Journal Logo

Original Research

Possible Evidence for Re-regulation of HPA Axis and Brain Reward Systems Over Time in Treatment in Prescription Opioid-Dependent Patients

Bunce, Scott C. PhD; Harris, Jonathan D. BS; Bixler, Edward O. PhD; Taylor, Megan BA; Muelly, Emilie MD, PhD; Deneke, Erin PhD; Thompson, Kenneth W. MD; Meyer, Roger E. MD

Author Information
doi: 10.1097/ADM.0000000000000087
  • Open


In the early 1960s, researchers at the US Public Health Service's Addiction Research Center described a state of heightened response to drug cues, anhedonia, and physiological dysregulation in opiate-dependent patients that persisted for some time after withdrawal (Martin et al., 1963, 1973). They posited that this persisting state of dysregulation was a key factor in vulnerability to relapse among abstinent opioid patients, a theory that was consistent with the observation of Hunt et al. (1971) that the peak period for risk of relapse was in the first 3 months of abstinence. More recently, systemic dysregulation related to opiate withdrawal has been investigated by behavioral neuroscientists in animal models of addiction. Koob and Volkow (2010) postulated that persistent changes in the reward and memory circuits, thought to be new homeostatic set-points brought about by developing dependence on opiates (termed allostasis), underlie the sensitivity to drug-related cues and diminished appetitive responses to natural rewards (subjective anhedonia) in the early stages of recovery. In animal models, this postwithdrawal state is characterized by increased reward thresholds in drug-free rats after a period of opiate addiction (Kornetsky and Bain, 1990). In humans, the postwithdrawal state is characterized by anhedonia, and the relative devaluation of nondrug rewards (Luo et al., 2011) among addicted patients in the early months of recovery (Meyer, 1986). In their recent meta-analysis of functional neuroimaging studies, Goldstein and Volkow (2011) emphasized that, in humans, in addition to the dorsal striatal and mesolimbic areas, the prefrontal cortex plays a key role in the addiction and recovery process “through its regulation of limbic reward regions, and its involvement in higher-order executive function (ie, self-control, salience attribution and awareness).”

In addition to the central nervous system (CNS) factors, Koob and Kreek's (2007) summary of the literature on dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis associated with opiate dependence (Kreek et al., 1983, 1984) highlights the enhanced sensitivity to stressors found during the early recovery period. Salivary cortisol concentrations are elevated during withdrawal from opiate dependence and are correlated with withdrawal symptom severity, indicative of dysregulated HPA axis involvement in vulnerability to relapse (Bearn et al., 2001). Persistent sleep abnormalities are also a well-documented risk factor for relapse in patients with substance dependence disorders (Brower et al., 1998; Foster and Peters, 1999; Brower and Perron, 2010; Morgan et al., 2010). Sleep duration is inversely related to cortisol levels, making it a sensitive marker of cortisol rhythm and HPA axis function. In summary, heightened CNS sensitivity to drug-related cues (Wikler et al., 1963; Zhao et al., 2012), reduced response to natural rewards, and increased sensitivity to stress mediated by the HPA axis (Kreek, 2000; Stewart, 2003) are generally considered to be proximal determinants of relapse in recovering patients.

Despite the well-documented dysregulation of CNS and HPA axis systems in opioid addiction, neither the recent animal literature on allostasis nor the patient-oriented clinical studies has addressed the question of re-regulation or reversibility of the new set points of the CNS and HPA axis systems with abstinence over time (Kreek et al., 1984; Markou et al., 1993; Grigson et al., 2000; Stewart, 2003; Koob and Kreek, 2007; Grigson, 2008). Experienced clinicians understand “recovery” as a process that requires sustained abstinence, but the field lacks objective indices of “recovery” that mirror the neuroscientific literature in a clinically relevant and applicable context. This study represents an initial effort to define objective indices of “recovery” that can be assessed in clinical settings and that might later be examined in terms of their predictive validity.

