1

Citation: Evans-Lacko S, Corker E., Williams P, Henderson C, Thornicroft G. Effect of the Time to Change anti-stigma campaign on trends in mental-illness-related public stigma among the English population in 2003–13: an analysis of survey data. Lancet Psychiatry. 2014. 1 (2): 121-28

Link to article online:

http://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(14)70243-3/fulltext

Effect of the Time to Change anti-stigma campaign on trends in mental-illness-related public stigma among the English population in 2003–13: an analysis of survey data

Sara Evans-Lacko, Elizabeth Corker, Paul Williams,

Claire Henderson*, Graham Thornicroft*

*authors contributed equally to the work

Health Service and Population Research Department, Institute of Psychiatry, King's College

London, UK

Corresponding Author Full Contact Details:

Sara Evans-Lacko, PhD

King’s College London

Institute of Psychiatry

Health Service and Population Research Department

Box P029

De Crespigny Park

2

London SE5 8AF

United Kingdom

Telephone: + 44 (0) 20 7848 5076

Facsimile: + 44 (0) 20 7277 1462

Email: Sara.Evans-Lacko@kcl.ac.uk

Author Information:

Elizabeth Corker, MSc

King's College London

Institute of Psychiatry

Health Service and Population Research Department

Box P029

De Crespigny Park

London SE5 8AF

United Kingdom

Paul Williams, MPH

King's College London

Institute of Psychiatry

Health Service and Population Research Department

Box P029

De Crespigny Park

London SE5 8AF

United Kingdom

Claire Henderson, PhD

King's College London

Institute of Psychiatry

3

Health Service and Population Research Department

Box P029

De Crespigny Park

London SE5 8AF

United Kingdom

Graham Thornicroft, PhD

King's College London

Institute of Psychiatry

Health Service and Population Research Department

Box P029

De Crespigny Park

London SE5 8AF

United Kingdom

4

Abstract

Background: Given the consequences of public stigma for people personally affected by

mental health problems, it is important to understand trends and also effective mechanisms

which may reduce stigma and discrimination.

Methods: We investigated ten-year trends in public attitudes across England before and

during the Time to Change anti-stigma campaign, and whether there is a dose-effect

response between campaign awareness and regional outcomes.

Findings: There was a significant increase in positive attitudes following the Time to Change

campaign. Our findings demonstrated a significant increase in relation to prejudice /

exclusion following the launch of Time to Change (B=0.02 95% CI: 0.01, 0.05), though not for

tolerance and support for community care (B=0.01 95% CI: -0.01, 0.03). We also found

evidence for a dose-effect relationship between campaign awareness and regional

improvement in knowledge and attitudes, but not intended behaviour.

Interpretation: The positive effects of Time to Change appear significant and moderate.

Although attitudes may be more at risk of deterioration during times of economic hardship,

anti-stigma programmes may still play an active role in long-term reduction of stigma and

discrimination especially in relation to prejudice / exclusion of people with mental health

problems.

Funding: SEL, EC, PW and CH were supported by a grant to Time to Change from the

Government Department of Health in England and Comic Relief. CH and GT were funded by

National Institute for Health Research (NIHR) Programme Grant for Applied Research (RP-

PG-0606-1053) (PI: GT) awarded to the South London and Maudsley NHS Foundation Trust.

5

Introduction

Public stigma and discrimination in relation to mental illness, i.e., the mental health related

knowledge, attitudes or behaviour, of the general public can have important consequences

for people personally affected by mental health problems. There is a growing evidence base

for wide ranging and significant consequences of public stigma for individuals with mental

illness. Recent research provides evidence linking the degree of public stigma with self-

stigma among people with mental health problems,1 disadvantage in relation to

employment,2 higher prevalence of suicide,3 lower perceived health status4 and lower rates

of help-seeking for mental health problems5, all of which shape and maintain health

inequalities.6 Given the persistence of public stigma over recent decades, in the UK7 and

abroad,8,9 there is also, a recognised need for effective interventions to reduce public

stigma.10

The Time to Change anti-stigma campaign is England’s largest programme ever aimed at

reducing stigma and discrimination against people with mental health problems. Evaluation

of phase one (2009-2012) of the Time to Change anti-stigma and social marketing campaign

showed that important gains were made in terms of reducing public stigma and

discrimination in England. Specifically, improvement was shown in intended behavior

among the general public i.e., increased willingness to live with, work with, live nearby or

continue a relationship with someone with a mental health problem11 and reduction in

experiences of discrimination among service users.12,13 Moreover findings suggested that

the campaign represented a potentially cost-effective and low cost intervention, especially

when taking into account effects on the wider economy such as service utilisation and

employment.14 Overall the campaign was associated with modest but significant changes in

public stigma. The evaluation was limited, however, in that it included only one year of

6

baseline data and thus, was not able to establish a context for public attitude trends before

the launch of the campaign.

This paper sets out to improve on our understanding of public attitude trends in England in

three ways: (i) we examine the longer-term trends in population attitudes beginning in 2003

(six years before the initiation of the Time to Change anti-stigma campaign in 2009); (ii) we

investigate whether the time trend has changed significantly after the launch of the

campaign; and (iii) we examine whether there is a dose-effect relationship whereby regions

in England with higher campaign awareness also demonstrate greater improvement in

mental health related knowledge, attitudes and intended behaviour.

