437676

2012
ISP0010.1177/0020764012437676Aoki et al.International Journal of Social Psychiatry

E CAMDEN SCHIZOPH

Article

International Journal of

Mental illness among journalists:

Social Psychiatry
0(0) 1­–14
© The Author(s) 2012
A systematic review Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0020764012437676
isp.sagepub.com

Yuta Aoki,1,2,3 Estelle Malcolm,1 Sosei Yamaguchi,1

Graham Thornicroft1 and Claire Henderson1

Abstract
Background: Mass media depictions of people with mental illness have a strong influence on public attitudes, to the
extent that changes in these depictions can reduce public stigmatization of people with such illness. Journalists’ mental
health may influence their depiction of those with mental illness, but little is known about this.
Aims: To investigate mental illness among journalists in five key areas: (1) journalists’ mental health status; (2) journalists’
personal attitudes towards mental illness; (3) attitudes and support journalists expect or have experienced from col-
leagues when they have a mental health problem; (4) effect of journalism’s professional culture on the course of mental
illness; and (5) effect of journalism’s professional culture on mass media depictions of people with mental illness.
Methods: We performed a systematic screening of MEDLINE, PsycINFO, EMBASE, Web of Science and the Cochrane
Library regarding the study aims.
Results: We identified 19, 12, seven and four studies for aims 1, 2, 3, and 4, respectively. No articles were found for aim 5.
Conclusions: The prevalence of post-traumatic stress disorder (PTSD) among journalists is higher than that among the
general population. Journalists have positive personal attitudes towards mental illness, but there are perceived workplace
disincentives to disclose mental health problems.

Keywords
journalist, journalism, mental disorder, stigma, prejudice

Introduction
The mass media is the main source of information on men- the general population in every country, why do many jour-
tal health and mental illness for the general population nalists continue to portray mental illnesses with so little
(Yankelovich, 1990). A great deal of research conducted on sympathy or understanding?
newspaper portrayals of mental illness has concluded that In this context, the aim of this paper was to conduct a
most often stories about mental illness focus on negative systematic review of mental illness among journalists. Our
aspects, especially danger and violence, and include pejorative objectives were five-fold and assessed: (1) the mental health
and discriminatory terminology (Lawrie, 2000). These repre- status of journalists; (2) journalists’ personal attitudes towards
sentations play an active part in shaping and sustaining what mental health problems; (3) how journalists are supported
mental illness means in society (Wahl, 1995). However, one when mentally unwell; (4) the effect of journalism’s profes-
study showed that changes in mass media coverage of mental sional culture on the course of mental illness; and (5) the
health could change attitudes towards mental health among
the general population (Henderson & Thornicroft, 2009).
Why do journalists describe mental illness in such nega- 1Health Service and Population Research Department, Institute of
tive ways? One survey reported that their primary concern Psychiatry, King’s College London, UK
2Department of Psychiatry, Tokyo Metropolitan Health and Medical
is to attract and maintain readership (Henderson, 1996).
Treatment Corporation, Ebara Hospital, Tokyo, Japan
Therefore, news reports emphasize the newsworthy rather 3Department of Neuropsychiatry, Graduate School of Medicine,

than the worthy. University of Tokyo, Japan

On the other hand, journalism can be a stressful occupa-
Corresponding author:
tion (Kelly, 2008). Journalists need to be on site at scenes of
Yuta Aoki, Department of Neuropsychiatry, Graduate School of
disaster or brutality that may traumatize them psychologi- Medicine, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo
cally. When there is reason to suspect that the rate of mental 113-8655, Japan
illness among journalists may be at least as high as among Email: yuaoki-tky@umin.ac.jp

Downloaded from isp.sagepub.com at PENNSYLVANIA STATE UNIV on September 15, 2016

2 International Journal of Social Psychiatry 0(0)

effect of journalism’s professional culture on mass media 1980 and June 2010. We also used Google, checked and
depictions of people with mental illness. emailed the relevant official associations’ websites (Dart
In particular, we aimed to test the following five inter- Center for Journalism & Trauma, International Federation
linked hypotheses: of Journalists, European Journalism Centre, Association of
European Journalists, and Vlaamse Verenigning van
1. Journalists have higher prevalence rates of mental ill- Journalisten) to find grey literature and ongoing studies.
ness, especially post-traumatic stress disorder (PTSD),
depression and alcohol dependence, compared to the
general population. Search strategy
2. Journalists have positive personal attitudes towards The lead reviewer initially screened all studies. A selection
people with mental illness. of excluded studies and all that met the inclusion criteria
3. The professional culture of journalism makes the dis- were screened by the second reviewer. Agreement was
closure of mental illness difficult. reached with the first researcher regarding any contested
4. Concealment has a negative impact on the course of studies. Figure 1 shows the flow chart of the search strategy.
mental illnesses among journalists by delaying recogni- The search terms used are provided in Appendix 1.
tion and treatment.
5. Journalism’s professional culture contributes to nega-
tive portrayals of mental illnesses. Results
Exclusions
Method Our search strategy identified 2,250 studies. Most that were
To achieve these objectives we performed a systematic review excluded were studies about the ‘influence of journalism’ or
in the following way. ‘depictions of mental health’ rather than papers focusing
directly on mental illness among journalists.
Eligibility criteria
Objective 1: Prevalence of mental health
Participants. Identified as journalists or journalism students
problems among journalists
in any relevant field.
Study characteristics. Cohort or cross-sectional studies with Exclusions. From a review of abstracts and references,
original quantitative data and no restrictions on the location, 42 studies were considered eligible for hypothesis testing.
number of participants or response rate. Narrative descriptions Among these, 23 were excluded because they were reviews
were excluded. We included studies that contained the follow- (n = 8), narratives (n = 11), not about mental health (n = 2),
ing information related to hypotheses 1–5, respectively: or did not meet participant criteria (n = 2).
Included studies. Nineteen studies (Table 1), all of which
1. Mental health problems diagnosed/evaluated using were cross-sectional surveys, met the inclusion criteria
standardized criteria. and were published between 1982 and 2010. Seven studies
2. Journalists’ personal attitudes towards people with were from North America, five from Europe, four from
mental health problems. Asia, and one each from Australia and Africa. Sample sizes
3. Journalists’ experiences or expectations about the atti- ranged from 12 to 876. The proportion of men ranged from
tudes of their colleagues or employers regarding mental 48.5% to 90%. Random sampling was adopted in three
illness. studies (Campbell, Heath, Bouknight, Rudd, & Pender,
4. Any quantitative data or source regarding the influence 2009; Feinstein, Owen, & Blair, 2002; Finzen, 1996), strat-
of journalistic culture on mental health problems ified sampling in one (McMahon, 2001), and one study
among journalists. (Feinstein et al., 2002) covered the target population almost
5. The relationship between the culture of journalism and completely. In almost all studies, the authors evaluated
how journalists portray mental illness. participants using self-report measures. For the diagnosis of
PTSD, the Impact of Event Scale (IES) (Mastroianni & Noto,
2008; Pyevich, Newman, & Daleiden, 2003; Reed, 2008), the
Information sources Impact of Event Scale-Revised (IES-R) (Engelhard et al.,
The PRISMA statement (Liberati et al., 2009) for a systematic 2007; Feinstein et al., 2002; Hatanaka et al., 2008; Liberati
review was followed. A systematic search of abstracts written et al., 2009; Marais & Stuart, 2005; Smith, 2009), the PTSD
in English, Italian and Japanese was conducted in MEDLINE, Checklist Civilian Version (PCL) (Cosper & Hughes, 1982;
PsycINFO, EMBASE, Web of Science and the Cochrane McMahon, 2001; Newman, Simpson, & Handschuh, 2003;
Library. Potentially relevant studies written in German and Pyevich et al., 2003; Simpson & Boggs, 1999) and the
Chinese were translated by colleagues. Reference lists were Post-traumatic Diagnostic Scale (PDS) (Yankelovich, 1990)
also scanned. All studies were published between January were used. For depression, the Center for Epidemiologic