Studies that have evaluated HPA axis and CNS function among patients in methadone maintenance programs highlight the need for a multidimensional approach to characterize recovery from opioid addiction. Kreek et al. (1984) observed that high-dose methadone maintenance normalized patient responses to a metyrapone challenge of HPA axis function. They postulated that the normalization might account for successful treatment outcomes. In contrast, in a functional magnetic resonance imaging (fMRI) study of methadone maintenance patients, Langleben et al. (2008) reported heightened CNS responses to heroin-related stimuli at the end of an interdose interval in the orbitofrontal cortex, insulae, and the left hippocampal complex, suggesting continued vulnerability to relapse. Lubman et al. (2009) reported that, despite the stable doses of agonist maintenance, some patients manifested reduced responsiveness to stimuli of natural rewards across a range of psychophysiological measures. These patients were more likely to use heroin over the next 6 months relative to those who showed more normative responses to appetitive stimuli. Together, these studies suggest that dysregulated CNS and HPA axis systems may pose relatively independent sources of vulnerability to relapse that re-regulate at different time intervals within individuals recovering from opioid dependence.

Amidst a rapidly escalating epidemic of prescription opioid-dependence in the United States (Substance Abuse and Mental Health Services Administration, 2008), most of the clinical research has been focused on studies of the patients dependent on heroin. There is a need to study the patients dependent on prescription opioids to identify and assess physiological responses as potential mediators and/or moderators of treatment efficacy in this population. Characterizing patients who differ in their duration of supervised abstinence is an important step in developing evidence-based treatments for prescription-opioid dependence. Whereas fMRI, PET, and endocrine challenge strategies offer highly granular data in relatively small patient samples, there is a need to apply the insights of the animal model and the patient-centered translational research in real-world clinical settings. One important element of this study was the utilization of functional near-infrared spectroscopy (fNIRs) to image dorsolateral/ventrolateral prefrontal cortex in response to drug-related, neutral, and natural rewarding images in a cue reactivity task. In an analogous application of the methodology, Bunce et al. (2012) demonstrated differences in the dorsolateral/frontopolar prefrontal cortex (DLPFC) reactivity to alcohol and natural reward cues in actively drinking alcohol-dependent individuals relative to both social drinkers and previously alcohol-dependent individuals who had been abstinent for 90 to 120 days. The literature is currently divided on the question of the DLPFC response to natural reward images in recently detoxified drug and alcohol-dependent patients (Heinz et al., 2007; Zijlstra et al., 2009). To help determine positive neural responses to naturally rewarding stimuli, we utilized a well-documented measure of emotional evaluation, the affect-modulated acoustic startle response (Lang, 1995; Bradley et al., 1999). In brief, the defensive blink in response to an abrupt acoustic stimulus (a negative stimulus) is amplified when the participant is processing a negative stimulus, such as an aversive picture, and attenuated when processing an appetitive stimulus, providing an objective measure of stimulus evaluation.

In this study, 3 groups of participants were compared: prescription opioid-dependent (POD) patients who were 7 to 10 days after medically assisted withdrawal (recently withdrawn; RW); POD patients who had been in supervised residence for 2 to 3 months (extended care; EC), and nonaddicted healthy control (HC) participants. On the basis of the literature and the specific measures used in this study, we hypothesized that, relative to the non–drug-dependent HC participants, the RW patients were expected to show (1) increased activation in the DLPFC in response to drug-related cues; (2) differential activity in the DLPFC in response to stimuli depicting natural rewards; (3) a neutral or augmented affect-modulated startle response to natural reward cues, indicating anhedonia or a negative hedonic response; and (4) abnormal diurnal cortisol rhythm and abnormal objective measures of sleep. The responses of EC patients who had been opioid free under supervision for 60 to 90 days were expected to fall in between those of the RW patients and the HC participants. Finally, we postulated that time away from opiates (days since last use) would be correlated with more normalized responses on each of these measures.



Patient samples (all right-handed) were recruited at the Caron Treatment Center, a residential drug and alcohol treatment facility in Wernersville, Pennsylvania. The Center provides drug rehabilitation and extended residential care for up to 3 months. The RW-POD participants had completed medically assisted withdrawal at Caron within the previous 7 to 10 days (Table 1). The EC POD research participants were patients in residence at the Center and abstinent for 58 to 95 days (mean = 79.3 ± 14.3 days). To allow measurement of the control sample using the same laboratory set-up as the enrolled patient research participants, nonaddicted HC participants were recruited from among the facility staff.

