Abnormal Brain Structure Implicated in Stimulant Drug Addiction

See allHide authors and affiliations

Science  03 Feb 2012:
Vol. 335, Issue 6068, pp. 601-604
DOI: 10.1126/science.1214463


Addiction to drugs is a major contemporary public health issue, characterized by maladaptive behavior to obtain and consume an increasing amount of drugs at the expense of the individual’s health and social and personal life. We discovered abnormalities in fronto-striatal brain systems implicated in self-control in both stimulant-dependent individuals and their biological siblings who have no history of chronic drug abuse; these findings support the idea of an underlying neurocognitive endophenotype for stimulant drug addiction.

Drug dependence is increasingly recognized as a “relapsing brain disorder” (1) and, in support of this view, marked structural changes in striatal and prefrontal brain regions have been reported in people dependent on stimulant drugs (2). These reports, however, raise the question of whether these brain abnormalities may have predated drug-taking, rendering individuals vulnerable for the development of dependence.

Individuals at risk for drug dependence typically have deficits in self-control (3, 4), which may reflect a diminished ability to recruit prefrontal networks for regulating behavior (5). Stimulant drugs are highly reinforcing, because they directly affect brain systems implicated in motivated behavior, such as the basal ganglia and the limbic system (6), and they modulate control systems in the prefrontal cortex (7). Malfunction of these circuitries may increase the susceptibility for stimulant-induced neuroadaptive changes and facilitate the development of drug dependence.

As brain structure is, to a large extent, inherited (8) and drug dependence runs in families (9), a genetic or epigenetic influence on addictive behaviors seems plausible. Yet, we know very little about the mechanisms through which risks for drug dependence might be inherited. Endophenotypes are quantitative traits, mediating between the predisposing genes (genotypes) and the clinical symptoms (phenotypes) in complex disorders (10). As heritable traits, endophenotypes can be measured objectively in both patients and their unaffected first-degree relatives. We compared brain structure and the ability to regulate behavior in 50 biological sibling pairs; within each pair, one sibling satisfied the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) criteria for dependence on stimulant drugs and the other had no history of chronic drug or alcohol abuse (11). The sib-pairs were also compared with 50 unrelated healthy volunteers matched for age and intelligence quotient (table S1). Tobacco was smoked by members of all groups, but smoking rates were significantly higher in the sib-pairs than in the unrelated volunteers (table S1), which is not surprising for individuals with a greater-than-normal genetic risk of drug dependence.

We used the stop-signal task (12), one of the most widely used measures of inhibitory control, which requires individuals to rapidly suppress an ongoing, well-established response whenever an auditory signal is suddenly presented. The stop-signal reaction time (SSRT) estimates the time that an individual needs to withhold an ongoing response and can predict the onset of substance abuse in vulnerable individuals (4). The neural circuitry underlying stop-signal task performance has been well-characterized and there is ample evidence for the association of SSRT with both the functional and structural integrity of brain systems known to be compromised in stimulant drug dependence (13, 14). We observed marked impairments in the regulation of behavior in both drug-dependent individuals and their biological siblings who have no history of chronic drug abuse (Fig. 1A). Indeed, the deficits in SSRT were as pronounced in the nondependent siblings as in the stimulant-dependent patients. Moreover, the variance in SSRT within sib-pairs was significantly smaller than in unrelated sib-pairs (permutation test, P = 0.033), which suggests that poor inhibitory control is a familial trait in vulnerable individuals and not a result of long-term drug abuse.

Fig. 1

Deficits of motor inhibitory control and white matter organization in stimulant-dependent individuals and their nondependent siblings. (A) SSRT differed significantly between the three groups (F2,141 = 9.9, P < 0.001). SSRT was significantly prolonged in both the stimulant-dependent individuals and their siblings compared with unrelated healthy volunteers (Bonferroni correction, P ≤ 0.005, for both comparisons). (B) The skeleton of group differences in mean FA is colored in blue (F2,141= 26.3, P < 0.001); on the basis of previous studies, regions of interest were selected within the blue skeleton, which included the IFG and the presupplementary motor area (colored in orange). (C) Scatterplot showing that participants with greater mean FA in the right IFG had better inhibitory performance (shorter SSRT) on the stop-signal task (the correlation coefficient, r = 0.24; the linear correlation constant, R2 = 0.057, P < 0.005).

