Table 5.

Summary of main findings from studies of EBC in schizophrenia.

Study	Summary of major findings
Taylor and Spence (54)	CR – Trend for increased percent visual CRs in “psychotics” vs. “neurotics.”
O’Connor and Rawnsley (55)	CR – No significant difference in number of CRs in response to CS-alone trials between groups (chronic paranoid SZ, chronic non-paranoid SZ, control).
	Extinction – Chronic paranoid SZ had significantly smaller “extinction scores” than controls.
Spain (56)
Auditory	CR – Increased overall number of CRs in SZ vs. control, but effect not significant when examining subgroups matched for skin potential. SZ had significantly more visual than auditory CRs; opposite relationship in HNs. CRs for auditory EBC fewer in SZ vs. HN (but no statistical test reported).
Visual	CR – Increased overall number of CRs in SZ vs. control, but effect not significant when examining subgroups matched for skin potential. SZ had significantly more visual than auditory CRs; opposite relationship in HNs. CRs for visual EBC greater in SZ vs. HN (but no statistical test reported).
Sears et al. (57)	CR – SZ had significantly higher %CRs than controls and reached 70% CR learning criterion significantly faster (i.e., earlier in the experiment) than controls. Significantly shorter onset latency of all blinks in SZ vs. controls during paired trials (however, difference is not significant when group differences in conditioning level were accounted for and for CS-alone trials). CR amplitude significantly increased in SZ vs. controls in CS-alone trials.
	UR – Significantly longer UR latency in SZ vs. controls on US-alone trials.
Hofer et al. (58)	CR – Trend for controls to develop first CR before SZ. No significant difference between S+ and S− in SZ; there was a significant difference in controls for increased CRs to S+ vs. S−. Significantly greater %CRs in controls vs. SZ for S+ but no significant difference for S−. Significant group x reinforcement type (S+ or S−) x block interaction indicated controls showed increased %CRs in response to S+ as the experiment progressed.
Stevens et al. (59)	CR – No significant differences between groups in number of trials to reach learning “criterion” (i.e., 5 consecutive trials with an eyeblink response <500 ms pre-US onset to S+ but not S−).
Marenco et al. (60)
Trace	CR – Analysis using the entire CR window appeared to be contaminated by spontaneous blinks (especially in SZ). A second analysis examining when in the CR window responses occurred revealed that SZ demonstrated increased early conditioned responses vs. controls, and slightly fewer later responses vs. controls. Frequency of early responses did not increase over time for SZ; control participants demonstrated trend-level increases in early responses over time. No significant effects when examining the last 500 ms as the CR window.
Delay	CR – No group differences in %CRs. Longer CR onset and peak latency for SZ vs. controls in “conditioners” during paired trials and CS-alone trials. More efficient “workratio” (a measure of CR efficiency of closing the eye at the time of US onset) in SZ vs. control “conditioners” during paired trials and CS-alone trials.
	UR – %URs significantly lower in SZ vs. controls in entire sample during paired trials. UR amplitude did not decrease across blocks in SZ vs. control “conditioners” during paired trials. For CS-alone trials, %UR-range responses significantly decreased in SZ vs. controls for entire sample (even larger effect when examining “conditioners” only).
Brown et al. (61)	CR – Significantly fewer %CRs overall in SZ vs. controls, and a trend for controls acquiring more CRs over time than SZ. Significantly shorter CR onset and peak latency in SZ vs. controls. Controls demonstrated decreased CR onset variability over time; SZ did not.
	UR – Trend for longer UR peak latency in SZ vs. control. Extinction – Significantly shorter CR onset and peak latency for SZ vs. controls.
Edwards et al. (62)	CR – Marginally significant difference between groups in learning, as indexed by the difference between mean %CRs in the last two blocks and mean %CRs in the first two blocks. Significantly higher %CRs in controls vs. SZ in block 9 of conditioning.
	UR – No significant differences in UR peak amplitude for paired or unpaired trials. No significant correlation between unpaired UR peak amplitude and mean CR amplitude.
Bolbecker et al. (63)	CR – Significantly decreased %CRs and shorter CR peak latency in SZ vs. controls.
	UR – Significantly slower UR peak latency in SZ vs. controls during paired trials. Significantly higher UR peak amplitude in SZ vs. controls for paired and unpaired trials.
	Extinction – Trend for fewer CRs during extinction for SZ vs. controls.
Bolbecker et al. (64)	CR – Decreased %CRs in SZ vs. controls across ISIs and later (i.e., closer to US) CR onset latency in SZ vs. controls across ISIs.
	UR – Significantly shorter UR latency in controls vs. SZ when first ISI presentation examined only (effect not significant when both first and second ISI presentations are considered).
Forsyth et al. (65)	CR – Decreased %CRs in SZ and SPD vs. controls, specifically in later blocks of conditioning. Trend for shorter CR peak latency in SZ and SPD vs. controls. CR amplitudes larger in a few later blocks in controls vs. SPD and SZ.
	UR – Significantly higher UR peak amplitude in SZ vs. controls and SPD.
Parker et al. (66)	CR – Significantly greater %CRs in controls compared to SZ in middle and late phases of conditioning. CR peak latency significantly shorter in SZ in middle phase of conditioning.
Bolbecker et al. (67)	CR – Significantly lower rate of learning in SZ and relatives compared to controls. Controls increase in %CRs over time more than relatives and SZ.
	UR – Larger UR amplitude during paired trials only in SZ vs. controls.
Coesmans et al. (68)	CR – Significantly fewer %CRs in SZ compared to controls, with a trend-level group x block interaction. Controls demonstrated significantly higher learning index (defined as the difference in first and last block number of CRs) vs. SZ.

SZ, individuals with schizophrenia; HN, healthy non-psychiatric controls; SPD, individuals with schizotypal personality disorder.