Previously, reduced perfusion in the prefrontal cortex has been observed in both clinically depressed (Drevets, 2000
) and opiate-dependent individuals (Danos et al., 1998
; Rose et al., 1996
). In this study, we examined the relationship between baseline rCBF and depressive symptoms in MM patients, who met or did not meet the diagnosis of major depression. We found that reduced baseline rCBF in the bilateral VLPFC and middle frontal regions were linked to higher depression scores on the BDI, a continuous measure of depression symptoms. These findings suggest that frontal paralimbic dysfunction may be linked to depressive symptoms in patients with and without opiate dependence.
Because opiate-dependent patients report fewer depressive symptoms over the course of methadone treatment (Hesse, 2006
), our patient sample, whose methadone treatment averaged 2.25 years, had a low, narrow range of depressive symptoms and did not meet the BDI score criterion for major depression of 21 as suggested by Geisser et al. (1997)
. Despite this limited range, the inverse relationship between rCBF and depression scores was robust.
In a sample of abstinent opiate abusers, Gerra et al (1998)
showed a significant negative correlation between depression symptoms and perfusion in the left temporal lobe, and near significant negative correlation in the frontal lobe. Our voxel-based fMRI approach, which provides more localized results, demonstrates the inverse relationship between frontal perfusion and depression symptoms. Interestingly, perfusion in the temporal lobe region was not inversely associated with depression symptoms in our study. This lack of consistency between our results and Gerra et al’s (1998)
may be due to the differences in participant samples. First, Gerra et al.’s (1998)
patients were former opiate-dependent patients who had been abstinent for 4 months. Additionally, using a sample of patients who entered a voluntary therapeutic community and did not seek methadone-maintenance treatment may help explain the differences in results. Alternatively, the relationship between perfusion in the temporal lobe and depression symptoms is demonstrably different in men and women (Videbech et al., 2002
), suggesting that the differences in sex distribution of the samples may help explain the discrepancy between the results.
Our findings are in agreement with two previous investigations that reported a negative correlation between depression symptoms as measured by BDI, and frontal brain metabolism (Dunn et al., 2002
) or rCBF (Bench et al., 1993
) in clinically depressed patients. However, two studies using Hamilton Rating Scale for Depression scores (HRSD) reported no association (Milak et al., 2005
) or a positive association (Graff-Guerrero et al., 2004
). The inconsistent results may reflect that BDI and HRSD, in part, assess different dimensions of depression; the BDI captures cognitive and affective dimensions whereas the HRSD captures somatic and behavioral dimensions (Brown et al., 1995
). Using cluster factors offered by Brown et al (Brown et al., 1995
), 70% of the total BDI scores from our sample derive from items that measure negative affect and cognition. Given that different depression subtypes show unique brain perfusion patterns (Fountoulakis et al., 2004
), our results may be most applicable to patients exhibiting a specific subset of depression symptoms. For example, in the present study, we used the BDI, a well-validated, stand-alone assessment tool with excellent psychometric properties, to measure severity of depression symptoms. An informal analysis between perfusion and SCL-90-R Depression subscale scores, which measure nonspecific depressive symptoms (Bonynge, 1993
; Clark et al., 1983
; Rauter et al., 1998
) and accounted for only 41% of the variance with our BDI scores, yielded non-significant findings. Future neuroimaging investigations on depression should consider using a specific depression scale relevant to a depression subtype of interest.
In addition to the frontal resting perfusion findings from the aforementioned studies, the VLPFC and medial frontal regions have been observed to be crucial in emotional regulation (Grimm et al., 2006
; Ochsner et al., 2002
; Ochsner et al., 2004
). For example, performing a task to regulate emotions was associated with increased activities in the VLPFC and middle frontal regions in healthy subjects. Given that depressed patients are poor affect regulators (Forbes et al., 2005
; Garber et al., 1995
), our results suggest that a similar relationship between poor affect regulation and frontal dysfunction is also present in our MM patients.