Methods

Data source

We used data from the 2003 and 2007-2013 national Attitudes to Mental Illness survey (the

survey was not carried out during the years 2004-2006 and so data are not available for

these years). This is an Omnibus Survey commissioned by the Department of Health and

carried out among a nationally representative sample of adults residing in England. There

were approximately 1700 respondents surveyed each year. Assessment of mental health

related attitudes was performed at each year; however, assessment of mental health

related knowledge and intended behaviour did not begin until 2009, in line with the

evaluation of the Time to Change anti-stigma campaign. The survey is carried out using a

quota sample, with sample points selected by a random location methodology. Census small

area statistics and the Postcode Address File were used to define sample points which were

stratified by Government Office Region and social grade. Interviews were carried out in the

participants' home by fully trained personnel using computer assisted personal interviewing

7

and demographic information was collected at the end of the interview. Additional

information regarding the survey methods can be found at:

https://catalogue.ic.nhs.uk/publications/mental-health/legislation/atti-ment-illn-2011/atti-ment-

illn-2011-apx.pdf. The study was classified as exempt by the King’s College London Psychiatry,

Nursing and Midwifery Research Ethics Subcommittee.

Measures

Mental health related knowledge

Mental health related knowledge was measured by the Mental Health Knowledge Schedule

[MAKS]).15 The MAKS comprises 6 items covering stigma-related mental health knowledge

areas: help seeking, recognition, support, employment, treatment, and recovery, and 6

items that inquire about classification of various conditions as mental illnesses.16 Overall

test-retest reliability of the MAKS is 0∙71 (Lin’s concordance statistic) and the overall internal

consistency among items is 0∙65 (Cronbach’s alpha).15 The MAKS intentionally includes items

of a multidimensional structure which investigate different types of mental health related

knowledge and this is reflected in an alpha score which is somewhat low by conventional

standards.15 The MAKS was not developed to function as a scale; however, it provides an

indication of overall trends in knowledge. The total score was calculated so that higher

MAKS scores indicate greater knowledge.

Mental health related attitudes.

8

The survey includes 26 attitude items derived from the Community Attitudes toward the

Mentally Ill scales (CAMI)17 and an added item on employment-related attitudes. A two-

factor structure has been established for this scale:18 one factor comprises items related to

prejudice and exclusion (M = 2∙1, SD = 0∙7; Cronbach’s alpha 0∙84) and the other factor

comprises items about tolerance and support for community care (M = 4∙0, SD = 0∙5;

Cronbach’s alpha 0∙78). Items were rated from 1 (strong disagreement) to 5 (strong

agreement). The total score for each subscale was calculated so that higher scores indicate

more positive and / or less stigmatizing attitudes.

Mental health related intended behaviour

Intended behaviour (the level of intended future contact with people with mental health

problems) was measured by the Reported and Intended Behaviour Scale (RIBS).19 The scale

assesses intended behavior across four domains: living with, working with, living nearby and

continuing a relationship with someone with a mental health problem. Assessment of

intended behaviour allowed for brief and feasible evaluation at the population level. Overall

test-retest reliability of the RIBS is 0∙75 (Lin’s concordance statistic). The overall internal

consistency of the scale is 0∙85 (Cronbach’s alpha). The total score was calculated so that

higher RIBS scores indicate higher levels of intended contact.

Socioeconomic status

Socioeconomic status of the respondent was categorized into one of four categories (i.e.,

AB, C1, C2 and DE) according to the Market Research Society’s classification system.

9

Classification was based on the occupation of the chief income earner in the household.

Category AB represents individuals with professional/managerial occupations, C1 represents

individuals with other non-manual occupations, C2 represents individuals having skilled

manual occupations and DE represents individuals with semi-/unskilled manual occupations

and people dependent on state benefits.

Regional prevalence of respondents reporting campaign awareness and familiarity with the

Time to Change campaign

Prompted awareness of the Time to Change campaign was assessed in 2012 and 2013.

Participants were first shown several pictures from different Time to Change attitudes that

appeared on television, radio, magazines or on the web. Overall awareness was then

determined via the following item: Do you think you have seen this advertising, or similar

during the last year? (Mental health advertising stills shown). Individuals who reported ‘yes’

or’ yes, seen similar’ were categorised as campaign aware while those who responded ‘no’

or ‘don’t know’ were categorised as ‘not campaign aware’. Respondents who reported some

campaign awareness were then asked: ‘How many times, before this interview, have you

seen or heard ANY of the advertising in the pictures’. Response options included: (i) Once or

twice, (ii) Three to five times and (iii) Six times or more.

We then calculated regional prevalence of respondents reporting overall awareness of the

Time to Change campaign. We also created an index of familiarity with the campaign (0=no

awareness, 1=seen/heard advertising once or twice, 2=seen/heard advertising three to five

times and 3= seen/heard advertising six times or more). Based on this index, we calculated

mean regional familiarity with the Time to Change campaign based on the mean score by

region.

