Author’s Accepted Manuscript

Empathy in schizophrenia: A meta-analysis of the

Interpersonal Reactivity Index

Kelsey A. Bonfils, Paul H. Lysaker, Kyle S.

Minor, Michelle P. Salyers

www.elsevier.com/locate/psychres

PII: S0165-1781(16)31016-2
DOI: http://dx.doi.org/10.1016/j.psychres.2016.12.033
Reference: PSY10169
To appear in: Psychiatry Research
Received date: 11 June 2016
Revised date: 18 December 2016
Accepted date: 24 December 2016
Cite this article as: Kelsey A. Bonfils, Paul H. Lysaker, Kyle S. Minor and
Michelle P. Salyers, Empathy in schizophrenia: A meta-analysis of the
Interpersonal Reactivity Index, Psychiatry Research,
http://dx.doi.org/10.1016/j.psychres.2016.12.033
This is a PDF file of an unedited manuscript that has been accepted for
publication. As a service to our customers we are providing this early version of
the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting galley proof before it is published in its final citable form.
Please note that during the production process errors may be discovered which
could affect the content, and all legal disclaimers that apply to the journal pertain.
EMPATHY IN SCHIZOPHRENIA 1

The four-factor conceptualization of empathy in schizophrenia: A meta-analysis

Kelsey A. Bonfils*a, Paul H. Lysakerb,c, Kyle S. Minora, Michelle P. Salyersa

a
Department of Psychology, Indiana University-Purdue University Indianapolis, 402 N.
Blackford St., Indianapolis, IN, United States
b
Psychiatric Rehabilitation and Recovery Center, Roudebush VA Medical Center, 1481 W. 10th
St., Indianapolis, IN, United States
c
Department of Psychiatry, Indiana University School of Medicine, 340 W. 10th St., Indianapolis,
IN, United States

*phone: 317-274-6767; fax: 317-274-6756; email: [email protected]

Abstract

Empathy is a complex construct, thought to contain multiple components. One popular four-

factor conceptualization has been used extensively to measure empathy in schizophrenia research

(empathic concern, perspective-taking, personal distress, and fantasy); however, no recent meta-

analysis has been conducted on the four factors together. The goal of this meta-analysis was to

examine self-reported empathy for each component in the four-factor conceptualization in people

with schizophrenia as compared to healthy controls. A literature search revealed 33

schizophrenia studies that utilized this conceptualization. The Hedges’ g standardized difference

effect size was calculated for each component using a random effects meta-analytic model.

Individuals with schizophrenia scored significantly differently from healthy controls on all

components, exhibiting lower scores for empathic concern, perspective-taking, and fantasy, as

well as greater scores for personal distress. Duration of illness significantly moderated the results

for perspective-taking such that those with a longer duration exhibited greater deficits. Future

work should examine in more detail the impact of heightened personal distress on empathic
EMPATHY IN SCHIZOPHRENIA 2

interaction and investigate the mechanism through which duration of illness impacts empathic

components in people with schizophrenia.

Keywords: social cognition, Interpersonal Reactivity Index, duration of illness, perspective-

taking, fantasy, personal distress

1. Introduction

Empathy is key to our interpersonal relationships, contributing to development of social

networks (Salovey and Mayer, 1989) and acquisition of relationship-maintaining behaviors such

as forgiveness and altruism (Eisenberg and Miller, 1987; Hoffman, 1981, 2000; McCullough et

al., 1998; McCullough et al., 1997). Empathy is a complex construct with multiple components,

most often measured in cognitive and affective domains. Research on empathy in people with

schizophrenia has revealed deficits in this population in both commonly measured empathic

components (Bonfils et al., 2016; Savla et al., 2013). However, reviews and meta-analyses

examining empathy in schizophrenia have omitted empathic components that do not cleanly fit

with conceptualizations of cognitive or affective empathy. Thus, although there is an abundance

of literature to show deficits in cognitive and affective empathy, the field has yet to gain an in-

depth understanding of deficits in other empathic components for people with schizophrenia.

Although cognitive and affective empathy have reached a general consensus in the field,

a four-factor model of empathy has also become increasingly prominent. This four-factor model

includes components that map onto affective empathy and cognitive empathy, but expands that

conceptualization to also include personal distress and fantasy components (Davis, 1983). The

empathic concern factor, which most clearly represents affective empathy, or the emotions felt in
EMPATHY IN SCHIZOPHRENIA 3

response to the situations or experiences of another, corresponds to experiences of warmth,

compassion, and concern for others. The perspective-taking factor, which most clearly represents

cognitive aspects of empathy, corresponds to one’s ability to take the perspective of others.

Personal distress corresponds to the amount of unpleasant emotion experienced upon witnessing

the negative situations of others (self-oriented distress), and fantasy corresponds to the tendency

to place oneself into fictional situations and empathically relate to characters, as in books,

movies, or daydreams (Davis, 1983).

Research in schizophrenia has investigated deficits in cognitive and affective empathy,

but the literature on personal distress and fantasy are less clear. While consistent deficits in

cognitive empathy (Savla et al., 2013) and affective empathy (Bonfils et al., 2016) have been

found in people with schizophrenia, research on personal distress seems to indicate the opposite

finding – that people with schizophrenia may experience an excess compared to healthy controls.

This finding has been reported in several studies (e.g., see Andrews et al., 2013; Fujiwara et al.,

2008; Gizewski et al., 2013; Montag et al., 2012a), but results across studies have yet to be meta-

analyzed. Studies examining the fantasy component, on the other hand, report more disparate

findings, with some finding significant deficits in schizophrenia samples (Derntl et al., 2012b;

Fujiwara et al., 2008; Hooker et al., 2011; Lee et al., 2010) while others find fantasy abilities to

be intact or trending toward greater levels in schizophrenia groups (Fischer-Shofty et al., 2013;

Matsumoto et al., 2015; McCormick et al., 2012).

The omission of the personal distress and fantasy components from meta-analyses of

empathy in schizophrenia has left a gap in the literature, considering the likely importance of

these factors in the ability to empathically interact for people with schizophrenia. Personal

distress, especially, may actually impede the ability to empathically respond to others. Studies
EMPATHY IN SCHIZOPHRENIA 4

finding increased personal distress in schizophrenia align with a previous meta-analysis that

found that people with schizophrenia experience heightened negative emotions compared to

healthy controls when faced with neutral or even positive stimuli (Cohen and Minor, 2010).

Some have argued that excess personal distress and negative emotion may work against

successful empathic interaction, possibly overwhelming the person with schizophrenia, leading

to a non-empathic experience or even disengagement from the social interaction (Horan et al.,

2015). Understanding how levels of personal distress differ in schizophrenia compared to healthy

controls is a key next step to identifying factors that negatively impact empathic interaction in

this population.