Participant Demographic Characteristics, Comorbid Disorders, and Medications for Patients Addicted to Prescription Opiates and Healthy Controls

Patient inclusion criteria were as follows: (1) capable of understanding and complying with protocol and giving informed consent; (2) age 18 years or older; (3) current DSM-IV-TR (Diagnostic and Statistical Manual of Mental Disorders–Fourth Edition–Text Revision) diagnosis of prescription opioid dependence, on the basis of clinical diagnosis at the Center, Structured Clinical Interview for DSM-IV-TR for axis I (SCID-I/P; First et al., 2002), and Form-90D (Westerberg et al., 1998). Exclusion criteria included (1) history of psychosis, bipolar disorder, or any psychiatric disorder that would compromise the patient's ability to complete the study; (2) current dependence (past year) on any drugs (or alcohol) other than prescription opiates, heroin, or nicotine; and (3) currently on an opiate agonist (eg, methadone or buprenorphine) or antagonist (naltrexone). HC participants had no current axis I disorder and no history of major mental illness or history of drug or alcohol abuse or dependence.

Data Collection

Clinical Assessment Instruments

Diagnoses were assigned using the SCID. Daily drug use for 180 days before intake was evaluated using the Form-90D.

fNIRs Cued Reactivity Task

Functional near-infrared spectroscopy uses near infrared light to monitor changes in the concentration of oxyhemoglobin (HbO2) and deoxyhemoglobin (HbR) as an indirect measure of regional brain activity (Villringer and Chance, 1997). Functional near-infrared spectroscopy offers several advantages for neuroimaging in a clinical setting; that is, it is highly portable, boasts rapid application times (5–10 minutes), and near zero run-time costs, enabling affordable, on-site neuroimaging that does not interfere with the patients' treatment regimen. Region of interest (ROI) optodes were chosen on the basis of previous results (Bunce et al., 2012). Using a standard cue reactivity paradigm, 125 slides were presented to participants while fNIRs was used to monitor frontopolar, dorsolateral and ventrolateral prefrontal cortical responses to the stimuli. Stimuli consisted of 25 color slides in 5 categories: prescription opioid pills (eg, close-ups of pills, labels, people taking pills), pill controls (eg, ball bearings, band aids), and positive hedonic stimuli resembling natural rewards (palatable foods, puppies, people having fun). Natural reward stimuli were drawn from the International Affective Picture System (Lang et al., 2008). Pill and pill-control stimuli were created for this study by the principal investigator. Stimulus presentation consisted of a pseudo-random block design composed of 5 epochs. Epochs contained five 25-second blocks composed of 5 pictures, each displayed for 5 s. A cross-hair was displayed for 10 s between blocks. The order of individual pictures within the blocks was randomized across participants. Before and after the cued-response study, participants rated their craving (desire to use opiates), desire to refrain from using, and sense of control over using opiates on 100-point visual analog scales.

Affect-Modulated Acoustic Startle Response Task

Because cued-response activation of the prefrontal cortex cannot be used to infer whether a response is appetitive or aversive per se, we employed an affect-modulated acoustic startle response (Lang, 1995; Bradley et al., 1999) to assess the emotional valence of reward-related stimuli. Participants viewed 12 pictures each of emotionally positive, neutral, and negative stimuli drawn from the International Affective Picture System. No stimuli were repeated from the fNIRs' cue reactivity task. The acoustic startle probe, a 50-ms burst of 104 dB white noise with instantaneous rise time, was presented at variable points during the 6-second slide viewing period, ranging from 3.5 to 5.5 s after slide onset. To minimize predictability, the probe was presented on 9 of the 12 slides for each type, and 4 startle probes were presented in the intervals between picture presentations.

Startle probes were presented binaurally through stereo headphones with presentation and timing of stimuli controlled by the STIM (Neuroscan, Inc) data presentation system. The eye-blink component of the startle reflex was measured by recording electromyographic (EMG) activity from 4-mm Beckman miniature Ag/AgCl electrodes positioned over the orbicularis oculi muscle beneath the left eye. Startle responses were standardized within participants; Z scores were converted to T-scores [50 + (Z × 10)] for analysis and plotting. One EC participant did not complete the startle task because of an allergy to the electrode adhesive.