10

Regional unemployment

Rates of unemployment among adults by region were obtained from the Labour Force

Survey commissioned by the UK Office for National Statistics20 and we look specifically at

change in regional unemployment rate. In this case ‘region’ refers to nine former

government regions. These regions were established across England in 1994. They reflect

administrative boundaries and to date are the primary classification for the presentation of

regional statistics used by the Office for National Statistics. Some research suggests that

negative attitudes towards minority groups, including people with mental health problems

may be associated with economic hardship, especially where there is greater competition

for employment.2,21 There was a differential impact across regions of England in relation to

the macroeconomic recession, with London experiencing less change in unemployment

rates compared to the North of England. As the macroeconomic recession coincided with

the Time to Change campaign and because changes in regional unemployment rates could

have been associated with changes in stigma outcomes, change in regional unemployment

rate was included as a covariate in the regression model when examining the relationship

between regional campaign awareness changes in mental health related knowledge,

attitudes and intended behavior.

Statistical analysis

We present the annual mean item score of the English population for attitude items related

to (i) prejudice and exclusion and (ii) tolerance and support for community care. We also

present an extrapolated linear trend line for the years 2009-2013 which project what

11

population attitude scores may have been without the Time to Change campaign. The trend

is based on earlier time points continuing along the same trend. Unadjusted and adjusted

linear regression models examine factors associated with (i) prejudice / exclusion and (ii)

tolerance and support for community care. Both attitude outcomes variables were

standardised according to the z-score. Independent variables included: gender, age

(categorical: 16-24, 25-44, 45-64 and 65+). To examine whether there was a change over

time, we included year as a continuous variable. For these models, we did not impute

attitude data for years where data were not collected. We assigned each year a value in

relation to the number of years following baseline data collection. To investigate whether

there was a change in attitude score after relative to before the campaign, we included a

binary variable pre / post. In order to better understand whether the time trend had changed

after relative to before the campaign, we also included an interaction term between year

and pre / post.

In addition to examining trends over time at the national level, to better understand if there

was a dose effect relationship between campaign awareness and reduction in stigma we

investigated factors associated with regional improvements in mental health related

knowledge, attitudes and intended behaviour, within each geographical region of England

(i.e., North East, North West, Yorkshire and Humberside, East Midlands, West Midlands,

East of England, London and South East). We examined changes in stigma between 2009

and 2013. Multivariable models were fitted to data aggregated by region. Multivariable

linear regression models adjusted by region were used to estimate the mean change in

standardised knowledge, attitudes (prejudice / exclusion and tolerance / support for

community care) and intended behaviour scores by region. Estimates and 95 percent

12

confidence intervals are presented for each of the outcomes. Model covariates included:

campaign awareness, regional unemployment rate and within region baseline knowledge,

attitude and intended behaviour score. Survey sampling weights were applied in all analyses

so that respondents reflected a nationally representative sample in terms of socio-

demographic characteristics within each region of England. For all regression models, we

formally tested whether the data met the regression assumptions. We tested for normality

of residuals and heteroscedasticity using scatterplots and quantile-quantile plots. To check

for multicollinearity, we examined tolerance (reciprocal of the variance inflation factor) and

we did not identify any values less than 0.1. We also examined non-linearity by plotting the

independent vs. dependent variable and found no indication of violation of the

assumptions. Akaike’s Information Criteria (AIC) was used to evaluate the goodness of fit of

the regression models. Analyses were carried out using SAS version 9.3.

Results

Table 1 provides a summary of survey respondent characteristics in terms of gender, age,

and socioeconomic group, by survey year.

TABLE 1 ABOUT HERE

Figure 1 displays the mean weighted population attitude scores for the prejudice / exclusion

factor and the tolerance for community care factor. Actual population attitudes from 2003-

2013 are depicted alongside simulated linear trend lines for the years 2009-2013 which

project what population attitude scores may have been without the Time to Change

campaign.

13

FIGURE 1 ABOUT HERE

Table 2 describes factors associated with more positive attitudes regarding; (i) tolerance and

(ii) prejudice and / or exclusion. The unadjusted analyses suggest that attitudes related to

prejudice / exclusion are becoming more positive over time (year) at 0∙014 standard

deviations per year (p <0∙001). The analysis also shows that attitudes were significantly

more positive after the Time to Change campaign compared to before the campaign (p

<0∙01). In the multivariable analysis, we also investigated the interaction between year and

pre / post (i.e., before and after the Time to Change campaign) to see whether the change

was significantly more positive following the campaign compared to before the campaign

was launched and the interaction term was positive and significant (p <0∙05). Additional

factors associated with more positive attitudes in relation to prejudice / exclusion include:

female gender, higher social class and being less than 65 years of age.

Attitude trends regarding tolerance, were similar to those related to prejudice / exclusion,

though the adjusted interaction term for time by pre / post was not significant in the

multivariable analysis (p=0∙27). The unadjusted analyses suggest that attitudes related to

tolerance improved over time, to a slightly greater degree compared to prejudice /

exclusion , i.e., 0∙021 standard deviations per year (p<0∙001) and that attitudes regarding

tolerance were significantly more positive after the Time to Change campaign (p<0∙01).

