EEG to Primary Rewards: Predictive Utility and Malleability by Brain Stimulation

Abstract

Theta burst stimulation (TBS) is thought to affect reward processing mechanisms, which may increase and decrease reward sensitivity. To test the ability of TBS to modulate response to strong primary rewards, participants hypersensitive to primary rewards were recruited. Twenty men and women with at least two opposite-sex, sexual partners in the last year received two forms of TBS. Stimulations were randomized to avoid order effects and separated by 2 hours to reduce carryover. The two TBS forms have been demonstrated to inhibit (continuous) or excite (intermittent) the left dorsolateral prefrontal cortex using different pulse patterns, which links to brain areas associated with reward conditioning. After each TBS, participants completed tasks assessing their reward responsiveness to monetary and sexual rewards. Electroencephalography (EEG) was recorded. They also reported their number of orgasms in the weekend following stimulation. This signal was malleable by TBS, where excitatory TBS resulted in lower EEG alpha relative to inhibitory TBS to primary rewards. EEG responses to sexual rewards in the lab (following both forms of TBS) predicted the number of orgasms experienced over the forthcoming weekend. TBS may be useful in modifying hypersensitivity or hyposensitivity to primary rewards that predict sexual behaviors. Since TBS altered the anticipation of a sexual reward, TBS may offer a novel treatment for sexual desire problems.

Fig 1. Monetary and Vibratory Incentive Delay (MID/VID) task and reward schedule.

Fig 2. Experiment setup.

Fig 3

Alpha in response to reward for MID and VID during (A) Anticipation and (B) Receipt (both TBS conditions included).

Fig 4

Alpha in response to sexual reward (win) versus non-reward (miss) (A) Anticipation and (B) Receipt (both TBS conditions included).

Fig 5. Prediction of orgasm count over the coming Friday, Saturday, and Sunday by EEG alpha in anticipation of sex reward (both brain stimulation conditions included).

The plot reflects the ordinal assumptions of conditional logit models (CLM). CLM do not require interval spacing, only that each “level of orgasm” be ordered For example, no value of 5 exists because no one reported 5 orgasms, and each logit is calculated only relative to the next value in the ordered list of orgasm counts.

Fig 6

Alpha during vibratory incentive delay to (A) anticipation and (B) receipt.