To investigate the relation between inhibitory control and brain structure, we calculated from each individual’s diffusion tensor images the fractional anisotropy (FA) values that serve as a general index of the integrity of white matter fiber tracts (15) and analyzed them using tract-based spatial statistics (TBSS) implemented in FSL software (11). We compared the mean FA values between the groups within a tract-based skeleton (16) and found evidence for a significant reduction of FA in the sib-pairs compared with healthy unrelated volunteers (indicated in blue in Fig. 1B). Again, the variance of FA within biological sib-pairs was significantly smaller compared with the variance within unrelated sib-pairs (permutation test, P = 0.004), which suggested that the observed white matter abnormalities were shared among members of the same family and may have predisposed them to drug-taking. In the stimulant-dependent individuals, reduced FA was associated with the duration of stimulant abuse (fig. S2A), which suggests that white matter changes can also result from drug-taking (although these effects of drug exposure were less anatomically extensive than the effects of familial risk for dependence).

Previous research shows that stop-signal task performance is subserved by a neural network, including the inferior frontal gyrus (IFG), the basal ganglia (caudate-putamen), and the presupplementary motor area (pre-SMA) (17, 18). On the basis of previous studies, we created a region of interest using Hammer’s probabilistic atlas (19), as indicated in orange in Fig. 1B. Within this region of interest, we regressed each participant’s SSRT score on their white matter FA. Reduced FA in fiber tracts adjacent to the right IFG was significantly associated with poorer inhibitory control and accounted for ~6% of the variance in SSRT (Fig. 1C). The variability of the abnormality in the right prefrontal white matter was significantly similar within biological siblings compared with unrelated sib-pairs (permutation test, P = 0.004). Together, these results support the idea that self-control deficits are subserved by white matter disorganization in the right prefrontal lobe; this finding provides an objective vulnerability marker for an increased, possibly inherited, risk for developing stimulant drug dependence.

We also compared the gray matter volume maps of healthy volunteers with those of the drug-dependent individuals (Fig. 2A) and of the non–drug abusing siblings (Fig. 2B) using voxel-based morphometry analysis implemented in FSL software (11). The brains of the sib-pairs showed distinct abnormalities relative to the healthy control volunteers. Specifically, key structures previously implicated in addiction, such as the medial temporal lobe (20) and the basal ganglia (21, 22), were significantly enlarged in the sib-pairs. We also identified a significant reduction of gray matter volume in the posterior postcentral gyrus and adjacent areas, such as the superior temporal gyrus and the posterior insula, in both drug-dependent individuals and their siblings compared with healthy volunteers (Fig. 2C). The within-pair variability of the gray matter volume increase in the putamen (permutation test, P = 0.013) and of the volume decrease in the posterior insula (permutation test, P = 0.012) were both significantly smaller in biological siblings than in randomly paired siblings, which indicated that this abnormality is shared between members of the same family. By contrast, the enlargement of the amygdala in the sib-pairs did not survive the test of familiality (permutation test, P = 0.144), which suggested that other, nonfamilial factors may account for this abnormality. Note that higher tobacco exposure in the nondependent siblings is unlikely to explain the pattern of gray matter abnormalities identified as an endophenotype for stimulant dependence, as shown by similar results from a separate analysis restricted to nonsmokers (fig. S1). Moreover, gray matter regions associated with the duration of stimulant drug exposure differed clearly in location from the regions identified as markers of familial risk for stimulant drug dependence (fig. S2B)

Fig. 2

Similar abnormalities in gray matter brain volume were found in stimulant-dependent individuals (A) and in their nondependent siblings (B) when separately compared with healthy volunteers. Brain areas of abnormally increased gray matter (relative to healthy volunteers) are colored in red, and brain areas of abnormally decreased gray matter are colored in blue. (C) Three clusters of shared abnormality were identified in both the stimulant-dependent individuals and their siblings when compared with unrelated healthy volunteers: left amygdala and left putamen were both significantly enlarged; whereas the left postcentral gyrus, superior temporal gyrus, and posterior insula were significantly reduced in gray matter volume. (Left side of the brain is shown on the left side of each slice; the numbers denote Z coordinates for each slice in standard stereotactic space.)