HPA Axis Function and Sleep

Diurnal Cortisol

HPA axis function was assessed noninvasively utilizing salivary measures of diurnal cortisol. Saliva was collected (via chewed cotton swabs) for cortisol measurements at 5 time points for 2 consecutive days, that is, 8:00 AM before breakfast, 12:00 PM before lunch, 3:00 PM, 6:00 PM, and 9:00 PM. Cortisol levels were determined using a commercially available, high-sensitivity salivary cortisol enzyme immunoassay kit (Salimetrics, State College, Pennsylvania) designed to standardize the measurement of salivary cortisol across research and biomedical laboratories.

Sleep Data

Sleep data were collected using sleep wrist actigraphy readings (GT3x+ Actigraph, Pensacola, Florida) for 8 consecutive nights. The 2 patient samples slept in a structured residential environment, whereas the HC slept in a setting according to their individual preferences.


All the participants signed an informed consent statement approved by the Penn State College of Medicine Institutional Review Board. Study tasks were completed in the following manner: Day 1, cue reactivity and startle response tasks were completed; days 1 to 8, sleep actigraphy was collected; days 5 to 6 (of 8-day period), cortisol was collected 5 times daily. All data were collected during rest periods in the patient's scheduled treatment program.

Statistical Analysis

Analyses for fNIRs, startle, cortisol, and sleep data consisted of a general linear model ANCOVAs across the 3 groups, controlling for age and sex. Post hoc analyses were conducted using average least-squares differences. fNIRs analyses were conducted on oxygenation changes from baseline in response to pill images minus neutral images, or natural reward images minus neutral images. Initial ROI analyses (optodes 11, 12, 13, 14, over right DLPFC, based on Bunce et al., 2012; Fig. 1) were conducted on each optode and then averaged across contiguous significant optodes. Hypotheses for the startle response data were tested using T-scores for Natural Reward stimuli minus Neutral stimuli. All statistical analyses were conducted with SPSS 19.0.0 (IBM SPSS Statistics).

(A) Functional near-infrared spectroscopy region of interest; right dorsolateral prefrontal cortex (DLPFC). Colored area represents optodes differentiating the recent withdrawal from the extended care groups. (B) Mean change in oxygenated hemoglobin in DLPFC (optodes 11, 12, 13) to pill cues minus control images. Oxy, oxygenated; μM, microMolar. Error bars represent ± 1 SEM. (C) The Pearsonr correlation between change in the DLPFC oxygenated hemoglobin to pill cues versus the number of days since the last drug use.


Participant Demographics

Table 1 describes the demographics, psychiatric comorbidity, addiction history, and concurrent medication use of the 3 participant groups. All opioid-dependent patients were addicted to prescription narcotics; 3/7 RW patients and 2/7 EC patients also had a history of intravenous heroin use.

Response to Drug-Related Stimuli

Self-ratings of craving increased in both groups of opioid-dependent research participants (RW and EC) after exposure to the drug slides (F[1,12] = 32.27, P < 0.001), providing evidence that the prescription opiate stimuli were eliciting craving among the patient groups. An ANCOVA conducted on fNIRs data, assessing the amplitude of oxygenation change in response to pill images minus pill control images, revealed a main effect for group (F[2,20] = 3.49, P = 0.05, ηP2 = 0.30; Fig. 1Aand 1B). Planned contrasts revealed that, as predicted, the RW patients showed greater activation to pill stimuli than the EC (P = 0.02; Cohen's d = 1.20); however, RW did not differ from the HC sample (P = 0.18; d = 0.56). The EC responses to pill cues did not differ from the HC (P = 0.32; d = 0.73). Inspection of the means showed that HC scores fell between those of the 2 patient groups. Optodes outside the ROI did not attain conventional levels of significance (all P > 0.10).

The fNIRs measures of increased oxygenation in response to pill images showed a marginal correlation with the number of days since last drug use (Fig. 1C).

Response to Stimuli Depicting Natural Rewards

For stimuli depicting natural rewards, an ROI analysis (Fig. 2A) revealed a significant main effect for group (F[2,16] = 6.61, P = 0.008, ηP2 = 0.45) for fNIR's measures of DLPFC activation. Planned contrasts revealed that RW patients showed greater activation of the FPC/DLPFC to natural reward images relative to both EC patients (d = 1.39) and HC participants (d = 1.60; Fig. 2B). The EC and HC participants did not differ from each other (d = 0.04). Inspection of the means revealed that both EC and HC participants showed lower activation to natural rewards than to the neutral stimuli, whereas RW patients showed relatively increased activation to the natural reward cues. The fNIR's measure of increased oxygenation in response to natural reward images was correlated with the number of days since last drug use (the Pearson r = −0.56; P = 0.04).