Additional characteristics associated with greater tolerance include: female gender, being of

middle age (i.e., 45-64 compared to 65+) and higher socioeconomic status. Individuals aged

16-44 were significantly less tolerant compared to individuals greater than or equal to 65

years of age.

14

TABLE 2 ABOUT HERE

Table 3 describes factors associated with regional changes in mental health related

knowledge, attitudes and intended behavior (during the years 2009-2013), specifically

addressing the question of whether there is a ‘dose effect’ relationship between campaign

awareness and intended reduction in stigma. In the multivariable models, level of regional

campaign awareness was associated with increased knowledge (p =0∙004) and tolerance

(p<0∙0001) and decreased prejudice (p =0∙001); but, not intended behaviour (p =0∙20). All

models controlled for baseline stigma (mean knowledge, tolerance, prejudice and intended

behaviour score per region in 2009). This was significant in all models suggesting that those

regions with lower levels of tolerance at baseline (in 2009) had greater increases in

tolerance over the period of the campaign (i.e., 2009-2013). We also examined how change

in unemployment rate (over the years 2007 to 2013) was related to change in stigma

outcomes. In the unadjusted models, an increase in unemployment rates was associated

with less positive changes in knowledge, prejudice and intended behaviour scores; however,

these relationships were not significant in the multivariable model.

TABLE 3 ABOUT HERE

We further investigated the relationship between campaign awareness and intended

reduction in stigma by looking at an index of regional familiarity with the Time to Change

campaign in relation to stigma outcomes. In the multivariable regression models, level of

regional campaign familiarity was associated with increased knowledge (z-score estimate =

0.58, p =0∙001), increased tolerance (z-score estimate = 0.58, p<0∙0001), and lower levels of

prejudice (z-score estimate = 0.51, p =0∙008). There was, however, no significant association

with intended behaviour (z-score estimate = 0∙32, p =0∙26).

15

Discussion

This study investigated trends in mental health related knowledge, attitudes and intended

behavior among a representative sample of the English population. We were able to look at

trends in population attitudes beginning in 2003 (six years before the initiation of the Time

to Change anti-stigma campaign in 2009). Our findings suggest that there was a step change

increase and also a change in the slope associated with prejudice / exclusion attitudes

following the launch of the Time to Change campaign, though this trend was not significant

for tolerance and support for community care attitudes. This follows an earlier trend of

decline in positive attitudes in England during the decade before 2003.7 We also identified

factors associated with regional variation throughout England for changes in knowledge,

attitudes and intended behavior from 2009 to 2013. These analyses support the hypothesis

of a dose-effect relationship between campaign awareness (regional prevalence of

respondents reporting awareness of the Time to Change campaign) and regional increases

in knowledge, prejudice / exclusion related attitudes and tolerance and support for

community care; but, not intended behavior.

In term of long terms trends, our findings suggest that population attitudes related to

prejudice / exclusion demonstrated a greater positive change following the Time to Change

campaign compared to those related to tolerance and support for community care (which

did not demonstrate a significant difference). This suggests that the campaign may have

been better at disconfirming negative ideas around prejudice and exclusion rather than

enhancing positive attitudes and support.18 The value of long term data, establishing

baseline trends and investigation of contextual factors is further emphasized in the context

16

of previous findings which suggested a decline in positive attitudes in England from 1993-

2003 which may have been influenced by implementation of mental health reforms.7 It is

possible that other factors in the sociopolitical context which coincided with the launch of

the Time to Change campaign contributed to the differential change or mediated the

influence of the campaign; however, given that the launch of Time to Change coincided with

an economic recession, it is not likely that this would have been associated with a natural

increase in positive attitudes.2,22,23 This, however, does not eliminate the possibility that

other factors may have mediated or influenced changes in public stigma. There was a

significant positive change in tolerance and support for community care attitudes between

2008 and 2009 and it is unclear what precipitated this change. There were community level

activities including: Get Moving! A Mass Participation Event and local and regional grass

roots projects mainly run by service users such as ‘Open Up’, which were launched during

2008 as a part of Time to Change. These, however, were small scale projects and thus it is

not likely that they account for the entire change. Stephen Fry also released a documentary

‘The Secret Life of the Manic Depressive’ which gained significant popularity and won an

Emmy award for the best documentary in 2007. It is possible that these factors contributed

toward the change, but these conclusions are post-hoc and so we cannot be certain about

the specific event or combination of events which influenced this change in attitudes related

to tolerance.

Similar to other studies5,17 this research showed that females and individuals from a higher

socio-economic group tended to have greater tolerance / support for community care and

less prejudice / exclusion compared to males, though to a greater extent for attitudes

17

related to prejudice / exclusion than those related to tolerance / support for community

care. Relationship between age and attitudes differed for the two constructs. Attitudes

related to tolerance / support for community care showed a u-shaped curve (with those

aged 45-64 having the best attitudes) and those related to prejudice / exclusion showed a

more linear trend with attitudes improving as age increased, except for those greater than

65 years; however, we cannot be certain whether these are due to fluctuations in attitudes

over the life span or a cohort effect.