Though not a priori ROI in our study, perfusion in the left inferior parietal lobe was also negatively correlated with BDI scores. Two studies have reported reduced perfusion (Galynker et al., 1998
) or glucose metabolism (Biver et al., 1994
) in the parietal regions in depressed patients compared to normal controls, although its correlational relationship with depressive symptoms was not significant. This region has been associated with opiate withdrawal (Rose et al., 1996
) and levomethadone dosage (Danos et al., 1998
) in MM patients. Based on our results, it is possible the previous findings from Rose et al. (1996)
and Danos et al. (1998)
might be a result of withdrawal symptoms secondary to depressive symptoms. A further examination of the region’s role in depressive symptoms for MM patients is warranted.
Resting rCBF in the anterior cingulate, an a priori ROI, was not significantly related to the depressive symptoms in our MM sample. Previous studies have reported abnormalities in the anterior cingulate regions in opiate-dependent patients (Forman et al., 2004
; Lee et al., 2005
; Yucel et al., 2007
). In particular, Yucel et al. (2007)
recently demonstrated that opiate-dependent patients had significantly lower concentrations of N-acetylaspartate (NAA), a marker of neuronal integrity, in the anterior cingulate when compared to healthy controls. Galynker et al. (2007)
previously found a significant association between cerebral glucose metabolism in the left perigenual anterior cingulate and dysthymic scores, although the participants engaged in an attention task. Given the existing evidence for the association between depressive symptoms and brain dysfunction in this region, a further examination of the anterior cingulate’s role in depressive symptoms for MM patients is warranted.
Our study has several limitations. First, our results do not explain whether dysfunction in the fronto-limbic systems and depressive symptoms contribute to the initiation/maintenance (e.g., relapse) of substance addiction, as compared to being a consequence of chronic drug use. One challenging task would be to investigate the relationship between depression (i.e., affect regulation) and addiction prospectively, with particular emphasis on neural mechanisms. Second, we used the BDI scores to assess the relationship between affect regulation and addiction. This relationship should be examined with a behavioral probe that can more directly measure the construct of affect regulation. Third, due to lack of comparable data from a matched cohort, we were not able to compare our present data to a healthy control group. Galynker et al. (2007)
recently reported that self-reported dysthymic scores were negatively correlated with cerebral glucose metabolism levels in the right perigenual anterior cingulate in controls. Comparing our present data to a healthy control group would be helpful in understanding the relationship between brain perfusion and depressive symptoms in MM patients, especially the unique ways in which MM patients process negative affect. Furthermore, the present study explicitly attempted to measure individual variations in mental or emotional activity during resting state and their functional relationship to brain perfusion at rest, and to relate them to the self-reported, depressive symptoms. Because our research depends on the natural, spontaneous variations in mental or emotional states during the resting condition, no explicit attempt was made to control for internal states during resting scans.
Lastly, future research should examine the relationship between the rCBF and gray matter concentration in the PFC regions. Lyoo et al. (2006)
reported that opiate-dependent patients have decreased gray matter density in the PFC. Further, Drevets et al. (1997)
found that decreased activity in the anterior cingulate in depressed individuals is partially due to a grey matter volume reduction in the PFC area. It would be important to explore this particular relationship in MM patients, to determine whether the observed rCBF abnormalities reflect functional and/or structural differences, and extend our current understanding of the relationship between the PFC and depression symptoms in opiate addiction.
In conclusion, our findings showed an inverse relationship between rCBF in the bilateral VLPFC and middle frontal regions, and depressive symptoms, suggesting that frontal paralimbic dysfunction contributes to depressive symptoms in MM patients with or without a diagnosis of major depression. A significant subgroup of MM patients has clinical and sub-clinical depression; our data identify brain substrates underlying these symptoms. Additionally, affect dysregulation is an important risk factor for addiction (Tarter et al., 1995
; Tarter et al., 1999
). Future investigations of affect dysregulation in addiction may enhance our understanding of the substance dependence and depression comorbidity. Finally, depressive mood and symptoms are significantly correlated with escalations in drug craving, opiate use and treatment attrition (Childress et al., 1994
; Hasin et al., 2002
; Havard et al., 2006
; Hesse, 2006
; Kosten et al., 2004
; Nunes et al., 2004
; Zilberman et al., 2007
). Treatment strategies, such as psychopharmacological or psychotherapeutic interventions, targeting these brain regions may improve the depression symptoms in addicted individuals.