Fig. 3

Effects of niacin or nicotinamide adenine dinucleotide (NAD) metabolite treatment (TMT) versus control on the standardized mean difference (95%CIs) (SMD) in NAD metabolite (Panel A) and microbe (Panel B) measures in blood or tissue in studies (author, year of publication) conducted in either mouse, rat, or hamster models. Animal numbers (n) for study groups are shown. Data employed for analysis are shown in Supplemental Table 2 and 3. Data from individual experiments that were pooled within studies based on nonsignificant heterogeneity (p ≥ 0.05) comparing the experiments are shown in Supplemental Figs. 3 and 4. In one study for NAD metabolite (Cros) and two for microbe (Li and Yan) levels, results of more than one experiment could not be combined. Treatment had effects on the on the side of increasing SMDs in NAD⁺ levels in 14 of 15 studies or experiments (93%) and in 10, these increases were significant (i.e. the 95%CI was entirely on the side of increases with treatment). (Panel A). The magnitude of increases differed across species (p < 0.01) and heterogeneity within species with more than one study or experiment was significant for mice (n = 12, I² = 79%, p < 0.01) but not rats (n = 2, I² = 0%, p = 0.74). Treatment had effects on the side of decreasing SMDs for microbe levels in 20 of 22 studies or experiments (91%), and decreases were significant in 14, although there was significant primarily quantitative heterogeneity (I² = 82%, p < 0.01) (Panel B).