There is less evidence linking abilities to relate to fantasy characters to empathic or

functional outcomes, but some literature shows a significant association between greater fantasy

abilities and increased hallucinations and delusions (Sparks et al., 2010), and heightened fantasy

abilities in relatives of people with schizophrenia have been associated with measures of

psychosis risk (Montag et al., 2012b). Further, although no research to our knowledge has

investigated this, it seems plausible that deficits in the ability to relate to fantasy characters may

add another layer of difficulty to interpersonal interaction, where conversations may center on

popular television shows or movies. That is, if the ability to empathize with a fictional character

is impaired, then everyday conversations about those experiences may also be negatively

affected. In addition, reduced ability to relate to fantasy characters may limit the experience or

enjoyment of recreational activities like reading or watching fiction, which could contribute to

poorer quality of life.

In addition to our lack of knowledge about deficits across components in the popular

four-factor empathy model, the body of literature has yet to inform whether empathic deficits are
EMPATHY IN SCHIZOPHRENIA 5

dependent on duration of illness. This question is of the utmost importance – if empathic skills

are found to be intact or considerably less impaired for those early in the course of illness, there

are ramifications for intervention design to better prevent decline in empathic abilities. A meta-

analytic framework offers the opportunity to assess any moderating role of duration of illness on

each of the empathic components. Although few studies have specifically examined the impact

of duration of illness on empathy, more literature has investigated the impact on broader social

cognitive deficits. This work has been mixed, with some studies indicating a longer course of

illness is associated with greater deficits in empathy (Montag et al., 2007) and others indicating

social cognitive abilities decline over the course of illness (Kucharska-Pietura et al., 2005);

however, others indicate deficits begin early, but remain stable (Green et al., 2011; Pinkham et

al., 2007).

Taken together, despite great advances in our knowledge about empathy in

schizophrenia, there are still important gaps in our understanding. The four-factor model has

been used predominantly in the form of the Interpersonal Reactivity Index (IRI; Davis, 1983), a

self-report instrument designed to assess empathic tendencies in each of the four components.

Yet, only one meta-analysis to our knowledge attempted to synthesize the literature on the IRI in

schizophrenia, and it included only six studies (Achim et al., 2011) and did not address

moderators. The current study includes a set of meta-analyses designed to explore empathic

differences between those diagnosed with schizophrenia and healthy controls using the popular

four-factor conceptualization of empathy (via the IRI). As research on empathy is on the rise,

and substantial literature using the IRI has accrued, synthesizing this literature is timely and

necessary to further our understanding of how we might most helpfully intervene on the empathy

construct.
EMPATHY IN SCHIZOPHRENIA 6

This project aimed to quantify the standardized mean difference between schizophrenia

and healthy controls for each component of the four-factor conceptualization of empathy (Davis,

1983): 1) Empathic concern; 2) Perspective-taking; 3) Personal distress; and 4) Fantasy. It

further aimed to examine the role of duration of illness for each of these components. We

hypothesized that healthy controls would score higher for empathic concern and perspective-

taking and lower for personal distress; due to mixed findings in the literature, analyses

examining the standardized mean difference for fantasy abilities were considered exploratory.

Moderator analyses examining duration of illness were also considered exploratory.

2. Method

The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses)

guidelines (Moher et al., 2009) were followed in order to maintain a high level of meta-analytic

quality. The checklist with descriptions of each item and locations where they are addressed in

the manuscript is available from the authors.

2.1. Literature Search

Studies were identified using a variety of methods. We searched PsycINFO,

PsycARTICLES, Web of Science Core Collection, Pubmed, and EMBASE for studies available up

to July 28th, 2015. To be included, studies were required to compare individuals with

schizophrenia to healthy controls using the IRI and be written in English. All searches used the

exploded terms “empath*” and “schizo.*” English language filters were applied when possible.

We also searched reference sections of key meta-analyses and conceptual articles in related areas

(Bora et al., 2009; Derntl and Regenbogen, 2014; Fett et al., 2011; Savla et al., 2013). Finally,

we contacted authors when additional information was needed in order to code an otherwise

eligible study.
EMPATHY IN SCHIZOPHRENIA 7

2.2. Coding

Studies were coded in accordance with a codebook developed based on suggestions from

Card (2012) and Lipsey and Wilson (2001). Sample-level information included year, publication

type, and country. Age, gender (percent female), and race were all coded, but only 7 (21%)

studies reported information for race or ethnicity, so the variable is not described further.

Variables coded only in the schizophrenia samples included diagnosis (percent schizophrenia,

percent schizoaffective disorder, and percent other psychotic disorder) and duration of illness,

when available.

2.2.1. Effect size.

2.2.2. Hedges’ g was calculated for each sample based on means and standard deviations of

each group, representing the standardized mean difference between the schizophrenia

group and healthy controls on the empathy components (as measured by the IRI

subscales). When means and standard deviations were not reported, but other

statistics such as independent samples t-values were available, Hedges’ g was

calculated from these values. Positive values of g signified higher scores in healthy

controls and negative values signified higher scores in the schizophrenia group. Data

were initially coded into Excel where effect sizes were calculated. All data were then

checked before being aggregated into SPSS version 23.0, and later imported into

Comprehensive Meta-Analysis, Version 2 (CMA; Borenstein et al., 2011).

2.3. Meta-Analytic Method

Prior to running meta-analytic statistics, descriptive statistics were conducted. We used

one-study removed sensitivity analyses to assess for the presence of outliers (Borenstein et al.,

2009). This analysis runs the meta-analysis repeatedly, each time with a single sample removed.
EMPATHY IN SCHIZOPHRENIA 8

If this produced a substantial change in the meta-analytic point estimate for any given study, it

was examined as a potential outlier. To assess risk of publication bias, we used Duval and

Tweedie’s (2000) trim and fill approach. This procedure examines observed effects and “trims”

extreme values from small samples. But, this artificially reduces variance, so extreme effects are

then replaced but with mirrored values on the other side of the mean to retain adjustments to the

effect size. This new effect size can be compared to meta-analytic results. If there are no

differences, or the difference is small, greater confidence can be had that publication bias has

minimal effects on results (Borenstein et al., 2009; Card, 2012).

2.3.1. Main analyses.

Effect sizes were calculated using a random effects model. This method accounts for both

within-study and between-study variability (Lipsey and Wilson, 2001) and allows

generalizations to be made from results beyond included studies. Separate meta-analyses were

conducted for each empathic component: empathic concern, perspective-taking, personal

distress, and fantasy. Individual effect sizes were weighted by the inverse variance to account for

standard error in effect size estimates (Card, 2012; Lipsey and Wilson, 2001). Hedges’ g is

similar to Cohen’s d, so mean effect sizes were interpreted with Cohen’s (1992) guidelines:

effect sizes ≤.20 were considered small, effect sizes of .50 were considered medium, and effect

sizes ≥.80 were considered large. All meta-analytic calculations were conducted using the CMA

computer program (Borenstein et al., 2011).

2.3.2. Heterogeneity and moderator analyses.

The Q-statistic was calculated to assess the presence of heterogeneity (Card, 2012) and

the I2 index was calculated (Higgins and Thompson, 2002) to assess how much of the variation

was due to between-studies variability (Huedo-Medina et al., 2006). Moderator analyses were
EMPATHY IN SCHIZOPHRENIA 9

conducted when Q was significant and I2 was 25% or greater. Due to the low power of the Q

statistic to detect heterogeneity when k is small, p < .10 was considered to suggest moderation

(Higgins and Thompson, 2002). If moderator analyses were warranted, meta-regressions were

conducted using a random effects model to examine the impact of duration of illness. Significant

beta weights (p < .05) and decrease in the I2 index indicated significant moderation. All

moderator analyses were conducted in CMA (Borenstein et al., 2011).