Downloaded from isp.sagepub.com at PENNSYLVANIA STATE UNIV on September 15, 2016

Aoki et al. 3

Figure 1. Process of study selection.

Studies Depression Scale (CES-D) (Simpson & Boggs, to be high among journalists who had faced extraordinary
1999; Weidmann, Fehm, & Fydrich, 2008) and the Beck events like war or a tsunami. Some studies showed that type
Depression Inventory-II (BDI-II) (Engelhard et al., 2007; and number of traumatic events were independent risk fac-
Feinstein et al., 2002) were used. The General Health tors for scoring higher on quantitative evaluations of symp-
Questionnaire (GHQ) (Engelhard et al., 2007; Mastroianni toms of PTSD and for the diagnosis of PTSD (Dworznik,
& Noto, 2008; Reed, 2008; Smith, 2009) was also com- 2008; Newman et al., 2003; Pyevich et al., 2003; Simpson
monly used for evaluation of general mental health. & Boggs, 1999). For example, journalists who developed
PTSD. The prevalence of PTSD among journalists PTSD tended to have reported on more events (Dworznik,
ranged from 0% to 33%. The overall prevalence of PTSD 2008). In addition, those who had reported on events involv-
from 11 studies (Dworznik, 2008; Feinstein et al., 2002; ing death or injury were more likely to develop PTSD
Feinstein & Nicolson, 2005; Fukuoka, 2008; Hatanaka (Pyevich et al., 2003).
et al., 2010; Newman et al., 2003; Pyevich et al., 2003; Sin, Major depression. There were only two studies by the same
Chan, & Huak, 2005; Smith, 2009; Teegen & Grotwinkel, author (Feinstein et al., 2002; Feinstein & Nicolson, 2005)
2001; Weidmann et al., 2008) with 15 subgroups (n = 3,327) about the prevalence of major depression among journalists.
was calculated to be 7.2%. The prevalence of PTSD tended Prevalence rates ranged from 5.3% to 21.4%. The samples in

Downloaded from isp.sagepub.com at PENNSYLVANIA STATE UNIV on September 15, 2016

 4
Table 1. Papers identified to test hypothesis 1.

Authors and Demographic information Sampling Measure Target Tool of Diagnostic Main findings
country method method disease evaluation time
frame

N Age Job Male Response

experience ratio rate
(years)
Weidmann & 81 34.1 8.06 49% N/A Convenience Self- PTSD IES-R Lifetime MS of intrusion = 7.01
Papsdorf, 2010 sampling report MS of avoidance = 8.22
Germany
GHQ-12 Past four MS = 11.4
weeks
65 35.7 8.71 48% N/A IES-R Lifetime N/A
GHQ-12 Past four MS = 11.25
weeks
Hatanaka 270 35.8 N/A 90% 86% Convenience Self- PTSD IES-R At the Prevalence of PTSD = 6%
et al., 2010 sampling report time of
Japan exposure
Beranuy, Oberst, 47 19.0 N/A 45% N/A Convenience Self- Psychological GSI At the MS 0.65:0.74
Carbonell, & sampling report distress time of male: female
Chamarro, 2009 exposure
Spain
PSDI 1.56:1.62
PST 33.9:33.5
Weidmann et al., 61 39.4 N/A 72% N/A Convenience Self- PTSD PDS Eight Prevalence of PTSD = 6.6%
2008 sampling report Depression months
Germany after
traumatic
event
CES-D N/A

Downloaded from isp.sagepub.com at PENNSYLVANIA STATE UNIV on September 15, 2016

Sin et al., 2005 12 N/A N/A N/A N/A Convenience Self- PTSD IES At the The prevalence of PTSD = 33%
Singapore sampling report time of
exposure
GHQ-28 M total score = 2.58
Feinstein & 38 36.6 13.3 84% 85% Random Self- PTSD IES-R At the Overall prevalence of MS of intrusion = 7.3
Nicolson, 2005 sampling report Depression time of PTSD = 15% MS of avoidance = 5.3
Canada Substance exposure Depression = 7% MS of arousal = 3.8
abuse
BDI-II MS = 7.2
GHQ MS = 4.3
47 38.9 14.5 75% IES-R MS of intrusion = 7.1
MS of avoidance = 5.8
International Journal of Social Psychiatry 0(0)

MS of arousal= 4.2
 Table 1. (Continued)

