The Safety and Comparative Effectiveness of Non-Psychoactive Cannabinoid Formulations for the Improvement of Sleep: A Double-Blinded, Randomized Controlled Trial

Figures & data

Table 1. Study arms and instructions for use in the Radicle Discovery^TM Sleep study.

Study arm	Product formula	Instructions for use
Arm 1	15mg CBD + 15 mg CBN + 5 mg melatonin	“Take 1 capsule each day with food before going to sleep. You may increase to 2 capsules taken at the same time, with food and before going to sleep. Do not take more than 2 capsules per day.”
Arm 2	15mg CBD + 15 mg CBN
Arm 3	15mg CBD
Arm 4	5mg melatonin
Arm 5	15mg CBD Full Spec + 15 mg CBN
Arm 6	15mg CBD + 15 mg CBN + 5 mg CBC

Figure 1. CONSORT diagram.

Table 2. Demographic characteristics of study participants in the Radicle Discovery^TM Sleep study.

	Arm 1		Arm 2		Arm 3		Arm 4		Arm 5		Arm 6		Total		p value^a
Participants [N]	298		298		300		300		299		298		1793
Age [mean (SD)]	45.4	(12.0)	46.7	(11.8)	44.7	(11.1)	44.7	(12.0)	45.5	(11.7)	45.9	(11.7)	45.5	(11.7)	0.7729
Race [N (%)]															0.559
Black	21	(7.0%)	21	(7.0%)	22	(7.3%)	29	(9.7%)	18	(6.0%)	19	(6.4%)	130	(7.3%)
Multi-racial	10	(3.4%)	10	(3.4%)	11	(3.7%)	8	(2.7%)	16	(5.4%)	11	(3.7%)	66	(3.7%)
Prefer not to say	3	(1.0%)	4	(1.3%)	5	(1.7%)	2	(0.7%)	5	(1.7%)	4	(1.3%)	23	(1.3%)
Some other race	12	(4.0%)	13	(4.4%)	23	(7.7%)	17	(5.7%)	10	(3.3%)	16	(5.4%)	91	(5.1%)
White	252	(84.6%)	250	(83.9%)	239	(79.7%)	244	(81.3%)	250	(83.6%)	248	(83.2%)	1,483	(82.7%)
Hispanic, Latino/a/x, or Spanish origin [N (%)]															0.722
No	270	(90.6%)	276	(92.6%)	263	(87.7%)	265	(88.3%)	274	(91.6%)	273	(91.6%)	1,621	(90.4%)
Yes	27	(9.1%)	20	(6.7%)	33	(11.0%)	33	(11.0%)	24	(8.0%)	24	(8.1%)	161	(9.0%)
Prefer not to say	1	(0.3%)	2	(0.7%)	4	(1.3%)	2	(0.7%)	1	(0.3%)	1	(0.3%)	11	(0.6%)
Education [N (%)]															0.881
Less than high school	2	(0.7%)	7	(2.3%)	6	(2.0%)	3	(1.0%)	8	(2.7%)	5	(1.7%)	31	(1.7%)
High school diploma	55	(18.5%)	63	(21.1%)	66	(22.0%)	59	(19.7%)	57	(19.1%)	59	(19.8%)	359	(20.0%)
Some college, no degree	70	(23.5%)	59	(19.8%)	66	(22.0%)	77	(25.7%)	66	(22.1%)	63	(21.1%)	401	(22.4%)
Trade/technical/vocational	25	(8.4%)	24	(8.1%)	23	(7.7%)	21	(7.0%)	25	(8.4%)	28	(9.4%)	146	(8.1%)
Bachelors or associates	107	(35.9%)	105	(35.2%)	96	(32.0%)	106	(35.3%)	106	(35.5%)	111	(37.2%)	631	(35.2%)
Masters or professional	38	(12.8%)	40	(13.4%)	41	(13.7%)	32	(10.7%)	34	(11.4%)	31	(10.4%)	216	(12.0%)
Prefer not to say	1	(0.3%)	0	(0.0%)	2	(0.7%)	2	(0.7%)	3	(1.0%)	1	(0.3%)	9	(0.5%)
Employment status															0.943
Employed	185	(62.1%)	199	(66.8%)	189	(63.0%)	198	(66.0%)	180	(60.2%)	198	(66.4%)	1,149	(64.1%)
Unemployed, not looking	20	(6.7%)	21	(7.0%)	28	(9.3%)	30	(10.0%)	24	(8.0%)	25	(8.4%)	148	(8.3%)
Unemployed, looking	11	(3.7%)	17	(5.7%)	20	(6.7%)	14	(4.7%)	26	(8.7%)	9	(3.0%)	97	(5.4%)
Student	6	(2.0%)	2	(0.7%)	5	(1.7%)	6	(2.0%)	4	(1.3%)	2	(0.7%)	25	(1.4%)
Homemaker	44	(14.8%)	32	(10.7%)	36	(12.0%)	31	(10.3%)	35	(11.7%)	35	(11.7%)	213	(11.9%)
Retired	28	(9.4%)	26	(8.7%)	21	(7.0%)	21	(7.0%)	25	(8.4%)	26	(8.7%)	147	(8.2%)
Prefer not to answer	4	(1.3%)	1	(0.3%)	1	(0.3%)	0	(0.0%)	5	(1.7%)	3	(1.0%)	14	(0.8%)
BMI category [N (%)]															0.522
Underweight	6	(2.0%)	5	(1.7%)	4	(1.3%)	5	(1.7%)	3	(1.0%)	7	(2.3%)	30	(1.7%)
Healthy weight	77	(25.8%)	81	(27.2%)	88	(29.5%)	71	(23.7%)	72	(24.1%)	78	(26.2%)	467	(26.1%)
Overweight	87	(29.2%)	104	(34.9%)	90	(30.2%)	101	(33.7%)	88	(29.4%)	98	(32.9%)	568	(31.7%)
Obese	128	(43.0%)	108	(36.2%)	116	(38.9%)	123	(41.0%)	136	(45.5%)	115	(38.6%)	726	(40.5%)
PROMIS Sleep Disturbance 8A baseline score [mean (SD)]^b	28.0	(6.6)	26.5	(6.9)	27.8	(6.7)	27.5	(6.7)	26.4	(7.6)	28.3	(6.9)	27.4	(6.9)	0.2143
PROMIS Sleep Disturbance baseline category [N (%)]^c															0.2532
None to slight	47	(31.8%)	56	(39.2%)	50	(33.6%)	54	(39.1%)	67	(45.9%)	38	(29.0%)	312	(36.5%)
Mild	40	(27.0%)	40	(28.0%)	39	(26.2%)	25	(18.1%)	24	(16.4%)	31	(23.7%)	199	(23.3%)
Moderate	49	(33.1%)	39	(27.3%)	46	(30.9%)	46	(33.3%)	45	(30.8%)	50	(38.2%)	275	(32.2%)
Severe	12	(8.1%)	8	(5.6%)	14	(9.4%)	13	(9.4%)	10	(6.8%)	12	(9.2%)	69	(8.1%)

