An evaluation of the evidence that methamphetamine abuse causes cognitive decline in humans

Abstract

Methamphetamine (MA) is one of the most commonly abused illicit substances worldwide. Among other problems, abuse of the drug has been associated with reduced cognitive function across several domains. However, much of the literature has not attempted to differentiate cognitive difficulties caused by MA abuse from preexisting cognitive difficulties that are likely caused by other factors. Here, we address this question, evaluating evidence for a priori hypotheses pertaining to six lines of research: (a) animal studies; (b) cross-sectional human studies; (c) a twin study; (d) studies of changes in cognition with abstinence from MA; (e) studies of changes in brain structure and function with abstinence from MA; and (f) studies of the relationship between the severity of MA abuse and the extent of cognitive deficits observed. Overall the findings were mixed, with some support for a causal relationship between MA abuse and cognitive decline, and other findings suggesting that there is no relationship. The preponderance of the data, however, does support the possibility that MA abuse causes cognitive decline, of unknown duration, in at least some users of the drug. When averaged across individuals, this decline is likely to be mild in early-to-middle adulthood. However, moderator variables are likely to contribute to the presence and/or severity of cognitive decline exhibited by a given individual.

Figure 1

Changes in gray matter for methamphetamine (MA)-dependent individuals after 1 month of abstinence compared to control subjects re-tested after an 1 month interval. (a) Control subjects did not show any significant changes in gray matter with retesting. (b) MA-dependent participants show increases in gray matter with 1 month abstinence. Note: N=12 per group; statistical threshold: p<0.001 uncorrected, cluster extent >100 voxels.

Figure 2

Relationship between age of first use of methamphetamine (MA) and the difference between predicted vs actual years of education. Note: N=36 MA-dependent participants. Pearson correlation. Predicted education was predicted from cognitive battery scores and demographic characteristics (age, gender, ethnicity), using a regression model developed in healthy comparison subjects (N=42). Positive difference scores indicate that predicted education was greater than actual years of education, whereas negative difference scores indicate that predicted education was less than actual years of education. Results show that larger (positive) difference scores were associated with a younger onset of MA use. Reprinted from Dean et al (2012) with permission from Elsevier.