Authors and Demographic information Sampling Measure Target Tool of Diagnostic Main findings
Aoki et al.

country method method disease evaluation time

frame

N Age Job Male Response

experience ratio rate
(years)
BDI-II MS = 8.6
GHQ MS = 4.3
Marais & Stuart, 50 31.9 N/A 64% N/A Convenience Self- PTSD IES-R At the 32% of participants scored 20 or higher
2006 sampling report time of
South Africa exposure
Pyevich et al., 866 35.8 11.5 51% 24% Convenience Self- PTSD PCL-C At the Prevalence of PTSD = 4.3%
2003 sampling report time of
USA exposure
Feinstein et al., 140 39.2 15.6 79% 83% Census Self- PTSD IES-R At the MS = 20.2
2002 survey report Depression time of
Canada Substance exposure
abuse
BDI-II MS = 10.1
GHQ MS = 19.8
28 N/A N/A N/A N/A Systematic Interview PTSD N/A Lifetime Prevalence = 28.6%
sampling
At the 10.7%
time
Pre-war 3.6%
Depression Lifetime 21.4%
At the 7.1%

Downloaded from isp.sagepub.com at PENNSYLVANIA STATE UNIV on September 15, 2016

time
Pre-war 3.6%
Substance Lifetime 14.3%
abuse At the 7.1%
time
Pre-war 3.6%
107 39.0 15.5 71% 80% Not Self- PTSD IES-R At the MS = 9.1
mentioned report Depression time of
Substance exposure
abuse
BDI-II MS = 6.4

(Continued)
5
 6
Table 1. (Continued)

Authors and Demographic information Sampling Measure Target Tool of Diagnostic Main findings
country method method disease evaluation time
frame

N Age Job Male Response

experience ratio rate
(years)
GHQ MS = 16.6
19 N/A N/A N/A N/A Systematic Interview PTSD N/A Lifetime Prevalence = 0%
sampling
(from the
population
above)
Depression 5.3%
Substance 5.3%
abuse
Teegen & 61 37 13 61% N/A Convenience Self- PTSD PCL-C At the Prevalence of PTSD = 13%
Grotwinkel, 2001 sampling report time of
Germany exposure
Depression CES-D N/A N/A
Freinkel, 15 37.6 15.2 60% 83% Random Self- Dissociative Original One At least 60% of participants experienced more
Koopman, & sampling report symptom questionnaire month than one dissociation symptom
Spiegel, 1994 after
USA exposure
Cosper & 25 35 N /A N/A N/A Random Interview Alcoholic Original At the Mean frequency of drinking was 6.84 drink
Hughes, 1982 sampling problems questionnaire time of occasions a week
Canada exposure
Dworznik, 2008 280 36.8 13.7 87% N/A Convenience Self- PTSD PCL-C At the Prevalence of PTSD = 9.3%
USA sampling report time of
exposure

Downloaded from isp.sagepub.com at PENNSYLVANIA STATE UNIV on September 15, 2016

Newman et al., 875 36.3 12.4 83% N/A Convenience Self- PTSD PCL-C At the Prevalence of PTSD = 6.7%
2003 sampling report time of
USA exposure
Original Prevalence of PTSD = 5.9%
questionnaire
McMahon, 2001 32 N/A 9.1 53% 21% Convenience Self- PTSD IES At the MS of intrusion = 19.9
Australia sampling report time of MS of avoidance =15.7
trauma
At the MS of intrusion = 4.7
time of MS of avoidance = 5.5
survey
International Journal of Social Psychiatry 0(0)
 Table 1. (Continued)

Authors and Demographic information Sampling Measure Target Tool of Diagnostic Main findings
Aoki et al.

country method method disease evaluation time

frame

N Age Job Male Response

experience ratio rate
(years)
GHQ-28 At the MS of somatic = 7.7
time of MS of anxiety = 9.2
trauma MS of social dysfunction = 8.1
MS of depression = 3.1
27 N/A 16.0 At the N/A
time of
survey
IES At the N/A
time of
survey
GHQ-28 At the N/A
time of
survey
Matsui, 753 N/A N/A N/A N/A Convenience Self- PTSD IES-R Lifetime Prevalence of PTSD
unpublished sampling report TV Manager 5.5%
Japan (n = 127)
Journalist 6.2%
(n = 177)
Newspaper Manager 5.0%
(n = 80)
Journalist 12.3%

Downloaded from isp.sagepub.com at PENNSYLVANIA STATE UNIV on September 15, 2016

(n = 219)
Hatanaka, 12 N/A N/A 58% N/A Convenience Interview Work- Original Lifetime All reported that they felt stress while working
unpublished sampling related questionnaire
Japan stress
Smith, 2009 167 42.2 18.5 48% 18% Stratified Self- PTSD PCL-C At the Prevalence of PTSD = 9.7%
USA sampling report time of
exposure
Simpson & Boggs, 131 N/A N/A 57% N/A Snowball Self- PTSD IES In the MS of intrusion = 12.0
1999 sampling report previous MS of avoidance = 12.6
USA week

GSI = Global Severity Index, PSDI = Positive Symptom Distress Index, PST = Positive Symptom Total, MS = mean score
7
8 International Journal of Social Psychiatry 0(0)