(A) Functional near-infrared spectroscopy region of interest; right dorsolateral prefrontal cortex (DLPFC). Red area represents optodes differentiating the recent withdrawal from the extended care and the healthy control groups. (B) Mean change in the oxygenated hemoglobin in DLPFC (optodes 11, 12, 13, 14) to natural reward cues minus control images. Oxy, oxygenated; μM, microMolar. Error bars represent ± 1 SEM. (C) The Pearson correlation (r = −0.48; P = 0.08) between change in the DLPFC oxygenated hemoglobin to the natural reward cues versus the number of days since last drug use.

Inspection of the results from the startle response task confirmed that HCs showed the expected pattern of results for negative, neutral, and positive stimuli (eg, Bradley et al., 1999). An ANCOVA conducted on the startle responses elicited while participants viewed natural reward cues revealed a significant difference for group (F[2,17] = 5.99, P = 0.01, ηP2 = 0.44). As predicted, RW patients showed increased startle amplitude to natural reward cues relative to the EC group (d = 2.17; Fig. 3A), but, despite a moderately large effect size, RW participants did not differ from the HC participants in this small sample (P = 0.11; d = 0.96). The EC and HC participants did not differ (P = 0.15; d = 0.72). Inspection of the means indicated that, as predicted, positive stimuli suppressed the startle response in the HC and EC groups (Fig. 3A), indicative of a positive evaluation. A Pearson correlation (Fig. 3B) suggested that greater number of days since last drug use was associated with more inhibited startle to (ie, a more positive evaluation of) natural reward cues. Finally, a marginal, positive correlation (the Pearson r = 0.39, P = 0.09) between startle response to natural reward cues and fNIRs-based prefrontal activation to natural reward cues across the 3 groups suggested that, in this paradigm, reduced prefrontal activation was associated with a more positive response to the pleasant stimuli.

(A) Mean affect-modulated startle amplitude T-scores to negative, neutral, and positive images for the recently withdrawn, the extended care, and the healthy control participants. Error bars represent ± 1 SEM. (B) The Pearson correlation (r[13] = −0.85, P < 0.001) between startle amplitude to natural rewards (ie, positive) minus neutral images and the number of days since last drug use.

Cortisol and Sleep

An ANCOVA revealed that a main effect for group was found for day-time cortisol (F[2,16] = 7.00, P = 0.007, ηP2 = 0.47). Post hoc analyses confirmed that RW patients had higher cortisol than either HC (d = 2.06) or EC (d = 1.42) participants, whereas HC and EC participants did not differ (d = 0.45; Fig. 4B). Although the planned contrast between RW and HC patients for evening cortisol was significant (d = 1.82), the main effect for group was marginal (F[2,16] = 3.06, P = 0.08, ηP2 = 0.28). A negative Pearson correlation between daytime cortisol and days since last drug use (r = −0.58, P = 0.03) indicated time in treatment away from drug was associated with reduced day-time cortisol.

(A) Eight-day average of time-in-bed and total-sleep-time (minutes) for the recently withdrawn, the extended care, and the healthy control participants, recorded via wrist actigraphy. (B) Mean day-time and evening salivary cortisol levels (ng/mL) averaged across 2 days (5 samples per day) for the normal control, the recently withdrawn, and the extended care participants. All error bars represent ± 1 SEM.

ANCOVAs revealed a main effect group for both time-in-bed (F[2,16] = 3.57, P = 0.05, ηP2 = 0.31; Figure 4A) and the total-sleep-time (F[2,16] = 4.58, P = 0.03, ηP2 = 0.36). Post hoc analyses confirmed that RW patients spent less time in bed than HC patients (d = 1.41), and marginally less time in bed than EC patients (d = 1.17; Fig. 4A). As predicted, post hoc analyses confirmed that RW patients had less total-time-asleep than either HC (d = 1.71) or EC (d = 1.29) patients; HC and EC patients did not differ (d = 0.38). The Pearson correlations between the days since last drug use, the time-in-bed, and the total-sleep-time were positive (r = 0.56, P = 0.04, and r = 0.58, P = 0.03, respectively).


