Low striatal dopamine receptor availability linked to caloric intake during abstinence from chronic methamphetamine abuse

Methamphetamine (MA) abuse, a prevalent problem with considerable public health and criminal justice costs,¹ resembles other addictive disorders in its association with deficits in striatal dopamine receptor availability.^2,3 This commonality is linked with the propensity to develop addictions via the Reward Deficiency Hypothesis⁴. Despite observations of low striatal dopamine receptor availability both in individuals with MA dependence⁵ and those who have obesity^3,6, empirical measures of eating dysregulation and propensity to obesity among abstinent MA-dependent individuals have not been reported. According to the Reward Deficiency Hypothesis, addicted individuals would seek out alternative rewards (e.g., food) in the absence of their preferred pathological rewarding outlet (e.g., MA abuse) to compensate for their deficient striatal dopamine signaling. To test the applicability of this hypothesis to MA dependence, we measured striatal dopamine D_2/3 receptor availability as well as caloric intake and weight gain in MA-dependent research participants during early abstinence from MA.

Twenty non-treatment-seeking MA-dependent research participants resided on a hospital inpatient research ward as previously described⁵, with approval from the UCLA Institutional Review Board. Aside from MA and nicotine dependence, the participants were otherwise healthy, and suffered from no other comorbid conditions. The participants were allowed to eat and smoke ad libitum during the study, with abstinence from alcohol and illicit substances confirmed by daily breathalyzer and urine drug screening. Complete daily calorie counts were assessed over days 3-7 and 28-32 of the inpatient stays. Participants were weighed on admission and weekly thereafter, in the morning. Positron emission tomography (PET) scanning with [¹⁸F]fallypride was performed, as previously described⁵, on days 4-8. Statistical comparisons across time-points were conducted using linear mixed effects modeling. As measures of D_2/3 receptor availability in different striatal regions were highly correlated (r's ≥ 0.7), they were averaged for regional correlations of receptor availability with caloric intake. Region of interest correlations were conducted using Spearman's rank order correlation. Striatal correlation maps (Figure 1) were prepared using non-parametric regression as previously described⁵.

Figure 1.

(a) Correlations of striatal dopamine D_2/D₃ receptor availability with eating behaviors in abstinent MA participants. Dots represent data from individual participants, with lines representing the slopes of the linear regressions. (b) Results from voxelwise regression of BPnd on calories/week (n=16). Whole-brain corrected threshold-free cluster estimation (TFCE) probability maps are overlaid on the mean spatially normalized anatomical image. Statistical maps are thresholded at TFCE-corrected p<0.05. Coronal and transverse slices are Y=10 and Z=0, respectively. Color bar indicates p-values. R=right.

Calorie counts during days 3-7 of abstinence were complete for 20 participants: 11 men (58%), 13 of whom were ethnically White (68%), average age of 35.6±8.3 years (mean±s.d.), mostly cigarette smokers (n=18, 90%). They had been using MA for 13.2±8.6 years on average, and 18.6±8.1 days of the last 30 before admission. On average, they consumed 4277±1121 Kcal/day over days 3-7 of hospitalization. Calorie counts were complete in both days 3-7 and 28-33 of the hospitalization for 7 participants; and participants consumed 3763±1073 Kcal/day over days 3-7, and 3588±621 Kcal/day over days 28-32, with no group difference between time-points (p=0.71).

Eight participants had BMI measurements through 3 weeks of hospitalization. The average admission BMI was 26.1±3.0 kg/m², with an increase to 28.2±3.3 kg/m² by the end of the third week (effect of time: p<0.001), an average increase of 8.0%.

Sixteen participants had both complete calorie counts for days 3-7 and [¹⁸F]fallypride PET scans while eight subjects had BMI data for days 28-32 and PET scans (Figure 1). Days 3-7 calorie counts were not correlated with scores on the Beck Depression Inventory, Three-Factor Eating Questionnaire, smoking status, sex, ethnicity, or measures of prior MA use. To control for age-related loss of striatal D_2/3 receptors, D_2/3 receptor availability was normalized to a common age, using a published estimate (4.9% loss/decade for nucleus accumbens and putamen, 5.5% loss/decade for caudate)⁷. Days 3-7 calorie counts were highly correlated with striatal D_2/3 receptor availability (r=-0.54, p=0.016, 1-tailed p-value; Figure 1). While third-week BMI increase did correlate with striatal D_2/3 receptor availability (r=-0.91) due to the small sample size, the results were not statistically robust (not shown).

The participants consumed more daily calories, on average, than hospitalized control subjects (e.g., ⁸), and gained a larger proportion (8%) of their BMI over 3 weeks than a group of psychiatric inpatients taking obesogenic antipsychotic medications for one month (only 3% BMI gain)⁹. While the findings are compelling, some limitations of the study include the small sample size, lack of a control sample, and lack of information about participants' dietary choices or smoking behavior. Nevertheless, taken together with the correlation between these measures of increased eating behavior and low striatal dopamine D_2/3 receptor availability (Figure 1), these observations provide support for the Reward Deficiency Hypothesis in MA dependence.