these studies were mainly war journalists, although there was hypothesis. Among these, one was excluded because par-
one group of domestic journalists as a comparison. ticipants did not meet the inclusion criteria.
Alcohol abuse or dependence. Not enough information Included studies. Seven studies (Cameron, 2007;
was obtained to discuss alcohol problems among journalists, Greenberg et al., 2009; Hazell, Hazell, Waring, & Sly, 1999;
although some studies (Cosper & Hughes, 1982; Feinstein Newman et al., 2003; Pieton, 2009; Reed, 2008; Simpson &
et al., 2002) suggested that journalists tended to be heavy drink- Boggs, 1999) were included in the current review. All stud-
ers and had high rates of abuse of substances such as cannabis. ies administered original questionnaires or conducted inter-
views. These studies comprised two groups. One group
explored the kinds of reactions journalists expect when
Objective 2: Journalists’ personal attitudes
they disclose their mental health problems (Cameron, 2007;
towards mental health problems
Greenberg et al., 2009; Reed, 2008). The other investigated
Exclusions. From a review of abstracts and references, support, knowledge or counselling from employers or edu-
14 studies were considered eligible to be read in full for this cators (Hazell et al., 1999; Newman et al., 2003; Pieton,
hypothesis. Among these, two were excluded for not meet- 2009; Simpson & Boggs, 1999).
ing participant criteria and not being quantitative. Expectations. Forty per cent of journalists feared losing
Included studies. Twelve studies (Table 2) were eligible the confidence of their employer and colleagues by con-
and were divided into three thematic categories. For the first fessing they had been traumatized (Reed, 2008). In one
category, ‘journalists’ attitudes towards their own mental health study, 40% felt some stigma attached to being traumatized
problems’, four studies were included (Dworznik, 2008; and thought it was evidence of weakness (Cameron, 2007)
Greenberg, Gould, Langston, & Brayne, 2009; Reed, 2008; (Table 3).
Simpson & Boggs, 1999). All of the studies targeted PTSD Support. Only 20% of journalists indicated that their
and used questionnaires that were not widely used or company had support policies for journalists who cover
accepted to assess favourable personal attitudes towards PTSD. trauma stories (Pieton, 2009), and 11% had been informed
For the second category, ‘journalists’ attitudes towards of the risk of being traumatized (Newman et al., 2003).
mental health problems among other people’, five studies were Thirty per cent of journalists had employers who offered
included (Abasiubong, Ekott, & Bassey, 2007; Campbell et al., them one-to-one meetings related to their stress. However,
2009; Finzen, 1996; Gutiérrez-Lobos & Holzinger, 2000; 70% of journalists felt ‘stressed out’ without any support
Holzinger, Kaup, & Gutierrez-Lobos, 2002). With the excep- from colleagues (Simpson & Boggs, 1999).
tion of a study from Nigeria (Abasiubong et al., 2007), jour-
nalists showed favourable and liberal attitudes towards other
Objective 4: The effect of journalism’s
people’s mental health problems (Gutiérrez-Lobos &
professional culture on the course of
Holzinger, 2000; Holzinger et al., 2002). One interventional
mental illness
cohort study was conducted. The survey involved journalism
students and psychiatric residents who attended weekend Exclusions. No studies were excluded.
workshops over a six-month period. Professors from both Included studies. Four studies (Table 4) were eligible.
journalism and psychiatry departments gave lectures, and Three reported lack of support (Hatanaka et al., 2010;
attitudes towards mental illness were evaluated using original Newman et al., 2003; Weidmann et al., 2008) as a risk factor
questionnaires before the lectures began and after they had for high symptom scores for PTSD, and the other (Smith,
been running for six months. Findings suggested that education 2009) reported that organizational stress was the strongest
can change the attitudes of journalists (Campbell et al., 2009). predictor of high PTSD scores.
For the third category, ‘journalists’ own help-seeking’,
four studies were eligible (Feinstein et al., 2002; Newman
Objective 5: The effect of journalism’s
et al., 2003; Ruggiero, Rheingold, Resnick, Kilpatrick, &
professional culture on mass media
Galea, 2006; Weidmann et al., 2008). In these four studies,
depictions of people with mental illness
13% to 39% of journalists stated they were seriously thinking
of seeking or had already sought professional help. Exclusions. From a review of abstracts, six studies were
eligible to be read in full for the hypothesis. All studies
were excluded because they did not include the original
Objective 3: Attitudes and support quantitative data (n = 6), were not quantitative (n = 2),
journalists expect or have experienced from were only about depiction and not the relationship between
colleagues or employers when they have a mental illness and mass media depiction (n = 2), were not
mental health problem related to depiction (n = 1), or did not meet participant
Exclusions. From a review of abstracts and references, eight criteria (n = 1).
studies were considered eligible to be read in full for this Included studies. None were eligible for the current review.

Downloaded from isp.sagepub.com at PENNSYLVANIA STATE UNIV on September 15, 2016

 Table 2. Papers identified to test hypothesis 2.
Authors and Demographic information Sampling method Measure Target disease Tool of evaluation Main findings
country method
Aoki et al.

N Age Male ratio Response rate

Journalists’ attitudes towards their own mental health problems
Greenberg et al., 124 N/A 64% N/A Convenience sampling Self-report PTSD Attitude to PTSD Schedule Attitudes to PTSD were favourable
2009 Help-Seeking Stigma Attitudes to seeking help were more neutral
England Questionnaire
Reed, 2008 808 N/A N/A N/A Convenience sampling Self-report PTSD Original questionnaire 62% of participants answered that traumatic
USA stress is an under-addressed problem in the news
industry
Dworznik, 2008 280 36.8 87% N/A Convenience sampling Self-report PTSD Original questionnaire 62.3% agreed they personally could be
USA psychologically affected
Simpson & Boggs, 131 N/A 57% N/A Snowball sampling Self-report PTSD Original questionnaire 59% of participants were aware of stress in other
1999 reporters who have covered a story in which
USA someone was hurt
Journalists’ attitudes towards mental health problems among other people
Gutiérrez-Lobo & 43 38.7 63% N/A Not mentioned Interview Compulsory Original questionnaire Compared with medical students, journalists
Holzinger, 2000 admission displayed a more liberal attitude towards the
Austria mentally ill
Campbell et al.,   5 N/A N/A N/A Not mentioned Self-report Schizophrenia Original questionnaire Among educated people, there is a probability
2009 16 N/A N/A N/A Depression that the attitude changes
USAa ADHD
Gutierrez-Lobos & 43 38.7 63% N/A Not mentioned Interview Not specified Original questionnaire Compared with the general population, journalists
Holzinger, 2000 markedly less often charged mentally disordered
Austria people with murder, rape and disorderly conduct
Finzen et al., 1996 20 N/A 25% N/A Not mentioned Self-report Schizophrenia Original questionnaire Journalists displayed a favourable attitude towards
Germany schizophrenia
Abasiubong et al., 210 39.4 44% N/A Random sampling Self-report Not specified Taylor and Dear Inventory There is a widespread negative attitude to mental
2007 of Community Attitude to illness among journalists
Nigeria Mental Health