This study demonstrates the feasibility of obtaining objective indices of “recovery” in patients who are involved in different stages of residential treatment. Despite the small sample size, the findings on RW patients, who have been addicted to prescription narcotics, are consistent with the animal model data on opioid-induced allostasis of the reward and stress systems (Koob and Kreek, 2007). This cross-sectional study compared measures of both CNS and HPA axis function at 2 time points in recovery from prescription opiate dependence, and relative to HCs. The results suggest that opioid-dependent patients who have been in supervised residence (with clinically documented abstinence) for 58 to 95 days differ in CNS and HPA axis function from individuals in the earlier stages of recovery. The CNS measures suggested heightened DLPFC responses to drug cues in RW patients compared with patients who had been abstinent for 2 to 3 months. A number of studies have shown increased activation of the DLPFC during drug cue reactivity tasks (Goldstein and Volkow, 2002, 2011; Wilson et al., 2004; Bunce et al., 2012). The increased PFC activation in response to opiates cues may be due to an attentional bias among patients in early recovery (Field and Cox, 2008; Field et al., 2009). A finding that requires further study involved the prefrontal cortical responses to drug stimuli. The RW patients showed the expected increased activation to pill cues relative to neutral stimuli (Wilson et al., 2004), whereas the responses of the HC to pill cues showed little differentiation from neutral stimuli, which was also expected. Although this difference was not statistically significant, a moderate-effect size suggests that HC responses may differ from RW in a larger sample. The attenuated response to pill cues among the EC patients, however, requires further study. Wilson et al. (2004), in their review, suggest that the prefrontal cortical responses of patients in treatment typically do not differ from HCs when confronted with the drug cues. A larger study would help determine if the EC responses were actually lower than those of the HC. One possibility is that, having been in 2 to 3 months of treatment, these patients were actively avoiding or suppressing their attention or response to the pill cues, resulting in decreased frontal activation (and potentially increasing activation in other areas of the brain). This hypothesis would require a different technology (eg, fMRI) to evaluate fully.

The between-groups comparison of the acoustic startle responses to natural rewards was consistent with our hypothesis that, relative to EC and HC participants, RW participants would show evidence of a reduced hedonic response to putatively positive stimuli. This finding is consistent with a growing literature on anhedonia, defined in relationship to responsivity to naturally rewarding stimuli, in opiate-dependent patients (eg, Lubman et al., 2009). The negative correlation between startle amplitude and greater time since last drug use suggests that there may be improved hedonic responses to naturally rewarding stimuli with continued abstinence and treatment.

The EC and HC participants showed relatively decreased fNIRs activation to the natural rewards in this study. Importantly, (1) decreased fNIRs activation was correlated with inhibited startle response to positive stimuli across the 3 groups, suggesting an interpretation of positive hedonic response, and (2) EC and HC participant responses were similar, and both diverged from the responses of the RW. Taken together, the neuroimaging and acoustic startle data in response to images depicting natural rewards differentiated between the RW and EC groups in this study.

Findings related to HPA axis and sleep disturbances in the 3 groups were consistent with the literature and suggested re-regulation of these disturbances with time in recovery. The RW research participants had elevated cortisol relative to HC participants, whereas the EC group showed only moderate elevations in cortisol, falling between the RW and HC groups. Consistent with their elevated cortisol, the RW group tended to spend less time in bed and less time asleep relative to the other 2 groups. Whereas it is likely that the HPA axis function in response to stress and the CNS reward system are linked, larger, longitudinal studies will be required to clarify the interactions and relative time course of recovery in these two systems. It is also possible that comorbid conditions (eg, depression, anxiety), or medications that influence the CNS function, could account for some of the reported findings. However, there were few differences on these potential confounds between the 2 patient groups (Table 1), where the group differences on this study's measures were strongest.


Overall, these results suggest an abstinence-related effect, that is, the CNS and HPA axis function may re-regulate with time and/or time in treatment away from drug. The magnitude of DLPFC activation to both drug cues and natural reward cues, startle responses to natural reward cues, day-time cortisol levels, time in bed and total time spent sleeping were all correlated with the number of days since last drug use, which was time spent in residential treatment.