BMI = body mass index.

^a p value based on the t-tests for continuous variables, and chi-squared tests for categorical variables.

^b p value based on univariate ANOVA model.

^c Severity thresholds determined after T-score conversion and mapping onto T-score severity thresholds established by the PROMIS Health Organization (PHO) and PROMIS Cooperative Group. p value based on univariate ordinal logistic regression model.

Table 3. Mean difference (MD) in PROMIS Sleep Disturbance 8a scores from baseline to study conclusion, based on adjusted^a linear mixed effects regression model.

	MD (SE)	Statistics	95% CI
Study arm#Week^b
Arm 1	−0.4 (0.2)	t = −1.89, p < 0.059	(−0.8, 0)
Arm 2	0.1 (0.2)	t = 0.62, p < 0.538	(−0.3, 0.5)
Arm 4	−0.3 (0.2)	t = −1.54, p < 0.123	(−0.7, 0.1)
Arm 5	0.1 (0.2)	t = 0.64, p < 0.525	(−0.3, 0.5)
Arm 6	−0.1 (0.2)	t = −0.27, p < 0.786	(−0.5, 0.4)

SE = standard error of the estimate; CI = confidence interval of the estimate.

Multiple imputed linear mixed-effects regression model with random intercept at the individual level and random slope at week level; F(20, 102788.8) = 46.56, p value < 0.0001.

^a Adjusting for sex, age, race, ethnicity and body mass index.

^b Joint-test of main effects significant; F(5, 9414.2) = 2.92, p value = 0.0433.

Table 4. Within-group mean difference (MD) in PROMIS Sleep Disturbance 8a scores from baseline to study conclusion, based on adjusted linear mixed effects regression model.

	MD (SE)	Within-group statistics	95% CI
Arm 1	−8.6 (0.6)	F(1, 595.9) = 185.98, p < 0.0001	(−9.8, −7.3)
Arm 2	−6.3 (0.6)	F(1, 560.0) = 98.74, p < 0.0001	(−7.6, −5.1)
Arm 3	−6.9 (0.6)	F(1, 578.8) = 119.50, p < 0.0001	(−8.1, −5.7)
Arm 4	−8.2 (0.7)	F(1, 546.7) = 157.90, p < 0.0001	(−9.5, −6.9)
Arm 5	−6.3 (0.6)	F(1, 570.4) = 95.39, p < 0.0001	(−7.6, −5.0)
Arm 6	−7.2 (0.7)	F(1, 524.3) = 117.75, p < 0.0001	(−8.5, −5.9)

SE = standard error of the estimate; CI = confidence interval of the estimate.

Table uses imputed data.