Downloaded from isp.sagepub.com at PENNSYLVANIA STATE UNIV on September 15, 2016

Journalists’ own help-seeking
Reed, 2008 808 N/A N/A N/A Convenience sampling Self-report PTSD Original questionnaire 13% of participants currently need counselling
USA services
Weidmann et al., 61 39.4 72% N/A Convenience sampling Self-report PTSD Original questionnaire 39.4% have sought help
2008 Depression
Germany
Feinstein et al., 140 39.2 79% 83% Census survey Self-report Not specified Original questionnaire 24.6% had received psychotropic medication,
2002 psychotherapy or a combination of both
Canada 107 39.0 71% 80% Not mentioned 16.2% had received psychotropic medication,
psychotherapy or a combination of both
Newman et al., 875 36.3 83% N/A Convenience sampling Self-report PTSD Original questionnaire 8% reported consulting mental health
2003 professionals for help specifically related to
USA responses to assignments
aAll studies but this one were designed as a cross-sectional study, or data from cross-sectional survey in the longitudinal study were included in the current study
9
10 International Journal of Social Psychiatry 0(0)

Table 3. Papers identified to test hypothesis 3.

Demographic information
Authors
and Male Response Sampling Measure Target
country N Age ratio rate methoda method disease Main findings
Expectation
Greenberg 124 N/A 64% N/A Convenience Self- PTSD 38% agreed with the statement: ‘I would be less
et al., 2009 sampling report likely to be given roles/tasks of responsibility’
England
Reed, 2008 808 N/A N/A N/A Convenience Self- PTSD 39% of journalists answered there is some stigma
USA sampling report attached to mental health in their workplace
Cameron, 20 N/A 75% N/A Not Interview PTSD All participants think seeking help is thought to
2007 mentioned Self- be a sign of weakness
USA report
Support
Pieton, 48 47.5 69% 10% Stratified Self- PTSD 20.7% of participants answered that their
2009 sampling report company has policies for journalists who cover
USA trauma stories
Newman et 875 36.3 83% N/A Convenience Self- PTSD 11% of journalists are warned by their employer
al., 2003 sampling report that gathering news might be hazardous to
USA mental health
Simpson 131 N/A 57% N/A Snowball Self- PTSD 30% of participants said that their employer
& Boggs, sampling report helped them or others handle stress
1999
USA
Hazell 23 N/A N/A 43%b Convenience Self- Suicide All schools of journalism were categorized into
et al., 1999 sampling report low category regarding knowledge and attitude
Australia about suicide and suicide prevention
aAll studies used original questionnaires as a tool of evaluation
bParticipants of this survey are faculty of journalism in the university

Table 4. Papers identified to test hypothesis 4.

Demographic information

Authors Job
and experience Male Response Sampling Measure Target Tool of
country N Age (years) ratio rate method method disease evaluation Main finding
Hatanaka 270 35.8 N/A 90% 86% Convenience Self- PTSD IES-R “Social support from
et al., 2010 sampling report Original supervisor” is statistically
Japan questionnaire significantly correlated
with the score of IES
Weidmann 61 39.4 N/A 72% N/A Convenience Self- PTSD PDS Symptom of depression
et al., 2008 sampling report Depression CES-D and PTSD are correlated
Germany Original with low degree of
questionnaire acknowledgment of their
supervisor and colleagues
Newman 875 36.3 12.4 83% N/A Convenience Self- PTSD Original Poor social support is a
et al., 2003 sampling report questionnaire risk factor for developing
USA PTSD
Smith, 174 42.2 18.5 48% 18% Stratified Self- PTSD PCL-C Organizational stressors
2009 sampling report Original were the strongest
USA questionnaire predictor of PTSD scores

Discussion PTSD in this profession tends to be higher than that among

the general population relevant for comparison. For example,
This is the first systematic review of studies addressing men- 11 subgroups of domestic journalists (n = 3,148 participants
tal health problems among journalists. The prevalence of in total) (Dworznik, 2008; Feinstein et al., 2002; Fukuoka,