The findings also suggest that there was a dose response relationship between campaign

exposure and stigma outcome in those regions which achieved greater awareness of the

Time to Change campaign also demonstrated greater positive increases in knowledge and

attitudes, when controlling for baseline stigma score and unemployment rates. This

suggests that even at the aggregate level, increased campaign activity was associated with

greater positive increases in knowledge and attitudes. It could be that increased campaign

activity and awareness / engagement could facilitate a type of herd immunity in which

communities become protected against stigma with increasing exposure. This dose

response relationship, however, was not evident for the intended behaviour outcome.

Although previous findings suggested that there was a positive and significant relationship

between awareness of the Time to Change campaign and intended behaviour score24 and a

significant overall positive increases in intended behavior from 2009 to 2012,11 a dose effect

relationship, in terms of campaign awareness at the regional level was not evident in these

data. It could be that we did not have sufficient data points to identify a relationship here.

However, it could also be that level of awareness of the social marketing campaign is less

important for improving population intended behavior compared to other types of more

18

active campaign engagement. Changes at the community level are likely to result both from

direct exposure to campaign media but also indirect effects facilitated by e.g., increasing

discussion and openness around mental health problems in the community. We were not

able to measure ‘grassroots activity’ at the local level or amount of ‘face-to-face contact’

which occurred. This is important as it has been suggested that this type of activity is the

most relevant when trying to achieve meaningful behaviour change.25

Strengths / Limitations

We had several years of data from a large nationally representative sample of individuals,

including population stigma indicators before and after the launch of the Time to Change

anti-stigma campaign which allowed us to establish a trend and to infer whether the pattern

changed significantly following the launch of the campaign. Although we collected data on

several stigma related outcomes between 2009 and 2013, only attitude data were available

before 2009 and these data were not collected annually. Additionally, campaign awareness

was not available before 2011 and so we cannot track this over time alongside the

indicators. Intensity of campaign activity varied throughout England and we were able to

look at the dose response relationship by region; however, we could not investigate

anything smaller than regional level. Moreover, there were no regions without anti-stigma

campaign activity which could act as a control group. Identifying an appropriate, non-

exposed control group is a significant challenge when evaluating such large scale national

interventions. Smaller scale controlled experiments would allow for a control group but

these would not tell us about the campaign’s potential to work in real life as a national

programme and the indirect effects which may result in relation to the intervention.

19

Although the lack of a control group makes it difficult to know to what extent changes are

attributable to the Time to Change campaign, we have used various methods, for example,

time trend and dose effect relationship, to triangulate our findings. We have also accounted

for secular trends, such as changes in unemployment rates associated with the economic

recession in the analysis to the extent possible. We assessed different individuals at each

time point to establish population trends, but were not able to look at change among the

same individuals over time. These data however, do represent an indicator, from the

perspective of the population. As all data were self-report data, it is possible that social

desirability influenced responses.26

Although our data comprise several different attitude statements, we do not look at how

each attitude changed over time. We focus on attitude trends at an aggregate level and so

we don’t report on how, for example, views about dangerousness have changed over time.

We do, however, explore specific attitudes in relation to the Time to Change social

marketing campaign elsewhere.24 Our data do not allow us to differentiate stigma outcomes

by type of mental illness. This is a limitation as attitudes may vary according to condition; for

instance, levels of stigma are generally lower for depression compared to schizophrenia.27

Additional research suggests that the general public associates the term 'mental illness' with

severe conditions28 and that variations in the conceptualisation of mental illness are related

to likelihood of disclosure and seeking help for a mental health problem.29 Future work

should consider how to better understand these variations in types of stigma and how these

may be addressed and evaluated in anti-stigma interventions.30;31

Conclusions

20

The effects of the Time to Change campaign to reduce stigma and discrimination among the

English adult population appear to be significant but moderate. It is encouraging that

attitudes continued to become more positive during a period of macroeconomic recession

and at a time when there is little evidence for positive increases elsewhere.9 Although

attitudes may be more at risk of decline during times of hardship, social marketing and anti-

stigma activity may provide an opportunity to reduce stigma and discrimination. Future

studies would benefit from more directly tying population indicators related to stigma and

discrimination against people with mental health problems to social, health and economic

consequences for people with mental health problems.1,6,12 and to investigate both direct

and indirect pathways to change in public stigma.

Panel: Research in context

Systematic Review

To identify scientific papers which have investigated trends in public stigma against people

with mental health problems and effectiveness of population level anti-stigma

interventions, we searched Medline and Web of Knowledge with search terms including:

public attitudes, mental health literacy, stigmatization, anti-stigma campaigns.

Interpretation

This study provides new evidence on trends in public attitudes in relation to people with

mental health problems across England over ten years. Findings suggest a step change

increase and also a change in the slope associated with prejudice / exclusion attitudes

following the launch of the Time to Change anti-stigma campaign, though this trend was not

21

significant for tolerance and support for community care related attitudes. This follows an

earlier trend of decline in positive attitudes in England during the decade before 2003. We

also found evidence for a dose-effect relationship between campaign awareness and

regional positive increases in knowledge and attitudes, but not intended behaviour. The

positive effects of Time to Change appear significant and moderate. Although attitudes may

be more at risk of deterioration during times of economic hardship, anti-stigma

programmes may still play an active role in long-term reduction of stigma and

discrimination.

