Safety Assessment of a Hemp Extract using Genotoxicity and Oral Repeat-Dose Toxicity Studies in Sprague-Dawley Rats

Dziwenka, Margitta; Coppock, Robert; Alexander, McCorkle; Palumbo, Eddie; Ramirez, Carlos; Lermer, Stephen

For hematology and clinical chemistry, the data from male and female rats were analyzed separately. Means and standard deviations were calculated for all quantitative clinical pathology parameters using Pristima® version 7 (Statistical Analysis, Xybion Corporation, Lawrenceville, NJ). These data were analyzed in a sequential manner. First, Bartlett's test for homogeneity and Shapiro-Wilk test for normality was done. If the Bartlett's test for homogeneity and Shapiro-Wilk test for normality were not significant, a one-way analysis of variance followed with Dunnett's test was performed. If the Bartlett's test for homogeneity and Shapiro-Wilk test for normality were significant then data transformations to achieve normality and variance homogeneity were done. The order of transformations attempted was log, square root, and rank-order. If the log and square root transformations fail, the rank-order was used. When an individual observation was recorded as being less than a certain value, e.g., below the lower limit of quantitation, calculations were performed on one-half of the recorded value. For example, if bilirubin was reported as < 0.1 or ≤0.1, then 0.05 was used for all calculations performed with that bilirubin data. When an individual observation was recorded as being greater than a certain value, e.g., above the upper limit of quantitation, then a greater value was used in place of the recorded value. For example, if specific gravity was reported as > 1.100 or ≥1.100, then 1.100 was used for all calculation performed using that specific gravity value. For all statistical testing, significance was a probability value of p < 0.05.