Downloaded from isp.sagepub.com at PENNSYLVANIA STATE UNIV on September 15, 2016

Aoki et al. 11

2008; Hatanaka et al., 2010; Pyevich et al., 2003; Smith, Because of the paucity of information about potential
2009; Weidmann et al., 2008) showed an average prevalence confounding factors except age and gender, we were not able
of PTSD of 6.5%, although the samples reported were pre- to investigate potential risk factors in the meta-analysis.
dominantly male. To compare this figure with that of the However, included studies reported length of career (Simpson
general population, it is necessary to consider the diagnostic & Boggs, 1999; Weidmann et al., 2008), type and number
time frame, the gender ratios in the samples, and their age of trauma exposures (Dworznik, 2008; Newman et al., 2003;
distribution. Pyevich et al., 2003; Simpson & Boggs, 1999), and poor
Regarding the diagnostic time frame, four studies with social support (Hatanaka et al., 2010; Nairn, Coverdale, &
five subgroups reported lifetime prevalence (Feinstein et al., Claasen, 2001; Sin et al., 2005; Weidmann & Papsdorf, 2010)
2002; Fukuoka, 2008; Hatanaka et al., 2008; Weidmann & as risk factors for symptoms of PTSD. These suggested risk
Papsdorf, 2010). Among the domestic journalist samples, factors correspond with the psychological resilience model
the weighted mean of the lifetime prevalence of PTSD of (King, Vogt, & King, 2004). This model is supported by
included studies (n = 622) was 7.9%. This rate was equal to previous studies (Connor, 2006; Gold et al., 2000) and implies
the 7.9% lifetime prevalence in the general population the importance of the role of journalists’ managing editors
(Kessler, Sonnega, Bromet, Hughes, & Nelson, 1995). in mitigating these risk factors by providing workplace sys-
On the other hand, six subgroups reported a point preva- tems to identify and respond to employee mental health
lence (Dworznik, 2008; Hatanaka et al., 2010; Newman problems. Some studies (Pieton, 2009; Simpson & Boggs,
et al., 2003; Pyevich et al., 2003; Smith, 2009; Teegen & 1999) have shown that journalists are neither sufficiently
Grotwinkel, 2001). The weighted mean of the point preva- informed of risks nor supported in terms of mental health
lence of PTSD of included samples (n = 2,526) was 6.4%. problems. Journalists’ personal attitudes towards mental ill-
These results should be interpreted cautiously because of ness are generally positive (Reed, 2008), but many think
possible participant gender bias. It has been reported that they will lose the confidence of employers and colleagues
women are twice as likely as men to have PTSD at some if they disclose that they were traumatized. Considering the
point in their lives. anxiety journalists have about being stigmatized as having
Unfortunately, the gender ratio of five domestic samples a ‘mental disorder’, the prevalence of mental health problems
assessed for lifetime prevalence was not available. Among among them should be interpreted with particular caution.
the six domestic groups assessed for point prevalence, 70% Because all included studies except one (Feinstein et al.,
of participants were male. For adult Americans, the lifetime 2002) relied on self-report questionnaires, the accuracy of
prevalence of PTSD in women and men has been reported the diagnosis is debatable compared with that obtained from
as 10.4% and 5.0%, respectively. We calculated the weighted interviews (Adkins, Weathers, McDevitt-Murphy, & Daniels,
mean point prevalence of PTSD among a 70% male popula- 2008; Brewin et al., 2002; Foa, Riggs, Dancu, & Rothbaum,
tion as 6.6%. To control for gender bias in the total sample 1993). The anxiety of being stigmatized could affect accuracy
of included studies, lifetime prevalence in the current analy- of self-report measures. A convenience sampling method
sis, 7.9%, should be compared with 6.6%. could also affect the estimated prevalence of PTSD.
We used a meta-analytic approach to evaluate gender bias Although we found no studies with quantitative data that
in each of the studies. To assess the effect of gender on the analysed the effect of journalism’s professional culture on the
prevalence of PTSD, we performed the Pearson correlation depiction of mental illness, some addressed this theme indi-
test using the male ratio and the prevalence of PTSD in studies rectly. They reported that the combination of reporting a few
that included that information (Dworznik, 2008; Feinstein & crimes committed by people with mental illness, creating a
Nicolson, 2005; Hatanaka et al., 2010; Marais & Stuart, 2005; stereotype of people with mental illness (Mastroianni & Noto,
Newman et al., 2003; Pyevich et al., 2003; Smith, 2009; Teegen 2008) and the lack of differentiation in other portrayals of
& Grotwinkel, 2001; Weidmann et al., 2008). Results revealed mental illness invite the reader to generalize the stereotype to
no correlation between them (p = .242). This suggests the other persons with mental health problems (Nairn et al., 2001).
importance of other known risk factors for developing PTSD Furthermore, media practices directed at engaging readers
such as perceiving a threat to one’s life (Voges & Romney require the use of cases and a style of writing that force readers
2003) or factors specific to journalists such as length of career. to draw upon common-sense knowledge of mental illness to
The prevalence of PTSD among foreign correspondents understand the text (Allen & Nairn, 1997). Last, to sell stories,
who cover extraordinarily disastrous situations such as wars danger is emphasized (Matas, el-Guebaly, Peterkin, Green, &
(Feinstein et al., 2002; Feinstein & Nicolson, 2005) and tsu- Harper, 1985). The Glasgow Media Group (Henderson, 1996)
namis (Sin et al., 2005; Weidmann et al., 2008) tends to be showed how news stories or dialogues closely resemble fic-
higher than among domestic journalists recruited in the tional writing in the first section after the headline. Although
included studies who mainly covered less extreme situations these findings were not evaluated quantitatively, they suggest
that citizens may face as well, such as fires or car accidents. that journalists write articles to emphasize fear using a stereo-
This prevalence ranged from 6.6% to 28.6%, which is similar type of mental illness to capture readers’ interest, even though
to the figure among war veterans (Engelhard et al., 2007). they personally know the facts are different.