Acknowledgments

We are grateful for the feedback and collaboration on the evaluation from Sarah Hamilton

and Vanessa Pinfold, both based at The McPin Foundation. We would also like to thank Sue

Baker, Paul Farmer, Maggie Gibbons and Paul Corry, for their collaboration.

22

Figure 1. Population attitude trends across England 2003 - 2013

Note: Estimates for population attitude scores for the years 2004, 2005 and 2006 are

based on a linear trend based on the closest year that data were available i.e., years

2003 and 2007.

* The prejudice / exclusion subscale was reverse coded so that all results could be

calculated in a way where going up indicated more positive and / or less stigmatizing

attitudes.

3.6

3.8

4.0

4.2

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

Prejudice / exclusion - Reverse coded*

Tolerance / support for community care

Prejudice / exclusion - Reverse coded* (simulated with no campaign)

Tolerance / support for community care (simulated with no campaign)

Official launch of Time to Change anti-

stigma social marketing campaign

23

Table 1. Participant characteristics stratified by survey year (unweighted frequency and weighted percentage)

2003 n=1897

n (%)

2007 n=1729

n (%)

2008 n=1703

n (%)

2009 n=1751

n (%)

2010 n=1745

n (%)

2011 n=1741

n (%)

2012 n=1717

n (%)

2013 n=1727

n (%)

Gender n (%)

Female 864 (48.8) 798 (48.7) 778 (48.3) 939 (51.5) 939 (51.7) 912 (51.5) 924 (51.3) 801 (49.0)

Male 1033 (51.3) 931 (51.3) 925 (51.7) 812 (48.5) 806 (48.3) 829 (48.5) 793 (48.7) 926 (51.0)

Age n (%)

16-24 275 (13.7) 206 (13.9) 188 (13.2) 247 (14.3) 240 (14.1) 322 (20.5) 258 (14.6) 289 (14.6)

25-44 693 (37.8) 581 (34.8) 562 (36.0) 633 (35.9) 540 (35.1) 450 (29.8) 580 (34.8) 568 (36.1)

45-64 543 (29.4) 526 (31.2) 525 (32.1) 512 (31.3) 549 (31.5) 340 (21.1) 506 (31.3) 486 (31.0)

65+ 386 (19.1) 416 (20.1) 428 (18.7) 359 (18.5) 416 (19.4) 629 (28.6) 373 (19.3) 384 (18.3)

Socioeconomic

status n (%)

AB 262 (16.1) 342 (21.3) 315 (21.0) 279 (19.4) 300 (20.2) 322 (20.5) 292 (19.3) 302 (20.5)

C1 499 (30.0) 448 (28.8) 433 (29.6) 454 (32.2) 464 (31.7) 450 (29.8) 456 (31.0) 445 (30.4)

C2 434 (23.0) 409 (20.6) 363 (21.1) 389 (20.8) 342 (19.2) 340 (21.1) 368 (21.6) 362 (20.8)

DE 702 (30.9) 530 (29.2) 592 (28.2) 629 (27.6) 639 (28.8) 629 (28.6) 601 (28.1) 618 (28.4)

24

Table 2. Attitude trends among the English population before and after the Time to Change anti-stigma campaign 2004-20131

Total tolerance/support for community care

attitude score (z-score, unadjusted)

Total tolerance/support for community care

attitude score (z-score, adjusted)

Total prejudice/ exclusion attitudes-

reverse codeda (z-score, unadjusted)

Total prejudice/ exclusion attitudes

reverse codeda (z-score, adjusted)

Gender Female Male

***0∙14 (0∙10, 0∙17)

Reference

***0∙14 (0∙11, 0∙18)

Reference

***0∙20 (0∙17, 0∙23)

Reference

***0∙22 (0∙19, 0∙25)

Reference

Age 16-24 25-44 45-64 65+

***-0∙35 (-0∙41, -0∙30)

***-0∙14, (-0∙18, -0∙09) ***0∙11 (0∙07, 0∙16)

Reference

***-0∙35 (-0∙40, -0∙29) ***-0∙14 (-0∙19, -0∙10)

***0∙10 (0∙06, 0∙15) Reference

**0∙09 (0.03, 0∙14)

***0∙23 (0∙19, 0∙28) ***0∙41 (0∙36, 0∙45)

Reference

***0∙10 (0∙05, 0∙16) ***0∙22 (0∙18, 0∙26) ***0∙39 (0∙35, 0∙44)

Reference

Class AB C1 C2 DE

***0∙24 (0∙19, 0∙29) ***0∙11 (0∙07, 0∙15) ***0∙08 (0∙04, 0∙13)

Reference

***0∙23 (0∙18, 0∙28) ***0∙14 (0∙10, 0∙18) ***0∙11 (0∙06, 0∙15)

Reference

***0∙55 (0∙50, 0∙59) ***0∙34 (0∙30, 0∙38) ***0∙16 (0∙12, 0∙21)

Reference

0∙54 (0∙49, 0∙59)

