Toward an objective characterization of an anhedonic phenotype: a signal-detection approach

Abstract

Background: Difficulties in defining and characterizing phenotypes has hindered progress in psychiatric genetics and clinical neuroscience. Decreased approach-related behavior and anhedonia (lack of responsiveness to pleasure) are considered cardinal features of depression, but few studies have used laboratory-based measures to objectively characterize these constructs.

Methods: To assess hedonic capacity in relation to depressive, particularly anhedonic, symptoms, 62 participants completed a signal-detection task based on a differential reinforcement schedule. Anhedonia was operationalized as decreased reward responsiveness.

Results: Unequal frequency of reward between two correct responses produced a response bias (i.e., a systematic preference to identify the stimulus paired with the more frequent reward). Subjects with elevated depressive symptoms (Beck Depression Inventory scores >/= 16) failed to show a response bias. Impaired reward responsiveness predicted higher anhedonic symptoms 1 month later, after controlling for general negative affectivity.

Conclusions: Impaired tendency to modulate behavior as a function of prior reinforcement might underline diminished hedonic capacity in depression. When applied to a clinical population, objective assessments of participants' propensity to modulate behavior as a function of reward might provide a powerful tool for improving the phenotypic definition of depression and thus offer a reliable behavioral screening approach for neuroscience studies of depression.

Figure 1

Schematic diagram of the task. After presentation of the mouth stimulus, subjects selected which type of mouth had been presented by pressing either the ‘z’ or the ‘/’ key (counterbalanced across subjects).

Figure 2

Overall effect of task manipulation. Mean accuracy (A), reaction time (B), response bias (C), and discriminability (D) for the entire sample (n = 62). Error bars represent standard errors. For accuracy and RT, the rich condition (black bars) refers to the stimulus associated with more frequent reward, whereas the lean condition (light gray bars) refers to the stimulus associated with less frequent reward.

Figure 3

Mean response bias (A) and (B) changes in response bias for high (black bars; n = 15) and low (light gray bars; n = 21) BDI subjects. Error bars represent standard errors.

Figure 4

Scatterplot and Pearson’s correlation between ΔResponse Bias (block 3 - block1) at Time 1 and BDI “melancholic” subscore at Time 2.

Figure 5

Scatterplot and Pearson’s correlation between ΔResponse Bias (block 3 - block1) at Time 1 and ΔResponse Bias (block 3 - block1) at Time 2.