Downloaded from isp.sagepub.com at PENNSYLVANIA STATE UNIV on September 15, 2016

12 International Journal of Social Psychiatry 0(0)

Limitations coverage of mental health-related topics; and (2) whether

educational interventions specifically targeted at this type
Several attributes of the studies reviewed showed that the of news coverage are needed.
overall scientific quality of these papers varied. Thus, the
current review has a number of limitations. Relying on self- References
report questionnaires is the major limitation. Previous studies
Abasiubong, F., Ekott, J. U., & Bassey, E. A. (2007). A comparative
have found that the prevalence of PTSD estimated this way
study of attitude to mental illness between journalists and nurses
is higher than the figure obtained by structured psychiatric in Uyo, Nigeria. African Journal of Medicine and Medical
interviews (Engelhard et al., 2007; Ruggiero et al., 2006). Science, 36, 345–351.
The accuracy of a diagnosis of major depression by self- Adkins, J. W., Weathers, F. W., McDevitt-Murphy, M., &
report has been questioned (Eaton, Neufeld, Chen, & Cai, Daniels, J. B. (2008). Psychometric properties of seven self-
2000). Furthermore, a convenience sampling method, low report measures of posttraumatic stress disorder in college
response rate or small sample size in some studies limited students with mixed civilian trauma exposure. Journal of
our ability to draw conclusions. Finally, some studies in the Anxiety Disorders, 22, 1393–1402.
current review included similar but different occupations, Allen, R., & Nairn, R. G. (1997). Media depiction of mental illness:
and the definition of a journalist was often unclear. An analysis of the use of dangerousness. Australian and
New Zealand Journal of Psychiatry, 31, 375–381.
Beranuy, M., Oberst, U., Carbonell, X., & Chamarro, A. (2009).
Conclusions Problematic internet and mobile phone use and clinical symp-
toms in college students: The role of emotional intelligence.
To test the five key hypotheses, we performed a systematic Computers in Human Behavior, 25, 1182–1187.
review. The studies included showed that prevalence of PTSD Brewin, C. R., Rose, S., Andrews, B., Green, J., Tata, P.,
among journalists was higher than among the general popula- McEvedy, C., & Foa, E. B. (2002). Brief screening instrument
tion. Some studies showed that journalists were aware of for post-traumatic stress disorder. British Journal of Psychiatry,
their risk of developing PTSD, and a minority had already 181, 158–162.
used psychiatric services. However, there were perceived Cameron, J. (2007). Journalism and trauma. MSc research report,
disincentives to disclosing mental health problems. It is not Institute of Work Psychology, University of Sheffield, UK.
Campbell, N. N., Heath, J., Bouknight, J., Rudd, K., & Pender, J.
clear whether employers are reluctant to inform journalists
(2009). Speaking out for mental health: Collaboration of future
about the potential for mental health problems and to provide journalists and psychiatrists. Academic Psychiatry, 33, 166–168.
help, or whether they lack the knowledge and resources to Connor, K. M. (2006). Assessment of resilience in the aftermath
do so. Some studies suggested that despite journalists’ posi- of trauma. Journal of Clinical Psychiatry, 67, 46–49.
tive attitudes towards people with psychiatric problems, they Cosper, R., & Hughes, F. (1982). So-called heavy drinking occu-
write articles in a way contrary to these attitudes to keep pations: Two empirical tests. Journal of Studies on Alcohol,
readers’ attention. However, the lack of evidence for this 43, 110–118.
meant we could not test our fifth hypothesis that the profes- Dworznik, J. (2008). The psychology of local news: Compassion
sional culture of non-disclosure adversely affects coverage fatigue and posttraumatic stress in broadcast reporters, photog-
of mental illness. raphers, and live truck engineers. PhD thesis, College of Com-
munication and Information / School of Communication Studies,
Kent State University, USA. Retrieved from http://etd.ohiolink.
Future research edu/view.cgi/Dworznik%20Gretchen%20J.pdf?kent1210513135
Eaton, W. W., Neufeld, K., Chen, L. S., & Cai, G. (2000). A com-
These findings indicate a clear direction for future research. parison of self-report and clinical diagnostic interviews for
To assess the prevalence of mental health problems among depression. Archives of General Psychiatry, 57, 217–222.
journalists and their attitudes towards people with mental Engelhard, I. M., van den Hout, M. A., Weerts, J., Arntz, A., Hox,
illness, studies of sufficient sample size and response rate J. J., & McNally, R. J. (2007). Deployment-related stress and
are required. Structured diagnostic interviews with well- trauma in Dutch soldiers returning from Iraq: Prospective study.
established properties should be considered. Research regard- British Journal of Psychiatry, 191, 140–145.
ing the effects of personal attitudes versus perceived Feinstein, A., & Nicolson, D. (2005). Embedded journalists in the
professional culture on mass media depiction of those with Iraq war: Are they at greater psychological risk? Journal of
mental illness requires prospective studies linking news cov- Traumatic Stress, 18, 129–132.
Feinstein, A., Owen, J., & Blair, N. (2002). A hazardous profession:
erage to participants. The interpretation of this would be
War, journalists, and psychopathology. American Journal of
further enhanced by qualitative interviews with journalists Psychiatry, 159, 1570–1575.
and editors on the apparent contradiction between personal Finzen, A. (1996). Meinungen zur Schizophrenie: Eine Befragung
attitudes and coverage. Finally, key issues for future research von Journalistinnen und Journalisten. Psychiatrische Praxis,
are: (1) whether development of workplace policies in media 23, 294–295.
organizations to facilitate disclosure and help-seeking for Foa, B. E., Riggs, D. S., Dancu, C. V., & Rothbaum, O. B. (1993).
mental health problems would promote positive change in Reliability and validity of a brief instrument for assessing