***0∙34 (0∙30, 0∙38) ***0∙17 (0∙13, 0∙21)

Reference

Year (continuous) ***0∙021 (0∙016, 0∙026) 0∙02 (0∙01, 0∙04) ***0∙014 (0∙01, 0∙02) *0∙03 (0∙01, 0∙05)

Pre/Post TTC2 Post Pre

***0∙12 (0∙09, 0∙16)

Reference

0∙03 (-0∙15, 0∙21)

Reference

**0∙07 (0∙04, 0∙10)

Reference

*0∙16 (-0∙02, 0∙34)

Reference

Year*pre/post -- 0∙01 (-0∙01, 0∙03) -- *0∙02 (0∙01, 0∙05)

*p<0∙05, **p<0∙01, ***p<0∙001

1 All results are given in the form of standardized regression coefficients with 95% confidence intervals in parenthesis.

2 The additional effect (mean difference) between pre and post at initiation of Time to Change (i.e. 2009)

a The prejudice / exclusion subscale was reverse coded so that all results could be calculated in a way where going up indicated more positive and / or less

stigmatizing attitudes.

25

Table 3. Factors associated with regional improvements in knowledge, attitude and

intended behaviour scores following the launch of the Time to Change campaign, 2009-

2013 (linear regression) 1

Change in knowledge score (z-score, unadjusted)

Change in knowledge score (z-score, adjusted)

Campaign awareness achieved in 2013

0∙002 (-0∙01, 0∙02)

**0∙01 (0∙004, 0∙02)

Baseline knowledge score **-0∙03 (-0∙05, -0∙002) ***-0∙03 (-0∙04, -0∙02)

Change in unemployment 2009-2013

*-0∙13 (-0∙27, 0∙01)

-0∙01 (-0∙07, 0∙05)

Change in Tolerance/ support for community

care attitudes (z-score, unadjusted)

Change in Tolerance/ support for community

care attitudes (z-score, adjusted)

Campaign awareness achieved in 2013

0∙004 (-0∙01, 0∙02)

***0∙01 (0∙007, 0∙013)

Baseline tolerance/ support for community care score

**-0∙08 (-0∙13, -0∙03) ***-0∙09 (-0∙11, -0∙08)

Change in unemployment 2009-2013

-0∙13 (-0∙29, 0∙03)

0∙02 (-0∙02, 0∙05)

Change in Prejudice/ Exclusion attitudes

(z-score, unadjusted)

Change in Prejudice/ Exclusion attitudes (z-score, adjusted)

Campaign awareness achieved in 2013

-0∙0004 (-0∙02, 0∙01)

*0∙01 (0∙001, 0∙02)

Baseline prejudice/ exclusion score

**-0∙04 (-0∙08, -0∙004) **-0∙05 (-0∙07, -0∙02)

Change in unemployment 2009-2013

*-0∙13 (-0∙25, -0∙02) -0∙03 (-0∙10, 0∙03)

Change in intended behaviour

(z-score, unadjusted)

Change in intended behaviour

(z-score, adjusted)

Campaign awareness achieved in 2013

-0∙01 (-0∙02, 0∙01)

0∙007 (-0∙004, 0∙02)

Baseline intended behaviour score

**-0∙17 (-0∙25, -0∙08) *-0∙20 (-0∙35, 0∙06)

Change in unemployment 2009-2013

*-0∙16 (-0∙28, -0∙05)

0∙02 (-0∙13, 0∙17)

* p<0∙05, ** p<0∙01, *** p<0∙001

1 All results are given in the form of standardized regression coefficients with 95%

confidence intervals in parenthesis.

26

Reference List

(1) Evans-Lacko S, Brohan E, Mojtabai R, Thornicroft G. Association between public views of mental illness and self-stigma among individuals with mental illness in 14 European countries. Psychol Med 2012 August;42(8):1741-52.

(2) Evans-Lacko S, Knapp M, McCrone P, Thornicroft G, Mojtabai R. The mental health consequences of the recession: economic hardship and employment of people with mental health problems in 27 European countries. PLoS One 2013;8(7):e69792.

(3) Schomerus G, Evans-Lacko S, Rusch N, Mojtabai R, Angermeyer MC, Thornicroft G. Collective levels of stigma and national suicide rates in 25 European countries. Epidemiol Psychiatr Sci 2014 February 27;1-6.

(4) Alvarez-Galvez J, Salvador-Carulla L. Perceived discrimination and self-rated health in europe: evidence from the European social survey (2010). PLoS One 2013;8(9):e74252.

(5) Mojtabai R. Mental illness stigma and willingness to seek mental health care in the European Union. Soc Psychiatry Psychiatr Epidemiol 2010 July;45(7):705-12.

(6) Hatzenbuehler ML, Phelan JC, Link BG. Stigma as a fundamental cause of population health inequalities. Am J Public Health 2013 May;103(5):813-21.