An important limitation of this pilot study was the small sample size. Several predicted findings failed to reach statistical significance and were likely underpowered. However, the moderate to large effect sizes found in this study suggest that these results are likely to be significant in a larger study. A longitudinal study is necessary to determine the within-subject relationship between the time in treatment (ie, time away from drug) and these measures, and the predictive validity of the findings in relationship to risk, and timing, of relapse. The current findings, and the technology employed in this study, indicate both the promise and the feasibility of applying research-based assessments of allostasis of the reward and stress response systems within a clinical environment. If the results can be validated in a longitudinal study design that examines intermediate-term treatment outcomes, the real-world significance of differential physiological re-regulation can be evaluated.


The authors thank the Caron Treatment Center for hosting the research study, Cheryl Knepper, MA, and Mike Early for their support; Dan Langleben, MD, for his helpful comments and contribution of the heroin stimuli; William Milchak for his ongoing collaborative involvement during the study; and Hasan Ayaz, PhD, and Kurtulus Izzetoglu, PhD, for their involvement in the fNIRs data processing.


Bearn J, Buntwal N, Papadopoulos A, et al. Salivary cortisol during opiate dependence and withdrawal. Addict Biol 2001;6(2):157–162.
Bradley M, Cuthbert B, Lang P. Affect and the startle reflex. In: Dawson ME, Schell AM, Boehmelt AH, eds. Startle Modification: Implications for Neuroscience, Cognitive Science and Clinical Science. Stanford, CA: Cambridge, 1999:157–183.
Brower K, Aldrich M, Hall J. Polysomnographic and subjective sleep predictors of alcoholic relapse. Alcohol Clin Exp Res 1998;22(8):1864–1871.
Brower K, Perron B. Sleep disturbance as a universal risk factor for relapse in addictions to psychoactive substances. Med Hypotheses 2010;74(5):928–933.
Bunce S, Izzetoglu K, Izzetoglu M, et al. Treatment status predicts differential prefrontal cortical responses to alcohol and natural reinforcer cues among alcohol dependent individuals. In: Zhang H, Hussain A, Liu D, Wang Z, eds. Advances in Brain Inspired Cognitive Systems. Berlin, Heidelberg, Germany: Springer, 2012:183–191.
Field M, Cox W. Attentional bias in addictive behaviors: a review of its development, causes, and consequences. Drug Alcohol Depend 2008;97(1/2):1–20.
Field M, Munafo M, Franken I. A meta-analytic investigation of the relationship between attentional bias and subjective craving in substance abuse. Psychol Bull 2009;135(4):589–607.
First MB, Spitzer RL, Gibbon M, Williams JBW. Structured Clinical Interview for DSM-IV-TR Axis I Disorders, Research Version, Patient Edition. (SCID-I/P). New York: Biometrics Research, New York State Psychiatric Institute, 2002.
Foster J, Peters T. Impaired sleep in alcohol misusers and dependent alcoholics and the impact upon outcome. Alcohol Clin Exp Res 1999;23(6):1044–1051.
Goldstein R, Volkow N. Drug addiction and its underlying neurobiological basis: neuroimaging evidence for the involvement of the frontal cortex. Am J Psychiatry 2002;159(10):1642–1652.
Goldstein R, Volkow N. Dysfunction of the prefrontal cortex in addiction: neuroimaging findings and clinical implications. Nat Rev Neurosci 2011;12:652–669.
Grigson P. Reward comparison: the Achilles' heel and hope for addiction. Drug Discov Today Dis Models 2008;5(4):227–233.
Grigson P, Twining R, Carelli R. Heroin-induced suppression of saccharin intake in water-deprived and water-replete rats. Pharmacol Biochem Behav 2000;66(3):603–608.
Heinz A, Wrase J, Kahnt T, et al. Brain activation elicited by affectively positive stimuli is associated with a lower risk of relapse in detoxified alcoholic subjects. Alcohol Clin Exp Res 2007;31(7):1138–1147.
Hunt WA, Barnett LW, Branch LG. Relapse rates in addiction programs. J Clin Psychol 1971;27(4):455–456.
