety, salivary cortisol and amylase, appetite, body measures, sleep quality and time to fall asleep did not differ from placebo. Three people left because of adverse events (two on Relora, one on placebo). A separate weight report from this trial (Garrison et al. 2006) found fewer Relora users gained 1 kg or more (37 versus 75 percent).