Table 5. Mean difference (MD) in PROMIS Sleep Disturbance SF 8a item score from baseline to study conclusion, based on linear mixed effects regression model.

	Estimate	z	p value	95% CI
In the past week….
My sleep quality was ….
Arm 1	0.0	−1.27	0.205	(−0.1, 0.0)
Arm 2	0.0	1.69	0.092	(0.0, 0.1)
Arm 3	(reference)
Arm 4	0.0	−0.29	0.774	(−0.1, 0.0)
Arm 5	0.0	1.02	0.308	(0.0, 0.1)
Arm 6	0.0	1.06	0.291	(0.0, 0.1)
My sleep was refreshing^a
Arm 1	−0.1	−1.97	0.049	(−0.1, 0.0)
Arm 2	0.0	−0.41	0.681	(−0.1, 0.0)
Arm 3	(reference)
Arm 4	0.0	−1.37	0.169	(−0.1, 0.0)
Arm 5	0.0	−0.45	0.653	(−0.1, 0.0)
Arm 6	0.0	−0.47	0.635	(−0.1, 0.0)
I had a problem with my sleep.
Arm 1	−0.1	−2.83	0.005	(−0.2, 0.0)
Arm 2	0.0	0.05	0.960	(−0.1, 0.1)
Arm 3	(reference)
Arm 4	−0.1	−1.81	0.070	(−0.1, 0.0)
Arm 5	0.0	1.38	0.168	(0.0, 0.1)
Arm 6	0.0	−0.26	0.798	(−0.1, 0.1)
I had difficulty falling asleep.
Arm 1	0.0	0.00	0.999	(−0.1, 0.1)
Arm 2	0.0	1.20	0.229	(0.0, 0.1)
Arm 3	(reference)
Arm 4	0.0	0.21	0.831	(−0.1, 0.1)
Arm 5	0.0	0.91	0.365	(0.0, 0.1)
Arm 6	0.0	0.98	0.327	(0.0, 0.1)
My sleep was restless.
Arm 1	−0.1	−2.34	0.019	(−0.1, 0.0)
Arm 2	0.0	0.20	0.838	(−0.1, 0.1)
Arm 3	(reference)
Arm 4	0.0	−1.15	0.249	(−0.1, 0.0)
Arm 5	0.0	0.34	0.732	(−0.1, 0.1)
Arm 6	0.0	−0.67	0.505	(−0.1,0.0)
I tried hard to get to sleep.
Arm 1	0.0	−0.69	0.491	(−0.1, 0.0)
Arm 2	0.0	1.37	0.171	(0.0, 0.1)
Arm 3	(reference)
Arm 4	0.0	−0.51	0.608	(−0.1, 0.1)
Arm 5	0.0	1.39	0.163	(0.0, 0.1)
Arm 6	0.0	0.01	0.994	(−0.1, 0.1)
I was worried about not being able to fall asleep.
Arm 1	−0.1	−1.76	0.078	(−0.1, 0.0)
Arm 2	0.0	0.48	0.630	(0.0, 0.1)
Arm 3	(reference)
Arm 4	−0.1	−1.59	0.113	(−0.1, 0.0)
Arm 5	0.0	0.58	0.561	(0.0, 0.1)
Arm 6	0.0	−1.33	0.184	(−0.1, 0.0)
I was satisfied with my sleep.^a
Arm 1	0.0	−1.39	0.164	(−0.1, 0.0)
Arm 2	0.0	1.12	0.262	(0.0, 0.1)
Arm 3	(reference)
Arm 4	0.0	−1.14	0.253	(−0.1, 0.0)
Arm 5	0.0	0.27	0.790	(−0.1, 0.1)
Arm 6	0.0	0.69	0.491	(0.0, 0.1)

CI = confidence interval of the mean difference.

Table uses unimputed data; for all items, N = 1288 and the number of unique responses was 4287.

Linear mixed-effects regression model with random intercept at the individual level and random slope at week level.

^a Item responses were reverse coded so that higher scores represented greater sleep disturbance.

Table 6. Adjusted^a odds ratio (aOR) of achieving a minimal clinically important difference in PROMIS Sleep Disturbance 8a scores.

	aOR	95% CI
Study arm^b
Arm 1	1.4	(0.8, 2.4)
Arm 2	0.7	(0.4, 1.2)
Arm 3	(reference)
Arm 4	1.3	(0.8, 2.2)
Arm 5	0.8	(0.5, 1.4)
Arm 6	1.1	(0.7, 1.8)

CI = confidence interval of the estimate.

Multiple imputed logistic regression; F(14, 22175.9) = 1.02, p value = 0.4272.

^a Adjusted for sex, age, race, ethnicity and body mass index.

^b Joint-test of main effects significant; F(5, 9414.2) = 2.29, p value = 0.0433.

Availability of data and materials

Data are proprietary to Radicle Science and Open Book Extracts, and so are not publicly available. However, data are however available from the authors upon reasonable request and with permission of Open Book Extracts.