Chronic marijuana use moderates the correlations of serum cholesterol with systemic mitochondrial function and fluid cognition

Abstract

Activating type 1 cannabinoid (CB1) receptor decreases the particle size of high-density lipoprotein (HDL) and inhibits reverse cholesterol transport (RCT). This study examined whether marijuana (MJ) use is associated with changes of RCT, and how the latter is associated with mitochondrial function and fluid cognition. We recruited 19 chronic MJ users and 20 nonusers with matched age, BMI, sex, ethnicity, and education. We measured their fluid cognition, mitochondrial function (basal and max respiration, ATP production) in peripheral blood mononuclear cells, cholesterol content in serum lipoprotein fractions, enterolactone/creatinine ratio in urine as a marker for dietary polyphenol intake, and lipase activity in serum. We found that higher percentage of large HDL cholesterol (HDL-C) correlated positively, while that of small HDL-C correlated inversely, with mitochondrial function among MJ users, but correlations of the opposite directions were found among nonusers. The concentrations of large and intermediate HDL-C correlated positively with mitochondrial function and fluid cognition among MJ users, but not among nonusers. Both percentage and concentration of large HDL-C correlated positively, while those of small HDL-C correlated inversely, with amounts of daily and lifetime MJ use. In all participants, higher urinary enterolactone/creatinine ratio and lower serum lipase activity were associated with higher large HDL-C/small HDL-C ratio, implying greater RCT. This study suggests that high MJ use may compromise RCT, which is strongly associated with mitochondrial function and fluid cognition among MJ users.

Keywords: Cognition; Lipase; Marijuana; Mitochondria; Polyphenol; Reverse cholesterol transport.

Figure 1.. Correlations between marijuana (MJ) use and serum cholesterol distribution in high-density lipoprotein (HDL) fractions.

Log transformed daily MJ use inversely correlated with percentage of large HDL cholesterol (large HDL-C) (A), but positively correlated with the percentage (B) and concentration (C) of small HDL-C. Log transformed lifetime cumulative MJ use inversely correlated with percentage of large HDL-C (D), positively correlated with percentage of small HDL-C (E), and inversely correlated with the ratio of large HDL-C/ small HDL-C (F). Correlation coefficient and p value of Spearman correlation are shown.

Figure 2.. Marijuana (MJ) use status moderated the correlations between fractions of high-density lipoprotein cholesterol (HDL-C) and mitochondrial (mt) basal respiration rate in peripheral blood mononuclear cells.

MJ use status significantly moderated the relationships of mt basal respiration rate with (A) percentage of total HDL-C, (B) percentage of large HDL-C, (C) percentage of small HDL-C, (D) concentration of total HDL-C, (E) concentration of large HDL-C, and (F) concentration of intermediate HDL-C. The moderating effect of MJ use status was determined by two-way ANOVA, and post hoc Pearson correlation was used to evaluate the correlations between the variables in each study group. Bold fond indicates statistical significance.

Figure 3.. Marijuana (MJ) use status moderated the correlations between fractions of high-density lipoprotein cholesterol (HDL-C) and fluid cognition composite score.

MJ use status significantly moderated the relationships of fluid cognition composite score with (A) concentration of total HDL-C, (B) concentration of intermediate HDL-C, and (C) concentration of large HDL-C. The moderating effect of MJ use status was determined by two-way ANOVA, and post hoc Pearson correlation was used to evaluate the correlations between the variables in each study group. Bold fond indicates statistical significance.

Figure 4.. Ratio of Enterolactone/Creatinine (Ent/Cre) in random urine samples and its correlations with fractions of high-density lipoprotein cholesterol (HDL-C).

The Ent/Cre ratio decreased by 80% in marijuana (MJ) users compared with nonusers (A), individual data point and median of each group are shown. Regardless of MJ use status, the log transformed Ent/Cre ratio positively correlated with the percentage of large HDL-C (B), had a trend to inversely correlate with percentage of small HDL-C (C), and positively correlated with the ratio of large HDL-C/small HDL-C (D). U-test was used to evaluate the group difference of Ent/Cre ratio in Panel A, Pearson correlation was used to assess the correlations between variables in Panels B-D. Bold fond indicates statistical significance.

Figure 5.. Serum lipase activity and its correlations with urinary Enterolactone/Creatinine (Ent/Cre) ratio, fractions of high-, low-, and very low-density lipoprotein cholesterol (HDL-C, LDL-C, and VLDL-C, respectively).

Serum lipase activity inversely correlated with Ent/Cre ratio in the urine (A). Marijuana (MJ) use status was a (weak) moderator to the correlation between serum lipase activity and BMI (B). Regardless of MJ use status, serum lipase activity inversely correlated with percentages of total HDL-C (C) and large HDL-C (D), correlated with percentage of small HDL-C (E) and inversely correlated with the ratio of large HDL-C/small HDL-C (F). Serum lipase activity also inversely correlated with the concentrations of total (G), large (H) and intermediate (I) HDL-C, inversely correlated with the percentage of VLDL-C (J), and had a trend to inversely correlate with percentage of LDL-C (K). In Panel A, t-test was used to evaluate the group difference of serum lipase activity. In Panel B, the moderating effect of MJ use status was assessed by two-way ANOVA, followed by post hoc Pearson correlation. Correlations in Panels C-K were evaluated using Pearson correlation. Bold fond indicates statistical significance.

Figure 6.

Hypothetical mechanistic pathways that link marijuana use with reverse cholesterol transport (RCT), mitochondrial function, and fluid cognition.