3. Results

3.1. Study Selection & Characteristics

A total of 33 independent samples were included in the meta-analyses of empathic

concern, perspective-taking, and personal distress; 31 were included for fantasy. See Figure 1 for

a flow chart of article identification. A total of 1,260 participants with schizophrenia and 1,086

healthy controls were included in the meta-analysis. See Table 1 for detailed characteristics at

the individual sample level, and Table 2 for aggregated study characteristics.

3.2. Sensitivity Analyses

Examination of one-study removed sensitivity analyses and forest plots (available upon

request from the authors) indicated, across meta-analyses, that no study needed to be removed as

an outlier. Point estimates of effect sizes with studies removed did not greatly differ from overall

mean effect sizes, indicating all samples could be retained for main analyses.

3.3 Main Analyses

All meta-analyses exhibited significant main effects, indicating differences between

healthy controls and people with schizophrenia on all four empathy components, though these

differences varied in magnitude and direction. See Table 3 for summary results of the four

components, and see Figures 1-4 in the supplemental online material for forest plots of each
EMPATHY IN SCHIZOPHRENIA 10

meta-analysis. As hypothesized, participants with schizophrenia evidenced significant deficits

(compared to healthy controls) in empathic concern (0.29, 95% CI [0.18, 0.41]) and perspective

taking (0.55, 95% CI [0.43, 0.67]), while showing elevated personal distress (-0.72, 95% CI [-

0.86, -0.58]). These were all medium effects. In exploratory analyses, fantasy exhibited a

positive mean effect size of small magnitude (0.19, 95% CI [0.08, .030]), indicating people with

schizophrenia report being less able to relate to fantasy characters. Across meta-analyses,

moderate to high heterogeneity was detected, with all I2 estimates surpassing the 25% threshold

to examine moderating variables (see Table 3).

Duration of illness was tested as a continuous moderator. Based on the results of meta-

regression analyses, duration of illness significantly moderated the relationship between sample

(i.e., schizophrenia vs. healthy control) and the perspective-taking component such that for every

one year increase in duration of illness, the standardized mean difference effect size is

strengthened by 0.022, indicating those with a more chronic course have greater impairments in

emotional perspective-taking than those earlier in the course of illness. This significant finding

was accompanied by a decrease in the I2 index of 27.21%, indicating a substantial reduction in

heterogeneity when duration of illness is controlled. There was also a trend (p = 0.08) in the

same direction for empathic concern, such that the standardized mean difference is strengthened

by .016 with every one year increase in duration. This trend-level association was accompanied

by a 26.85% decrease in the I2 index. Duration of illness did not significantly moderate personal

distress or fantasy (Table 4).

3.4. Publication Bias

Trim and fill analyses found no evidence of publication bias for empathic concern,

perspective-taking, or fantasy. For personal distress, the trim and fill procedure imputed two
EMPATHY IN SCHIZOPHRENIA 11

values (the imputed funnel plot can be seen in the supplemental online material, Figure 5);

however, revised summary statistics using these imputed values indicated a very similar effect

size and confidence interval (Hedges’ g = -.69; 95% CI [-.83, -.54]), differing from the non-

corrected value by only .03. As the imputed and non-corrected effect sizes are nearly the same,

publication bias can be considered minimal for the personal distress meta-analysis.

4. Discussion

This meta-analysis, including a total of 33 samples reporting results for empathy using

the four-factor conceptualization (Davis, 1983), represents a substantial extension of the past

effort to synthesize this literature in schizophrenia (Achim et al., 2011). Results of all four

components were significant and appear robust to effects of outliers, suggesting that people with

schizophrenia experience deficits in empathic concern, emotional perspective-taking, and the

ability to relate to fictional characters, and at the same time experience heightened personal

distress. Effects were similar to results of Achim and colleagues’ (2011) meta-analysis, though

the fantasy effect size was considerably smaller (g = .19) in this meta-analysis compared to the

level previously reported (d = .45). Differences from that meta-analysis are not surprising, and in

fact we might have expected more differences, considering that this meta-analysis includes five

times as many samples as the meta-analysis by Achim and colleagues (2011).

The empathic concern component displayed an effect of small to moderate magnitude (g

= 0.29), similar to a recent meta-analysis of affective empathy, which included many of the same

studies in its self-report category (g = 0.22; Bonfils et al., 2016). Impairments in affective

empathy are important to note, as affective empathy is key in the development of social networks

(Salovey and Mayer, 1989) and altruistic behavior (Eisenberg and Miller, 1987; Hoffman, 1981,
EMPATHY IN SCHIZOPHRENIA 12

2000), and aspects of empathy have been linked in schizophrenia specifically to social

functioning (Michaels et al., 2014; Shamay-Tsoory et al., 2007; Smith et al., 2014).

The personal distress component, on the other hand, revealed a medium to large negative

effect, indicating that compared to healthy controls, people with schizophrenia report heightened

personal distress when confronted with the experiences and emotions of others. The personal

distress empathic component has been debated in the literature, with some advocating that it

should not be considered empathy at all (Corbera et al., 2013; Horan et al., 2015; Michaels et al.,

2014) because it assesses self-oriented distress (Davis, 1983) rather than emotional-matching,

which many consider key to the empathic experience of emotion (De Vignemont and Singer,

2006; Decety and Jackson, 2004; Derntl and Regenbogen, 2014). However, when observed as a

separate factor as originally intended; (Davis, 1983)), findings indicate that individuals with

schizophrenia perceive experiencing more personal distress than healthy controls.

This aligns with findings from a meta-analysis of laboratory studies investigating

emotional experience conducted by Cohen and Minor (2010) in which people with schizophrenia

displayed heightened aversive emotion in response to positive or neutral lab-based stimuli as

compared to healthy controls. Though respondents did not differ substantially from healthy

controls on experience of expected hedonic emotions, they were simultaneously experiencing

negative emotions that were absent or lessened in healthy controls. This may reflect some aspect

of emotion dysregulation in which individuals with schizophrenia are less able to downregulate

negative emotion in situations that would otherwise be considered neutral or even pleasant, (as

suggested by Cohen and Minor, 2010; Horan et al., 2006a; Horan et al., 2015; Strauss et al.,

2013). In the case of personal distress, it may be normative to experience some distress when

faced with the unpleasant experiences of others (as exhibited by non-zero means of healthy
EMPATHY IN SCHIZOPHRENIA 13

control participants for this scale), but a failure to downregulate that emotion may characterize

those with schizophrenia, leading to increased unpleasant emotion when dealing with others’

situations.