The involvement of the putamen is consistent with its implication in fronto-striatal circuits for stop-signal performance and proposed antecedent problems in response control (23). However, the additional regions identified as showing changes in the sib-pairs may be related to other psychological processes underlying addiction. Thus, brain abnormalities observed in the sib-pairs in neural systems underlying learning and memory [such as the medial temporal lobe (24)], and habit formation [such as the putamen (25)] are intriguing, given that some forms of drug addiction are thought to develop through maladaptive acquisition and control of habits (26). Enlargement of limbic and striatal structures has been reported previously in patients with obsessive-compulsive disorder (OCD) (27), and like addiction, OCD is characterized by dysfunctional habits and “out-of-control” behavior. Our findings may indicate markers of neural vulnerability for pathological habit formation, which could further facilitate the effects of drugs of abuse by interfering with limbic-striatal functions. Pathological habits in drug addiction typically result in compulsive drug-taking behaviors when prefrontal control fails to regulate behavior (26). Our data are also in keeping with preclinical research indicating that impairments in response control are predictive of cocaine reinforcement and dopamine receptor dysfunction in the striatum (28). Deficits in inhibitory prefrontal control were evident in both drug-dependent individuals and their siblings who do not abuse drugs, which may reflect an increased risk for out-of-control drug-seeking or drug-taking behaviors, which could pave the way for the development of drug dependence.

The identified profile of familial abnormalities remarkably resembles the developmental changes of brain structure during adolescence, i.e., limbic-striatal structures mature before prefrontal brain systems. This developmental asynchrony has been suggested to create an imbalance between mesolimbic reward and prefrontal control systems, which predisposes adolescents to sensation-seeking and impulsive behavior, rendering them potentially vulnerable to drug-taking (29). Our previous data on biological siblings of stimulant-dependent individuals indicated a propensity for increased impulsivity, as measured on the Barratt Impulsivity Scale, contrasting with normal scores on measures of sensation-seeking traits (30). The present findings show that even stronger effects in the sib-pairs are observed with an objective measure of impulse control, the SSRT. These findings are also related to changes in brain structure, including the inferior frontal cortex and putamen, which are key nodes in a neural network that mediates response regulation.

Our findings thus indicate that gray matter changes in the dorsal striatum, together with abnormal inferior prefrontal cortical connectivity, underlie an increased risk for developing stimulant drug dependence. However, the almost equivalent impairments in SSRT in both the stimulant-dependent individuals and their unaffected siblings need careful interpretation, as they do not reflect the classic pattern for endophenotypes, i.e., that the first-degree relatives have trait values intermediate between the patients and the unrelated healthy volunteers (10). Presumably, the siblings must have some other resilience factors that counteract the familial vulnerability to drug dependence. The identification of these brain and behavioral biomarkers for familial risk of drug dependence demonstrates that an individual’s predisposition to become addicted to stimulant drugs may be mediated by brain abnormalities linked to impaired self-control.

Supporting Online Material

Materials and Methods

Figs. S1 to S2

Tables S1 to S3

References (3142)

References and Notes

  1. Materials and methods are available as supporting material on Science Online
  2. Acknowledgments: The authors thank all volunteers for their participation in this study and the clinical study officers at the Mental Health Research Network for their dedicated support with volunteer recruitment. The authors are grateful to the staff at the Wellcome Trust Clinical Research Facility at Addenbrooke’s Hospital and the Wolfson Brain Imaging Centre for their help and support throughout the study. This work was funded by the U.K. Medical Research Council (MRC) and conducted within the Behavioral and Clinical Neuroscience Institute at the University of Cambridge, which is jointly funded by the MRC and the Wellcome Trust. K.D.E., P.S.J., and A.J.T. are supported by a research grant of the MRC. T.W.R. and A.J.T. consult for Cambridge Cognition. E.T.B. is employed 50% by GlaxoSmithKline and 50% by the University of Cambridge. K.D.E., P.S.J., and G.B.W. declare that they have no conflicts of interest.
View Abstract

Stay Connected to Science

Navigate This Article