TABLE 3

Summary of the trials of complementary and alternative medicines (CAM) with significant positive outcomes

CAM	Intervention group	Control group	Treatment duration	Sample size n	Positive outcome^#	Intervention group	Control group	p-value	Overall risk of bias score^¶	First author [ref.]	Country
Curcumin	“Standard” asthma treatment + 500 mg curcumin twice daily	“Standard” asthma treatment	4 weeks	60	Change in FEV₁	−0.72±0.69	−0.83±0.63	<0.05	Low	Abidi [11]	India
New Zealand green-lipped mussel	Two 150 mg capsules of New Zealand green-lipped mussel twice daily	Two 150 mg placebo capsules twice daily	8 weeks	46	Mean morning PEFR	47.0±11.7	−33.4±6.2	<0.001	Low	Emelyanov [12]	Russia
*Solanum xanthocarpum*	Solanum xanthocarpum 300 mg orally once daily	Salbutamol 4 mg once daily or deriphylline 220 mg	N/A (1 day)	60	% increase predicted FVC, % increase predicted FEV₁, % increase PEFR	22.4±2.3 25.3±2.7 20.8±2.4	32.2±3.7 41.9±6.9 37.5±6.0	<0.05 <0.05 <0.01	Unclear	Govindan [13]	India
Aqueous extract of propolis	13% aqueous extract of propolis in water orally once daily + oral theophylline	Placebo sachet in water orally once daily + oral theophylline	8 weeks	45	% increase FVC, % predicted FEV₁, % increase PEFR	18.7% 29.5% 29.8%	Not available Not available Not available	0.0001 0.0001 0.0001	Unclear	Khayyal [18]	Egypt
n-3 Polyunsaturated fatty acids	10–20 g of perilla seed oil daily (n-3 fatty acids)	10–20 g of corn oil daily (n-6 fatty acids)	4 weeks	14	FVC, % predicted FEV₁	Exact data not available 96.7 (85.4–108.0)	Exact data not available 90.9 (75.9–105.8)	<0.05 <0.05	Unclear	Okamoto [28]	Japan
ASHMI	Four oral ASHMI capsules 3 times a day + prednisone placebo	Prednisone 20 mg once daily + ASHMI placebo	4 weeks	91	Mean change in FEV₁ and PEFR	19.4±5.5 20.1±5.6	23.2±8.9 23.0±7.5	0.02 0.04	Unclear	Wen [30]	China

ASHMI: anti-asthma herbal medicine intervention; FEV₁: forced expiratory volume in 1 s; PEFR: peak expiratory flow rate; FVC: forced vital capacity. ^#: lung function only; ^¶: the risk of bias was assessed using the Cochrane Risk of Bias Assessment Tool [10]. An “unclear” judgement was made if insufficient detail was provided. The mean judgement from the seven domains was used to determine the “overall” score.