Though personal distress may not directly measure the empathic experience, it is relevant

to empathy research, as increased self-oriented negative emotions may impede empathic

responding. For example, increased personal distress may require one to exert self-control in

order to respond appropriately despite negative feelings. Research in the general population has

shown that self-control is a limited resource that can be depleted (Baumeister et al., 1998;

Muraven and Baumeister, 2000; Muraven et al., 1998), and this has been replicated in

schizophrenia (Leung et al., 2014). Thus, if people with schizophrenia experience heightened

negative emotions in response to the experiences of others, they may have to exert self-control in

order to handle their internal experiences, reducing their cognitive resources available to display

empathic responses and potentially build social connections. A complementary theory from

Corbera et al. (2013) suggests that increased personal distress can push individuals with

schizophrenia to withdraw from social situations entirely, completely negating any opportunity

to respond empathically to the other. Future research should investigate these ideas to better

understand how personal distress impacts empathic responding.

The third empathic component, perspective-taking, was also consistent with hypotheses,

revealing a moderate deficit for individuals with schizophrenia. The perspective-taking

component most closely reflects cognitive empathy. Theory of mind is one aspect of cognitive

empathy, and meta-analyses of this construct have consistently shown large deficits in people

with schizophrenia (Bora et al., 2009; Savla et al., 2013; Sprong et al., 2007). Our effect is

somewhat smaller (g = .55, compared to 1.10, .96, and 1.26 found in the previous 3 cited meta-
EMPATHY IN SCHIZOPHRENIA 14

analyses). The smaller effect may be related to the content of the perspective-taking subscale,

which primarily assesses emotional perspective-taking, whereas many theory of mind

assessments focus only on the ability to discern thoughts and intentions, neglecting the emotional

aspect of knowing the other (see Shamay-Tsoory et al., 2005 for a discussion of this issue).

Alternatively, it could be that people with schizophrenia perceive themselves to be better at

perspective-taking than is reflected in actual performance. Because most meta-analyses of theory

of mind use only performance-based measures, deficits in performance that people with

schizophrenia may not perceive could explain the larger effects found in those meta-analyses.

This is consistent with literature showing that people with schizophrenia report themselves to be

more empathic than do observers (Lysaker et al., 2013) or family members (Bora et al., 2008).

Analyses for the final empathic component, fantasy, were considered exploratory, as the

literature shows mixed findings regarding the ability to relate to fictional characters for people

with schizophrenia. Results revealed a small, but significant, deficit in fantasy abilities in the

schizophrenia group. However, the size of the effect calls into question whether this has clinical

relevance. Especially as compared to effects evident for other types of empathy (and IRI

subscales), the fantasy effect is small, and ability to relate to fictional characters may not be

necessary for empathic interaction. In fact, Davis (1983) asserted that associations between the

fantasy subscale and interpersonal functioning were not expected because one’s ability to get

involved in fictional scenarios from books or movies was not relevant to social relationships, but

rather may reflect aspects of emotionality. Thus, the fantasy component may be less useful in

directly assessing empathic abilities than the other empathic components. However, some

research indicates that reading fiction is associated with increased empathy and prosocial

behavior (Johnson, 2012). In addition, given some literature linking fantasy abilities to increased
EMPATHY IN SCHIZOPHRENIA 15

psychotic symptoms (Sparks et al., 2010) and psychosis risk (Montag et al., 2012b), future

research may try to glean a better understanding of how fantasy abilities affect (or are affected

by) illness course and symptoms.

Across empathy components, moderate to large heterogeneity was observed, indicating

moderators were at work. For perspective-taking, those with a longer duration of illness

exhibited a greater deficit in emotional perspective-taking, such that for every decade of illness

we might expect a decrease in emotional perspective-taking on the order of a small effect size –

0.22. This finding is consistent with some literature asserting greater length of illness negatively

impacts empathic abilities (Achim et al., 2011; Montag et al., 2007), but inconsistent with the

meta-analysis conducted by Savla and colleagues (2013) that found duration of illness was not

significantly associated with theory of mind. It may be that duration of illness has a greater

impact on self-reported perspective-taking abilities (perhaps decreasing self-perception due to

depleted confidence after years of illness) than on performance-based theory of mind

assessments. Alternatively, emotional perspective-taking may be more directly impacted by

duration of illness, as opposed to other forms of perspective-taking usually assessed in theory of

mind measurement. Further research is needed to parse apart these possibilities.

There was also a trend (p = 0.08) in moderator analyses such that those with a greater

duration of illness exhibited less empathic concern. Considering the notoriously low power of

moderator analyses (Borenstein et al., 2009; Hedges and Pigott, 2004), we consider this trend as

pointing toward possible future avenues for additional research. As the empathic concern and

perspective-taking components map most closely onto affective and cognitive empathy, this pair

of findings indicates duration of illness may be impactful for the most commonly measured and

reported components of empathy. There are a multitude of potential reasons for increased deficits
EMPATHY IN SCHIZOPHRENIA 16

with extended course of illness. First, it is possible that the reduced social network size

experienced by many with schizophrenia (Horan et al., 2006b) results in a lack of opportunity to

practice empathic skills, leading to empathic atrophy over time. Second, it could be that

discrimination from others over time as a result of stigmatizing societal views contributes to

reduced empathy felt for others. It might also be that cortical regions involved in empathic

capacity are affected by long-term symptoms, or by use of antipsychotics over time. For

example, research shows reductions in brain volume with extended duration of illness for some

with schizophrenia (Haijma et al., 2013). However, these possibilities are speculative, and future

research is needed to understand how duration of illness might affect empathic abilities.

Results should be interpreted in light of some limitations. This study focused on the most

commonly used four-factor conceptualization of empathy and the IRI as a self-report measure of

that conceptualization. Results may not generalize to other empathic conceptualizations and may

not represent the complete picture regarding empathy deficits. However, while other meta-

analyses have examined cognitive and affective empathy (Bonfils et al., 2016; Savla et al.,

2013), only one other study has examined all four components measured here (Achim et al.,

2011), and that was with a much reduced sample size. Additionally, not all moderators of interest

could be examined here. For example, symptoms and medications were reported variably (both

with regard to symptom assessment and scoring method), precluding examination of these

potentially important moderators. Finally, this meta-analysis is not exempt from limitations of all

meta-analyses; that is, there is always the threat of the “file drawer” problem, and meta-analytic

results are limited by methodological shortcomings of the primary studies (Card, 2012).

Regarding the former, there was evidence in the personal distress funnel plot of potential for

missing study values. However, the effects of publication bias appear minimal, as the corrected
EMPATHY IN SCHIZOPHRENIA 17

effect size computed with the trim and fill procedure differed from the observed effect by only

0.03. Regarding the latter, our results were limited by small samples that employed convenience

sampling methods and often reported incomplete moderator data – these issues should be

considered when interpreting our meta-analytic results.

Taken together, our results indicate significant deficits in empathic concern, perspective-

taking, and fantasy, but heightened scores for personal distress, in people with schizophrenia as

compared to healthy controls. Considering the extensive and ongoing use of this empathic

conceptualization and the IRI, these results point to several avenues for future research. First, the

role of emotion regulation in personal distress and subsequent empathic interaction should be

investigated. Second, interventions based in the reading of fiction to enhance empathy could

provide benefit to people with schizophrenia – future studies may consider investigating the use

of these interventions to enhance social interactions in this group. However, research is also

needed to further assess the relationship between fantasy skills and psychotic symptoms. Third,

research is needed to determine the mechanism through which duration of illness impacts

affective and cognitive empathy, and to identify ways to mitigate the negative impact of longer

duration on those constructs. Finally, additional work is needed to examine further potential

moderating variables, such as symptoms, medication, or services received.