Koob G, Kreek M. Stress, dysregulation of drug reward pathways, and the transition to drug dependence. Am J Psychiatry 2007;164(8):1149–1159.
Koob G, Volkow N. Neurocircuitry of addiction. Neuropsychopharmacology 2010;35(1):217–238.
Kornetsky C, Bain G. Brain-stimulation reward: A model for drug-induced euphoria. In: Adler MW, Cowan A, eds. Modern Methods in Pharmacology, 6: Testing and Evaluation of Drugs of Abuse. New York: Alan R. Liss, 1990:211–231.
Kreek M. Methadone-related opioid agonist pharmacotherapy for heroin addiction: history, recent molecular and neurochemical research and future in mainstream medicine. Ann N Y Acad Sci 2000;909:186–216.
Kreek M, Ragunath J, Plevy S, et al. ACTH, cortisol and β-endorphin response to metyrapone testing during chronic methadone maintenance treatment in humans. Neuropeptides 1984;5(1-3):277–278.
Kreek M, Wardlaw S, Hartman N, et al. Circadian rhythms and levels of β-endorphin, ACTH, and cortisol during chronic methadone maintenance treatment in humans. Life Sci 1983;33(suppl 1):409–411.
Lang P. The emotion probe. Studies of motivation and attention. Am Psychol 1995;50(5):372–385.
Lang P, Bradley MM, Cuthbert BN. International Affective Picture System (IAPS): Affective Ratings of Pictures and Instruction Manual, Technical Report A-8. Gainesville: University of Florida, 2008.
Langleben D, Ruparel K, Elman I, et al. Acute effect of methadone maintenance dose on brain fMRI response to heroin-related cues. Am J Psychiatry 2008;165(3):390–394.
Lubman D, Yücel M, Kettle J, et al. Responsiveness to drug cues and natural rewards in opiate addiction: associations with later heroin use. Arch Gen Psychiatry 2009;66(2):205–212.
Luo S, Ainslie G, Giragosian L, et al. Striatal hyposensitivity to delayed rewards among cigarette smokers. Drug Alcohol Depend 2011;116(1-3):18–23.
Markou A, Weiss F, Gold L, et al. Animal models of drug craving. Psychopharmacology 1993;112(2/3):163–182.
Martin W, Jasinski D, Haertzen C, et al. Methadone—a reevaluation. Arch Gen Psychiatry 1973;28(2):286–295.
Martin W, Wikler A, Eades C, et al. Tolerance to and physical dependence on morphine in rats. Psychopharmacologia 1963;4:247–260.
Meyer R. How to understand the relationship between psychopathology and addictive disorders: another example of the chicken and the egg. In: Meyer R, ed, Psychopathology and Addictive Disorders. New York: Guilford Press, 1986:3–16.
Morgan P, Pace-Schott E, Pittman B, et al. Normalizing effects of modafinil on sleep in chronic cocaine users. Am J Psychiatry 2010;167(3):331–340.
Stewart J. Stress and relapse to drug seeking: studies in laboratory animals shed light on mechanisms and sources of long-term vulnerability. Am J Addict 2003;12(1):1–17.
Substance Abuse and Mental Health Services Administration. Results From the 2007 National Survey on Drug Use and Health: National Findings, NSDUH Series H-34, DHHS Publication No. SMA 08-4343. Rockville, MD: Office of Applied Studies, 2008.
Villringer A, Chance B. Non-invasive optical spectroscopy and imaging of human brain function. Trends Neurosci 1997;20(10):435–442.
Westerberg V, Tonigan J, Miller W. Reliability of Form 90D: an instrument for quantifying drug use. Subst Abus 1998;19(4):179–189.
Wikler A, Martin WR, Pescor FT, et al. Factors regulating oral consumption of an opioid (etonitazene) by morphine-addicted rats. Psychopharmacologia 1963;5(1):55–76.
Wilson S, Sayette M, Fiez J. Prefrontal responses to drug cues: a neurocognitive analysis. Nat Neurosci 2004;7(3):211–214.
Zhao M, Fan C, Du J, et al. Cue-induced craving and physiological reaction in recently and long-abstinent heroin-dependent patients. Addict Behav 2012;37(4):393–398.
Zijlstra F, Veltman D, Booij J, et al. Neurobiological substrates of cue-elicited craving and anhedonia in recently abstinent opioid-dependent males. Drug Alcohol Depend 2009;99(1-3):183–192.

affect modulated startle response; allostasis; cue reactivity; functional near-infrared spectroscopy; natural rewards; prescription opioids

© 2015 American Society of Addiction Medicine