Validation of a method to assess ADHD-related impulsivity in animal models

Abstract

Background: Response inhibition capacity (RIC), the ability to withhold instrumentally reinforced responses, is compromised in ADHD. Most standard methods for assessing RIC in rodents potentially confound motivational, motor, learning, and inhibitory processes, lack sensitivity to pharmacological treatment, and have unknown reliability.

New method: The fixed minimum interval (FMI) schedule of reinforcement and its associated analytical techniques are designed to dissociate inhibitory processes from incentive-motivational and timing processes. This study is aimed at validating the FMI as a method for assessing RIC in animal models. FMI performance was compared across different withholding requirements (0.5, 3, 6 and 21s), deprivation levels, reinforcement rates, and reinforcer magnitudes.

Results and comparison with existing methods: Motivational manipulations differentially affected estimates of incentive motivation but not the FMI-derived index of RIC, θ. Changes in the withholding requirement influenced timed IRTs in a manner consistent with extant timing theories. Individual estimates of RIC were resilient to prolonged changes in motivation but not to changes in FMI schedule. Results indicate that the FMI schedule is not vulnerable to the same limitations associated with existing methods for assessing RIC.

Conclusions: These results support the use of the FMI schedule and associated analytic techniques as tools for assessing RIC in animal models.

Keywords: ADHD; Fixed minimum interval schedule; Impulsivity; Incentive motivation; Response inhibition capacity; Temporal Regulation model.

Figure 1

Schematic description of the FMI schedule of reinforcement and parameters of the Temporal Regulation model of withholding performance. In the FMI schedule, trial onset is signaled by the insertion of a lever. The first lever press initiates an inter-response time (IRT) that terminates with a head entry into the food hopper. IRTs longer than a programmed interval t are reinforced intermittently according to a conjunctive variable-interval (VI) 90-s schedule (for details, see Methods section). The Temporal Regulation model characterizes IRTs as sampled from a mixture of a gamma distribution (with mean μ and standard deviation σ) and an exponential distribution (with mean = standard deviation K). The model further characterizes μ as a linear function of t; the slope of that function, θ, serves as index of response inhibition capacity (RIC; for details see Temporal Regulation model section).

Figure 2

Mean (+/− SEM) median post-R (A) and post-N (B) latencies, mean IRTs (C), and standard deviation of IRTs (D) in Baseline, Pre-feeding, and Mediated conditions across FMI schedules. FMI 6-s data was pooled across instantiations.

Figure 3

Mean cumulative probability distributions of IRTs in each FMI schedule under Baseline conditions. FMI 6 s data are pooled from both instantiations. The solid curves correspond to the fits of the Temporal Regulation model (Equations 1–3).

Figure 4

Mean (+/− SEM) observed and predicted means (A) and standard deviations (B) of IRTs. Predicted values are estimated from the Temporal Regulation model (Equations 1–3).

Figure 5

Mean (+/− SEM) estimates of Temporal Regulation model parameters θ (A), δ (B), w (C), and β (D) in Baseline, Pre-feeding, and Mediated conditions. * Significant difference, p < .05.

Figure 6

Mean (+/− SEM) estimates of Temporal Regulation model parameter P in Baseline, Pre-feeding, and Mediated conditions in each FMI schedule. FMI 6-s estimates were derived from pooled instantiations.

Figure 7

Mean (+/− SEM) median post-R (A) and post-N (B) latencies, mean IRTs (C), and standard deviations of IRTs (D) in Baseline, HRR, and LR conditions. Baseline data is pooled from the two instantiations.

Figure 8

Mean cumulative probability distributions of IRTs in Baseline, HRR, and LR conditions. Baseline data are pooled from both instantiations. The solid curves correspond to the fits of the Temporal Regulation model (Equation 1).

Figure 9

Mean timed IRTs (A) and standard deviations of timed IRTs (B) across Baseline, HRR and LR conditions. * Significant difference, p < .05.