Acknowledgement

This research did not receive any specific grant from funding agencies in the public,

commercial, or not-for-profit sectors. The authors declare no conflicts of interest. The authors

would like to thank the many researchers who supplied data for the meta-analysis or who

corresponded with the authors regarding the details of their work.

EMPATHY IN SCHIZOPHRENIA 18

References
Achim, A.M., Ouellet, R., Roy, M.-A., Jackson, P.L., 2011. Assessment of empathy in first-
episode psychosis and meta-analytic comparison with previous studies in schizophrenia.
Psychiatry Res. 190 (1), 3-8.
Andrews, S.C., Enticott, P.G., Hoy, K.E., Fitzgerald, P.B., 2013. Mirror systems and social
cognition in schizophrenia. Schizophr. Bull. 39, S218.
Baumeister, R.F., Bratslavsky, E., Muraven, M., Tice, D.M., 1998. Ego depletion: is the active
self a limited resource? J. Pers. Soc. Psychol. 74 (5), 1252-1265.
Bonfils, K.A., Lysaker, P.H., Minor, K.S., Salyers, M.P., 2016. Deficits in affective empathy in
schizophrenia: a meta-analysis. Schizophr. Res., Advance online publication.
Bora, E., Gökçen, S., Veznedaroglu, B., 2008. Empathic abilities in people with schizophrenia.
Psychiatry Res. 160 (1), 23-29.
Bora, E., Yucel, M., Pantelis, C., 2009. Theory of mind impairment in schizophrenia: meta-
analysis. Schizophr. Res. 109 (1), 1-9.
Borenstein, M., Hedges, L.V., Higgins, J.P.T., Rothstein, H.R., 2009. Introduction to Meta-
Analysis. John Wiley & Sons, Ltd, Chippenham, Wiltshire, UK.
Borenstein, M., Hedges, L.V., Higgins, J.P.T., Rothstein, H.R., 2011. Comprehensive Meta-
Analysis (version 2). Biostat, Englewood, NJ.
Brown, E.C., Gonzalez-Liencres, C., Tas, C., Brune, M., 2016. Reward modulates the mirror
neuron system in schizophrenia: a study into the mu rhythm suppression, empathy, and
mental state attribution. Soc. Neurosci. 11 (2), 175-186.
Card, N.A., 2012. Applied Meta-Analysis for Social Science Research. Guilford Press, New
York, NY.
Chiang, S.-K., Hua, M.-S., Tam, W.-C.C., Chao, J.-K., Shiah, Y.-J., 2014. Developing an
alternative Chinese version of the Interpersonal Reactivity Index for normal population
and patients with schizophrenia in Taiwan. Brain Impairment 15 (2), 120-131.
Cohen, A.S., Minor, K.S., 2010. Emotional experience in patients with schizophrenia revisited:
meta-analysis of laboratory studies. Schizophr. Bull. 36 (1), 143-150.
Cohen, J., 1992. A power primer. Psychol. Bull. 112 (1), 155-159.
Corbera, S., Cook, K., Brocke, S., Dunn, S., Wexler, B.E., Assaf, M., 2014. The relationship
between functional deficits and empathy for emotional pain in schizophrenia. Biol.
Psychiatry 75 (9), 200S.
Corbera, S., Wexler, B.E., Ikezawa, S., Bell, M.D., 2013. Factor structure of social cognition in
schizophrenia: is empathy preserved? Schizophr Res Treat 2013, 1-13.
Davis, M.H., 1983. Measuring individual differences in empathy: evidence for a
multidimensional approach. J. Pers. Soc. Psychol. 44 (1), 113-126.
EMPATHY IN SCHIZOPHRENIA 19

De Vignemont, F., Singer, T., 2006. The empathic brain: how, when and why? Trends Cogn. Sci.
10 (10), 435-441.
Decety, J., Jackson, P.L., 2004. The functional architecture of human empathy. Behav. Cogn.
Neurosci. Rev. 3 (2), 71-100.
Derntl, B., Finkelmeyer, A., Voss, B., Eickhoff, S.B., Kellermann, T., Schneider, F., Habel, U.,
2012a. Neural correlates of the core facets of empathy in schizophrenia. Schizophr. Res.
136 (1-3), 70-81.
Derntl, B., Regenbogen, C., 2014. Empathy, in: Lysaker, P.H., Dimaggio, G., Brune, M. (Eds.),
Social Cognition and Metacognition in Schizophrenia: Psychopathology and Treatment
Approaches. Elsevier, Waltham, MA, pp. 69-81.
Derntl, B., Seidel, E.-M., Schneider, F., Habel, U., 2012b. How specific are emotional deficits?
A comparison of empathic abilities in schizophrenia, bipolar and depressed patients.
Schizophr. Res. 142 (1-3), 58-64.
Duval, S., Tweedie, R., 2000. Trim and fill: a simple funnel‐plot–based method of testing and
adjusting for publication bias in meta‐analysis. Biometrics 56 (2), 455-463.
Eisenberg, N., Miller, P.A., 1987. The relation of empathy to prosocial and related behaviors.
Psychol. Bull. 101 (1), 91-119.
Fett, A.-K.J., Viechtbauer, W., Dominguez, M.-d.-G., Penn, D.L., van Os, J., Krabbendam, L.,
2011. The relationship between neurocognition and social cognition with functional
outcomes in schizophrenia: a meta-analysis. Neurosci. Biobehav. Rev. 35 (3), 573-588.
Fischer-Shofty, M., Brüne, M., Ebert, A., Shefet, D., Levkovitz, Y., Shamay-Tsoory, S.G., 2013.
Improving social perception in schizophrenia: the role of oxytocin. Schizophr. Res. 146
(1-3), 357-362.
Fujino, J., Takahashi, H., Miyata, J., Sugihara, G., Kubota, M., Sasamoto, A., Fujiwara, H., Aso,
T., Fukuyama, H., Murai, T., 2014. Impaired empathic abilities and reduced white matter
integrity in schizophrenia. Prog. Neuropsychopharmacol. Biol. Psychiatry 48, 117-123.
Fujiwara, H., Shimizu, M., Hirao, K., Miyata, J., Namiki, C., Sawamoto, N., Fukuyama, H.,
Hayashi, T., Murai, T., 2008. Female specific anterior cingulate abnormality and its
association with empathic disability in schizophrenia. Prog. Neuropsychopharmacol.
Biol. Psychiatry 32 (7), 1728-1734.
Gizewski, E.R., Müller, B.W., Scherbaum, N., Lieb, B., Forsting, M., Wiltfang, J., Leygraf, N.,
Schiffer, B., 2013. The impact of alcohol dependence on social brain function. Addict.
Biol. 18 (1), 109-120.
Green, M.F., Bearden, C.E., Cannon, T.D., Fiske, A.P., Hellemann, G.S., Horan, W.P., Kee, K.,
Kern, R.S., Lee, J., Sergi, M.J., 2011. Social cognition in schizophrenia, part 1:
performance across phase of illness. Schizophr. Bull., 854-864.
Haijma, S.V., Van Haren, N., Cahn, W., Koolschijn, P.C.M., Pol, H.E.H., Kahn, R.S., 2013.
Brain volumes in schizophrenia: a meta-analysis in over 18,000 subjects. Schizophr. Bull.
39 (5), 1129-1138.
EMPATHY IN SCHIZOPHRENIA 20