Downloaded from isp.sagepub.com at PENNSYLVANIA STATE UNIV on September 15, 2016

Aoki et al. 13

post-traumatic stress disorder. Journal of Traumatic Stress, 6, Marais, A., & Stuart, A. D. (2005). The role of temperament in the
459–473. development of post-traumatic stress disorder among journal-
Freinkel, A., Koopman, C., & Spiegel, D. (1994). Dissociative ists. South African Journal of Psychology, 35, 89–105.
symptoms in media eyewitnesses of an execution. American Mastroianni, F. C., & Noto, A. R. (2008). Newsmaking on drugs:
Journal of Psychiatry, 151, 1335–1339. A qualitative study with journalism professionals. Journal of
Fukuoka, Y. (2008). Critical incident stress of journalists. Retrieved Psychoactive Drugs, 403, 293–300.
from http://www.human.tsukuba.ac.jp/~ymatsui/introduction. Matas, M., el-Guebaly, N., Peterkin, A., Green, M., & Harper, D.
html (1985). Mental illness and the media: An assessment of attitudes
Gold, P. B., Engdahl, B. E., Eberly, R. E., Blake, R. J., Page, W. F., and communication. Canadian Journal of Psychiatry, 30,
& Frueh, B. C. (2000). Trauma exposure, resilience, social sup- 12–17.
port, and PTSD construct validity among former prisoners of Nairn, R., Coverdale, J., & Claasen, D. (2001). From source mate-
war. Social Psychiatry and Psychiatric Epidemiology, 35, 36–42. rial to news story in New Zealand print media: A prospective
Greenberg, N., Gould, M., Langston, V., & Brayne, M. (2009). study of the stigmatizing processes in depicting mental illness.
Journalists’ and media professionals’ attitudes to PTSD and Australian and New Zealand Journal of Psychiatry, 35, 654–659.
help-seeking: A descriptive study. Journal of Mental Health, Newman, E., Simpson, R., & Handschuh, D. (2003). Trauma expo-
18, 543–548. sure and post-traumatic stress disorder among photojournalists.
Gutiérrez-Lobos, K., & Holzinger, A. (2000). Mentally ill and Visual Communication Quarterly, 10, 4–13.
dangerous? Attitudes of female journalists and medical students. Pieton, M. (2009). Media company policies concerning journalists
Psychiatrische Praxis, 27, 336–339. who cover traumatic events. MA thesis, College of Journalism
Hatanaka, M., Fukuoka, Y., Matsui, Y., Ando, K., Koshiro, E., and Mass Communication, Kent State University, USA.
Itamura, H., & Inoue, K. (2008). Exploratory study on traumatic Retrieved from http://etd.ohiolink.edu/send-pdf.cgi/Pieton%20
stress in journalists, Part 2: A report on the findings of an inter- Marla.pdf?kent1260654915
view survey on broadcasting journalists and managers. Uni- Pyevich, C. M., Newman, E., & Daleiden, E. (2003). The relation-
versity of Tsukuba, Japan. Retrieved from http://www.human. ship among cognitive schemas, job-related traumatic exposure,
tsukuba.ac.jp/~ymatsui/pdf/Hatanaka08.pdf and posttraumatic stress disorder in journalists. Journal of
Hatanaka, M., Matsui, Y., Ando, K., Inoue, K., Fukuoka, Y., Traumatic Stress, 16, 325–328.
Koshiro, E., & Itamura, H. (2010). Traumatic stress in Japanese Reed, E. (2008). Traumatic stress in photojournalism. MSc thesis,
broadcast journalists. Journal of Traumatic Stress, 23, 173–177. Brooks Institute of Photography, USA. Retrieved from http://
Hazell, P., Hazell, T., Waring, T., & Sly, K. (1999). A survey of www.leareed.com/EricWeb/EricReedThesisBrooksInst08.pdf
suicide prevention curricula taught in Australian universities. Ruggiero, K. J., Rheingold, A. A., Resnick, H. S., Kilpatrick, D.
Australian and New Zealand Journal of Psychiatry, 33, 253–259. G., & Galea, S. (2006). Comparison of two widely used PTSD-
Henderson, C., & Thornicroft, G. (2009). Stigma and discrimination screening instruments: Implications for public mental health
in mental illness: Time to change. Lancet, 373, 1928–1930. planning. Journal of Traumatic Stress, 19, 699–707.
Henderson, L. (1996). Selling suffering: Mental illness and media Simpson, R., & Boggs, J. (1999). An exploratory study of traumatic
values. In G. Philo (Ed.), Media and mental distress (pp. 18–36). stress among newspaper journalists. Journalism and Commu-
London: Longman Publishing. nication Monographs, 1, 1–26.
Holzinger, A., Kaup, B., & Gutierrez-Lobos, K. (2002). Potentially Sin, S. S., Chan, A., & Huak, C. Y. (2005). A pilot study of the
dangerous behavior in the mentally ill: Attitudes of journalists impact of the Asian Tsunami on a group of Asian media work-
and medical students toward compulsory admission. International ers. International Journal of Emergency Mental Health, 7,
Journal of Offender Therapy and Comparative Criminology, 46, 299–305.
445–452. Smith, R. J. (2009). Trauma and journalism: Exploring a model of
Kelly, W. (2008). High anxiety. American Journalism Review. risk and resilience. The Sciences and Engineering, 69, 5058.
Retrieved from http://www.ajr.org/Article.asp?id=4478 Teegen, F., & Grotwinkel, M. (2001). Traumatic exposure and
Kessler, R. C., Sonnega, A., Bromet, E., Hughes, M., & Nelson, C. B. post-traumatic stress disorder of journalists: An internet-based
(1995). Posttraumatic stress disorder in the national comorbid- study. Psychotherapeut, 46, 169–175.
ity survey. Archives of General Psychiatry, 52, 1048–1060. Voges, M. A., & Romney, D. M. (2003). Risk and resiliency factors
King, D., Vogt, D., & King, L. (2004). Risk and resilience factors in posttraumatic stress disorder. Annals of General Hospital
in the etiology of chronic posttraumatic stress disorder. In Psychiatry, 2, 4–13.
B. Lits (Ed.), Early intervention for trauma and traumatic loss Wahl, O. (1995). Media madness. New Brunswick, NJ: Rutgers
(pp. 34–63). New York: Guilford Press. University Press Publishing.
Lawrie, S. M. (2000). Newspaper coverage of psychiatric and Weidmann, A., Fehm, L., & Fydrich, T. (2008). Covering the tsu-
physical illness. Psychiatric Bulletin, 24, 104–106. nami disaster: Subsequent post-traumatic and depressive symp-
Liberati, A., Altman, D. G., Tetzlaff, J., Mulrow, C., Gotzsche, P., toms and associated social factors. Stress Health, 24, 129–135.
Ioannidis, J. P. A., . . .Moher, D. (2009). The PRISMA state- Weidmann, A., & Papsdorf, J. (2010). Witnessing trauma in the
ment for reporting systematic reviews and meta-analyses of newsroom: Posttraumatic symptoms in television journalists
studies that evaluate health care interventions: Explanation and exposed to violent news clips. Journal of Nervous & Mental
elaboration. Journal of Clinical Epidemiology, 62, 1–34. Disease, 198, 264–271.
McMahon, C. (2001). Covering disaster. A pilot study into second- Yankelovich, D. (1990). Public attitudes toward people with chronic
ary trauma for print media journalists reporting on disaster. mental illness: Final report. Princeton, NJ: Robert Wood John-
Australian Journal of Emergency Management, 16, 52–56. son Foundation Publishing.

Downloaded from isp.sagepub.com at PENNSYLVANIA STATE UNIV on September 15, 2016

14 International Journal of Social Psychiatry 0(0)

Appendix 1. Search strategy 25. cyclothymia

26. dysthymia
The following terms were included: 27. agoraphobia
28. phobias
01. journalist/s 29. anxiety disorder
02. journalism 30. mood disorder
03. news reporter 31. panic disorder
04. TV reporter 32. obsessive compulsive disorder
05. Media worker 33. adjustment disorder
06. 01 or 02 or 03 or 04 or 05 34. dissociative disorder
07. PTSD 35. conversion disorder
08. trauma 36. amnesia
09. mental health 37. dissociative
10. mental illness 38. somatoform disorder
11. mental disorder 39. hypochondriacal disorder
12. mentally ill 40. somatization disorder
13. schizophrenia 41. neurasthenia
14. schizophrenic 42. depersonalization
15. schizotypal 43. neurotic
16. delusional disorder 44. neurosis
17. paranoid 45. bipolar disorder
18. schizoaffective disorder 46. dependence
19. psychotic 47. misuse
20. psychosis 48. alcohol related disorder
21. depression 49. alcoholism
22. depressive 50. alcohol abuse
23. mania 51. 07 or 08 or 09 or….or 49 or 50
24. manic 52. 06 and 51

Downloaded from isp.sagepub.com at PENNSYLVANIA STATE UNIV on September 15, 2016