(7) Mehta N, Kassam A, Leese M, Butler G, Thornicroft G. Public attitudes towards people with mental illness in England and Scotland, 1994-2003. Br J Psychiatry 2009 March;194(3):278-84.

(8) Pescosolido BA, Martin JK, Long JS, Medina TR, Phelan JC, Link BG. "A disease like any other"? A decade of change in public reactions to schizophrenia, depression, and alcohol dependence. Am J Psychiatry 2010 November;167(11):1321-30.

(9) Schomerus G, Schwahn C, Holzinger A, Corrigan PW, Grabe HJ, Carta MG et al. Evolution of public attitudes about mental illness: a systematic review and meta-analysis. Acta Psychiatr Scand 2012 June;125(6):440-52.

(10) Evans-Lacko S, Courtin E, Fiorillo A, Knapp M, Luciano M, Park A et al. The state of the art in European research on reducing social exclusion and stigma related to mental health: a systematic mapping of the literature. European Psychiatry 2014.

(11) Evans-Lacko S, Henderson C, Thornicroft G. Public knowledge, attitudes and behaviour regarding people with mental illness in England 2009-2012. British Journal of Psychiatry 2012.

(12) Corker E, Hamilton S, Henderson C, Weeks C, Pinfold V, Rose D et al. Experiences of discrimination among people using mental health services in England 2008-2011. Br J Psychiatry Suppl 2013 April;55:s58-s63.

27

(13) Henderson C, Corker E, Lewis-Holmes E, Hamilton S, Flach C, Rose D et al. England's time to change antistigma campaign: one-year outcomes of service user-rated experiences of discrimination. Psychiatr Serv 2012;63(5):451-7.

(14) Evans-Lacko S, Henderson C, Thornicroft G, McCrone P. Economic evaluation of the anti-stigma social marketing campaign in England 2009-2011. Br J Psychiatry Suppl 2013 April;55:s95-101.

(15) Evans-Lacko S, Little K, Meltzer H, Rose D, Rhydderch D, Henderson C et al. Development and psychometric properties of the Mental Health Knowledge Schedule. Can J Psychiatry 2010 July;55(7):440-8.

(16) Rüsch N., Evans-Lacko S, Thornicroft G. What is a mental illness? Public views and their effects on attitudes and disclosure. Australian & New Zealand Journal of Psychiatry 2012.

(17) Taylor SM, Dear MJ. Scaling community attitudes toward the mentally ill. Schizophr Bull 1981;7(2):225-40.

(18) Rüsch N, Evans-Lacko SE, Henderson C, Flach C, Thornicroft G. Knowledge and attitudes as predictors of intentions to seek help for and disclose a mental illness. Psychiatr Serv 2011 June;62(6):675-8.

(19) Evans-Lacko S, Rose D, Little K, Flach C, Rhydderch D, Henderson C et al. Development and psychometric properties of the reported and intended behaviour scale (RIBS): a stigma-related behaviour measure. Epidemiol Psychiatr Sci 2011 September;20(3):263-71.

(20) Labour Force Survey. Office of National Statistics 2012 March;Available from: URL: http://www.statistics.gov.uk/hub/labour-market/people-not-in-work/unemployment

(21) Warner R. Recovery from schizophrenia: psychiatry and political economy. Third edition ed. Routledge; 2004.

(22) Wahlbeck K, McDaid D. Actions to alleviate the mental health impact of the economic crisis. World Psychiatry 2012 October;11(3):139-45.

(23) Warner R. Labor, Poverty and Schizophrenia. Recovery from Schizophrenia. 2004. p. 136-57.

(24) Evans-Lacko S, Malcolm E, West K, Rose D, London J, Rusch N et al. Influence of Time to Change's social marketing interventions on stigma in England 2009-2011. Br J Psychiatry Suppl 2013 April;55:s77-s88.

(25) Corrigan PW, Morris SB, Michaels PJ, Rafacz JD, Rusch N. Challenging the public stigma of mental illness: a meta-analysis of outcome studies. Psychiatr Serv 2012 October;63(10):963-73.

(26) Henderson C, Evans-Lacko S, Flach C, Thornicroft G. Responses to mental health stigma questions: the importance of social desirability and data collection method. Can J Psychiatry 2012 March;57(3):152-60.

(27) Pescosolido BA, Medina TR, Martin JK, Long JS. The "backbone" of stigma: identifying the global core of public prejudice associated with mental illness. Am J Public Health 2013 May;103(5):853-60.

28

(28) Jorm AF, Reavley NJ. Public confusion caused by differing understandings of the term 'mental illness'. Aust N Z J Psychiatry 2012 May;46(5):397-9.

(29) Rusch N, Evans-Lacko S, Thornicroft G. What is a mental illness? Public views and their effects on attitudes and disclosure. Aust N Z J Psychiatry 2012 July;46(7):641-50.

(30) Reavley NJ, Jorm AF. Stigmatising attitudes towards people with mental disorders: changes in Australia over 8 years. Psychiatry Res 2012 May 30;197(3):302-6.

(31) West K, Hewstone M, Holmes EA. Rethinking 'mental health stigma'. Eur J Public Health 2010 April;20(2):131-2.