Haker, H., Rössler, W., 2009. Empathy in schizophrenia: impaired resonance. Eur. Arch.
Psychiatry Clin. Neurosci. 259 (6), 352-361.
Hedges, L.V., Pigott, T.D., 2004. The power of statistical tests for moderators in meta-analysis.
Psychol. Methods 9 (4), 426-455.
Higgins, J.P.T., Thompson, S.G., 2002. Quantifying heterogeneity in a meta-analysis. Stat. Med.
21 (11), 1539-1558.
Hoffman, M.L., 1981. Is altruism part of human nature? J. Pers. Soc. Psychol. 40 (1), 121-137.
Hoffman, M.L., 2000. Empathy and moral development: implications for caring and justice.
Cambridge University Press, New York, NY.
Hooker, C.I., Bruce, L., Lincoln, S.H., Fisher, M., Vinogradov, S., 2011. Theory of mind skills
are related to gray matter volume in the ventromedial prefrontal cortex in schizophrenia.
Biol. Psychiatry 70 (12), 1169-1178.
Horan, W.P., Green, M.F., Kring, A.M., Nuechterlein, K.H., 2006a. Does anhedonia in
schizophrenia reflect faulty memory for subjectively experienced emotions? J. Abnorm.
Psychol. 115 (3), 496-508.
Horan, W.P., Pineda, J.A., Wynn, J.K., Iacoboni, M., Green, M.F., 2014. Some markers of
mirroring appear intact in schizophrenia: evidence from mu suppression. Cogn. Affect.
Behav. Neurosci. 14 (3), 1049-1060.
Horan, W.P., Reise, S.P., Kern, R.S., Lee, J., Penn, D.L., Green, M.F., 2015. Structure and
correlates of self-reported empathy in schizophrenia. J. Psychiatr. Res. 66-67, 60-66.
Horan, W.P., Subotnik, K.L., Snyder, K.S., Nuechterlein, K.H., 2006b. Do recent-onset
schizophrenia patients experience a 'social network crisis? Psychiatry 69 (2), 115-129.
Huedo-Medina, T.B., Sánchez-Meca, J., Marín-Martínez, F., Botella, J., 2006. Assessing
heterogeneity in meta-analysis: Q statistic or I² index? Psychol. Methods 11, 193-206.
Johnson, D.R., 2012. Transportation into a story increases empathy, prosocial behavior, and
perceptual bias toward fearful expressions. Pers. Individ. Dif. 52 (2), 150-155.
Kucharska-Pietura, K., David, A.S., Masiak, M., Phillips, M.L., 2005. Perception of facial and
vocal affect by people with schizophrenia in early and late stages of illness. Br. J.
Psychiatry 187 (6), 523-528.
Lam, B.Y.H., Raine, A., Lee, T.M.C., 2014. The relationship between neurocognition and
symptomatology in people with schizophrenia: social cognition as the mediator. BMC
Psychiatry 14 (138), 1-10.
Lee, J., Zaki, J., Harvey, P.O., Ochsner, K., Green, M.F., 2011. Schizophrenia patients are
impaired in empathic accuracy. Psychol. Med. 41 (11), 2297-2304.
Lee, S.J., Kang, D.H., Kim, C.-W., Gu, B.M., Park, J.-Y., Choi, C.-H., Shin, N.Y., Lee, J.-M.,
Kwon, J.S., 2010. Multi-level comparison of empathy in schizophrenia: an fMRI study of
a cartoon task. Psychiatry Res: Neuroimaging 181 (2), 121-129.
EMPATHY IN SCHIZOPHRENIA 21

Lehmann, A., Bahcesular, K., Brockmann, E.M., Biederbick, S.E., Dziobek, I., Gallinat, J.,
Montag, C., 2014. Subjective experience of emotions and emotional empathy in paranoid
schizophrenia. Psychiatry Res. 220 (3), 825-833.
Leung, C.-M., Stone, W.S., Lee, E.H.-M., Seidman, L.J., Chen, E.Y.-H., 2014. Impaired
facilitation of self-control cognition by glucose in patients with schizophrenia: a
randomized controlled study. Schizophr. Res. 156 (1), 38-45.
Lipsey, M.W., Wilson, D.B., 2001. Practical meta-analysis. SAGE Publications, Inc., Thousand
Oaks, California.
Lysaker, P.H., Hasson-Ohayon, I., Kravetz, S., Kent, J.S., Roe, D., 2013. Self perception of
empathy in schizophrenia: emotion recognition, insight, and symptoms predict degree of
self and interviewer agreement. Psychiatry Res. 206 (2-3), 146-150.
Matsumoto, Y., Takahashi, H., Murai, T., Takahashi, H., 2015. Visual processing and social
cognition in schizophrenia: relationships among eye movements, biological motion
perception, and empathy. Neurosci. Res. 90, 95-100.
McCormick, L.M., Brumm, M.C., Beadle, J.N., Paradiso, S., Yamada, T., Andreasen, N., 2012.
Mirror neuron function, psychosis, and empathy in schizophrenia. Psychiatry Res:
Neuroimaging 201 (3), 233-239.
McCullough, M.E., Rachal, K.C., Sandage, S.J., Worthington Jr, E.L., Brown, S.W., Hight, T.L.,
1998. Interpersonal forgiving in close relationships: II. Theoretical elaboration and
measurement. J. Pers. Soc. Psychol. 75 (6), 1586-1603.
McCullough, M.E., Worthington Jr, E.L., Rachal, K.C., 1997. Interpersonal forgiving in close
relationships. J. Pers. Soc. Psychol. 73 (2), 321.
McGuire, J., Barbanel, L., Brüne, M., Langdon, R., 2015. Re-examining Kohlberg's conception
of morality in schizophrenia. Cogn. Neuropsychiatry 20 (5), 377-381.
Michaels, T.M., Horan, W.P., Ginger, E.J., Martinovich, Z., Pinkham, A.E., Smith, M.J., 2014.
Cognitive empathy contributes to poor social functioning in schizophrenia: evidence from
a new self-report measure of cognitive and affective empathy. Psychiatry Res. 220 (3),
803-810.
Moher, D., Liberati, A., Tetzlaff, J., Altman, D.G., Grp, P., 2009. Preferred reporting items for
systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 6 (7),
e1000097.
Montag, C., Brockmann, E.M., Lehmann, A., Muller, D.J., Rujescu, D., Gallinat, J., 2012a.
Association between oxytocin receptor gene polymorphisms and self-rated 'empathic
concern' in schizophrenia. PLoS One 7 (12), e51882.
Montag, C., Heinz, A., Kunz, D., Gallinat, J., 2007. Self-reported empathic abilities in
schizophrenia. Schizophr. Res. 92 (1), 85-89.
Montag, C., Neuhaus, K., Lehmann, A., Krüger, K., Dziobek, I., Heekeren, H.R., Heinz, A.,
Gallinat, J., 2012b. Subtle deficits of cognitive theory of mind in unaffected first-degree
EMPATHY IN SCHIZOPHRENIA 22

relatives of schizophrenia patients. Eur. Arch. Psychiatry Clin. Neurosci. 262 (3), 217-
226.
Muraven, M., Baumeister, R.F., 2000. Self-regulation and depletion of limited resources: does
self-control resemble a muscle? Psychol. Bull. 126 (2), 247-259.
Muraven, M., Tice, D.M., Baumeister, R.F., 1998. Self-control as a limited resource: regulatory
depletion patterns. J. Pers. Soc. Psychol. 74 (3), 774.
Pinkham, A.E., Penn, D.L., Perkins, D.O., Graham, K.A., Siegel, M., 2007. Emotion perception
and social skill over the course of psychosis: a comparison of individuals 'at-risk' for
psychosis and individuals with early and chronic schizophrenia spectrum illness. Cogn.
Neuropsychiatry 12 (3), 198-212.
Regenbogen, C., Kellermann, T., Seubert, J., Schneider, D.A., Gur, R.E., Derntl, B., Schneider,
F., Habel, U., 2015. Neural responses to dynamic multimodal stimuli and pathology-
specific impairments of social cognition in schizophrenia and depression. Br. J.
Psychiatry 206 (3), 198-205.
Salovey, P., Mayer, J.D., 1989. Emotional intelligence. Imagin. Cogn. Pers. 9 (3), 185-211.
Savla, G.N., Vella, L., Armstrong, C.C., Penn, D.L., Twamley, E.W., 2013. Deficits in domains
of social cognition in schizophrenia: a meta-analysis of the empirical evidence.
Schizophr. Bull. 39 (5), 979-992.
Shamay-Tsoory, S.G., Shur, S., Harari, H., Levkovitz, Y., 2007. Neurocognitive basis of
impaired empathy in schizophrenia. Neuropsychology 21 (4), 431-438.
Shamay-Tsoory, S.G., Tomer, R., Berger, B.D., Goldsher, D., Aharon-Peretz, J., 2005. Impaired
“affective theory of mind” is associated with right ventromedial prefrontal damage.
Cogn. Behav. Neurol. 18 (1), 55-67.
Singh, S., Modi, S., Goyal, S., Kaur, P., Singh, N., Bhatia, T., Deshpande, S.N., Khushu, S.,
2015. Functional and structural abnormalities associated with empathy in patients with
schizophrenia: An fMRI and VBM study. J. Biosci. 40 (2), 355-364.
Smith, M.J., Horan, W.P., Cobia, D.J., Karpouzian, T.M., Fox, J.M., Reilly, J.L., Breiter, H.C.,
2014. Performance-based empathy mediates the influence of working memory on social
competence in schizophrenia. Schizophr. Bull. 40 (4), 824-834.
Sparks, A., McDonald, S., Lino, B., O'Donnell, M., Green, M.J., 2010. Social cognition,
empathy and functional outcome in schizophrenia. Schizophr. Res. 122 (1-3), 172-178.
Sprong, M., Schothorst, P., Vos, E., Hox, J., Van Engeland, H., 2007. Theory of mind in
schizophrenia: meta-analysis. Br. J. Psychiatry 191 (1), 5-13.
Strauss, G.P., Kappenman, E.S., Culbreth, A.J., Catalano, L.T., Lee, B.G., Gold, J.M., 2013.
Emotion regulation abnormalities in schizophrenia: cognitive change strategies fail to
decrease the neural response to unpleasant stimuli. Schizophr. Bull. 39 (4), 872-883.
Thirioux, B., Tandonnet, L., Jaafari, N., Berthoz, A., 2014. Disturbances of spontaneous
empathic processing relate with the severity of the negative symptoms in patients with
EMPATHY IN SCHIZOPHRENIA 23

schizophrenia: a behavioural pilot-study using virtual reality technology. Brain Cogn. 90,
87-99.
Wojakiewicz, A., Januel, D., Braha, S., Prkachin, K., Danziger, N., Bouhassira, D., 2013.
Alteration of pain recognition in schizophrenia. Eur J Pain 17 (9), 1385-1392.

Figure 1 Literature Search Diagram (PRISMA)

EMPATHY IN SCHIZOPHRENIA 24

Table 1 - Studies included in meta-analyses

Citation (K=33) Country SSD HC Duration Duration Empathy Hedges’
N N of illness of illness Component g
M SD
(Achim et al., 2011) Canada 31 31 1.7 1.2 IRI-EC 0.06
IRI-PD -0.47
IRI-PT 0.19
IRI-F 0.08
(Andrews et al., 2013)t Australia 18 18 22.1 3.3 IRI-EC 0.37
IRI-PD -0.90
IRI-PT 0.73
IRI-F 0.49
(Brown et al., 2016) Germany 17 17 9.3 6.9 IRI-EC -0.12
IRI-PD -0.90
IRI-PT 0.31
IRI-F 0.11
(Chiang et al., 2014) Taiwan 70 35 -- -- IRI-EC 0.85
IRI-PD 0.3
IRI-PT 0.59
IRI-F 0.13
(Corbera et al., 2013) United 30 24 22.2 10.3 IRI-EC 0.29
States
IRI-PD -0.57
IRI-PT 0.64
IRI-F 0.07
(Corbera et al., 2014)t United 21 26 -- -- IRI-EC -0.09
States
IRI-PD -0.44
IRI-PT 0.62
IRI-F 0.26
(Derntl et al., 2012b) Germany 24 24 11.5 7.6 IRI-EC 0.29
IRI-PD -0.88
IRI-PT 0.33
IRI-F -0.95
(Derntl et al., 2012a)t Germany 15 15 7.30 5.3 IRI-EC 0.15
IRI-PD -1.09
IRI-PT 0.06
IRI-F 0.57
(Fischer-Shofty et al., Israel 34 44 11.78 7.0 IRI-EC -0.10
2013)t
IRI-PD -0.69
IRI-PT 0.71
IRI-F -0.07
(Fujino et al., 2014) Japan 69 69 13.1 9.7 IRI-EC 0.04
IRI-PD -0.57
IRI-PT 0.54
IRI-F 0.29
(Fujiwara et al., 2008) Japan 24 20 10.4 8.4 IRI-EC 0.19
EMPATHY IN SCHIZOPHRENIA 25

Citation (K=33) Country SSD HC Duration Duration Empathy Hedges’

N N of illness of illness Component g
M SD
IRI-PD -0.71
IRI-PT 0.65
IRI-F 0.96
(Gizewski et al., 2013) Germany 24 12 14.2 7.4 IRI-EC 0.62
IRI-PD -1.56
IRI-PT 0.69
IRI-F 0.00
(Haker and Rössler, Switzerland 43 45 11 9.0 IRI-EC 0.20
2009) IRI-PD -0.56
IRI-PT 0.49
IRI-F 0.36
(Hooker et al., 2011) United 21 17 -- -- IRI-EC 0.35
States
IRI-PD -0.32
IRI-PT 0.58
IRI-F 1.00
(Horan et al., 2014) United 30 24 26.8 11.5 IRI-EC 0.84
States
IRI-PD -0.37
IRI-PT 0.76
IRI-F -0.03
(Horan et al., 2015)t United 145 45 19.9 -- IRI-EC 0.26
States
IRI-PD -1.04
IRI-PT 0.63
IRI-F 0.03
(Lam et al., 2014)t China 58 61 13.4 8.8 IRI-EC 0.29
IRI-PD -0.27
IRI-PT 0.23
IRI-F -0.19
(Lee et al., 2011) United 30 22 -- -- IRI-EC 0.73
States
IRI-PD -1.23
IRI-PT 0.58
IRI-F 0.15
(Lee et al., 2010) South 15 18 4.6 3.4 IRI-EC 0.65
Korea
IRI-PD -0.44
IRI-PT 0.35
IRI-F 0.91
(Lehmann et al., 2014) Germany 55 55 10 7.7 IRI-EC 0.25
IRI-PD -1.14
IRI-PT -0.19
(Matsumoto et al., Japan 17 18 15.2 7.9 IRI-EC -0.09
2015) IRI-PD -0.63
IRI-PT 0.22
IRI-F -0.22
EMPATHY IN SCHIZOPHRENIA 26

Citation (K=33) Country SSD HC Duration Duration Empathy Hedges’

N N of illness of illness Component g
M SD
(McCormick et al., United 16 16 15.8 8.8 IRI-EC -0.47
2012) States
IRI-PD -1.51
IRI-PT 0.26
IRI-F -0.32
(McGuire et al., 2015)t Australia 24 20 22.71 10.2 IRI-EC 0.32
IRI-PD -0.62
IRI-PT 0.36
IRI-F 0.10
(Michaels et al., 2014)t United 52 37 14.8 8.7 IRI-EC 0.53
States
IRI-PD -1.07
IRI-PT 1.02
IRI-F 0.32
(Montag et al., 2012a) Germany 145 145 10.4 9.5 IRI-EC 0.07
IRI-PD -0.99
IRI-PT 0.35
(Montag et al., 2007) Germany 45 45 11.6 9.6 IRI-EC -0.17
IRI-PD -1.04
IRI-PT 0.62
IRI-F 0.10
(Regenbogen et al., Germany 20 24 -- -- IRI-EC 0.13
2015) IRI-PD -0.36
IRI-PT 0.09
IRI-F 0.84
(Shamay-Tsoory et al., Israel 26 31 -- -- IRI-EC 0.50
2007) IRI-PD -0.32
IRI-PT 1.02
IRI-F 0.49
(Singh et al., 2015) India 14 14 9.26 6.4 IRI-EC 1.14
IRI-PD 0.14
IRI-PT 1.68
IRI-F 0.11
(Smith et al., 2014) United 60 45 14.4 9.3 IRI-EC 0.46
States
IRI-PD -0.90
IRI-PT 0.83
IRI-F 0.31
(Sparks et al., 2010) Australia 28 25 -- -- IRI-EC 1.29
IRI-PD -1.09
IRI-PT 1.51
IRI-F -0.02
(Thirioux et al., 2014) France 10 10 11.8 1.5 IRI-EC -0.04
IRI-PD -1.52
IRI-PT 0.75
IRI-F -0.32
(Wojakiewicz et al., France 29 27 8 8.0 IRI-EC 0.32
EMPATHY IN SCHIZOPHRENIA 27

Citation (K=33) Country SSD HC Duration Duration Empathy Hedges’

N N of illness of illness Component g
M SD
2013) IRI-PD -0.72
IRI-PT 0.08
IRI-F -0.14
Note. t Supplemental information was provided by authors to assist in coding for these studies.

Table 2 - Study and Sample Characteristics

Sample Characteristics Mean (SD)/Mean Percent Range K

Age, healthy controls 35.8 (5.5) 25.2-46.1 33
Age, schizophrenia spectrum 38.2 (5.7) 24.9-47.9 32
Female, healthy controls 35.9 (15.5) 0-60.0 33
Female, schizophrenia spectrum 31.9 (15.2) 0-53.3 33
Diagnosis
Schizophrenia 94.4 (12.1) 57.1-100 33
Schizoaffective 4.8 (11.4) 0-42.9 33
Other Psychosis 0.8 (3.5) 0-19.4 33
Years since onset 13.6 (6.1) 1.7-26.8 27

Study Characteristics Mean (SD)/Percent Range K

Sample type
Published article 31 (93.9) -- 33
Poster (data from author) 2 (6.1) -- 33
Year 2012 2007-2016 33
SZ Sample size 38.2 (32.0) 10-145 33
HC Sample size 32.9 (24.9) 10-145 33
Total Sample size 71.1 (54.1) 20-290 33
Location
United States 9 (27.3) -- 33
Abroad 24 (72.7) -- 33

Table 3 - Summary of Standardized Mean Difference Effect Size Results

Empathy k ES SE 95% CI z p Q p I2
EMPATHY IN SCHIZOPHRENIA 28

Component
Empathic concern 33 0.29 0.06 [0.18, 4.91 <0.001 56.68 0.005 43.54
0.41]
Perspective-taking 33 0.55 0.06 [0.43, 8.96 <0.001 58.72 0.003 45.50
0.67]
Personal distress 33 -0.72 0.07 [-0.86, - -9.87 <0.001 80.52 <0.001 60.26
0.58]
Fantasy 31 0.19 0.06 [0.08, 3.43 0.001 41.48 0.079 27.67
0.30]
Note. k = number of studies used in the calculation of the mean effect size. ES = Hedges’ g effect size
statistic. SE = standard error. CI = confidence interval. z = test for statistical significance of the mean
effect size. p = 2-tailed p-value associated with the test of statistical significance. Q = test for
heterogeneity. I2 = indicates the extent of between-study variability. Possible values range from 0-100%.

Table 4 - Duration of Illness Moderator Analyses

Empathy Component k B SE 95% CI z p I2

Empathic concern 27 0.016 0.009 [-0.002, 0.035] 1.75 0.080 16.69
Perspective-taking 27 0.022 0.010 [0.001, 0.042] 2.12 0.034 18.29
Personal distress 27 0.003 0.011 [-0.019, 0.025] 0.29 0.775 9.03
Fantasy 25 -0.004 0.011 [-0.026, 0.017] -0.40 0.688 17.10
Note. k = number of studies in the meta-regression. B = regression coefficient. R2= R2 analogue. SE =
standard error. 95% CI = 95% confidence interval. z = test for statistical significance of regression
coefficient, B. p = two-tailed p-value associated with the test of statistical significance. I2 = indicates the
extent of between-study variability.

Highlights

 Deficits in the Interpersonal Reactivity Index subscales were meta-analyzed

 People with schizophrenia were compared to healthy controls
 Deficits found for empathic concern, perspective-taking, and fantasy
 Schizophrenia group had heightened personal distress
 Duration of illness and percent female were significant moderators for some scales