Maya Semrau and her colleagues at the Institute of Psychiatry, Psychology and Neuroscience, London, have written a very elegant and helpful overview of the current knowledge about mental illness stigma and anti-stigma interventions in low- and middle-income countries (LAMICs). The paper is rooted in the authors’ strong interest in global mental health (Thornicroft & Tansella, Reference Thornicroft and Tansella2013; Votruba et al. Reference Votruba, Eaton, Prince and Thornicroft2014) as well as in recent reviews (Thornicroft et al. Reference Thornicroft, Mehta, Clement, Evans-Lacko, Doherty, Rose, Koschorke, Shidhaye, O'Reilly and Henderson2015) and original articles (Koschorke et al. Reference Koschorke, Padmavati, Kumar, Cohen, Weiss, Chatterjee, Pereira, Naik, John, Dabholkar, Balaji, Chavan, Varghese, Thara, Thornicroft and Patel2014) on mental illness stigma in LAMICs. Reading the paper by Semrau and colleagues we were impressed by the cultural differences to be taken into account in order to understand and address stigma in LAMICs as well as by the scarcity of available data, a consequence of difficult conditions for intervention research in poorer countries. Fortunately, the latter has begun to change.

One can think about stigma as an overarching concept consisting of biased stereotypes or ignorance; of prejudice or emotional negative reactions; and of discriminating behaviour. There is also the related concept of mental health literacy, referring to knowledge about mental illness, including available treatments (Jorm, Reference Jorm2012). These concepts may be generalised to low-income settings, but what we refer to as correct knowledge v. biased stereotypes may differ according to local circumstances. For example, where mental health services are not available, more sceptical public views about outcomes and treatability of severe mental illness may reflect the reality in those countries.

Multiple discriminations due to mental illness and to membership in another stigmatised group may be even more common in LAMICs as compared with high-income countries. Therefore many people with mental illness in LAMICs will have to struggle with multiple discriminations. Sometimes mental illness stigma may not be the most harmful one. Consider the example of Saudi Arabia (though not a LAMIC) and driving rights of women. There a woman with mental illness will not be allowed to drive – a restriction often cited as an example of structural discrimination. But in that case, gender-related discrimination trumps mental illness stigma. Another example would be gay men with mental illness in Uganda. Therefore anti-stigma interventions will have to choose their targets carefully and may often have to address multiple discriminations.

Individual poverty is an aspect of social exclusion and often a source of mental health problems (Hudson, Reference Hudson2005). Health care systems with less financial resources tend to have a reduced quantity and quality of mental health services. Different aspects of stigma, including structural discrimination such as poor funding of mental health care, are barriers to service use (Rüsch & Thornicroft, Reference Rüsch and Thornicroft2014; Clement et al. Reference Clement, Schauman, Graham, Maggioni, Evans-Lacko, Bezborodovs, Morgan, Rüsch, Brown and Thornicroft2015). It could be speculated that in rich countries with low structural barriers, internal aspects of stigma would be more relevant obstacles to mental health service use. Consistent with this assumption we found in a recent study in Switzerland that shame, an emotional aspect of self-stigma (Rüsch et al. Reference Rüsch, Hölzer, Hermann, Schramm, Jacob, Bohus, Lieb and Corrigan2006a ), as well as reduced subjective mental health literacy predicted more negative attitudes towards mental health care (Rüsch et al. Reference Rüsch, Müller, Ajdacic-Gross, Rodgers, Corrigan and Rössler2014b ). In poorer countries, on the other hand, external aspects of stigma might be more relevant barriers. That said, Mirja Koschorke and colleagues (Koschorke et al., Reference Koschorke, Padmavati, Kumar, Cohen, Weiss, Chatterjee, Pereira, Naik, John, Dabholkar, Balaji, Chavan, Varghese, Thara, Thornicroft and Patel2014) in their important study stress the fact that not only external consequences of stigma (such as, anticipated mental health-related discrimination with respect to employment or marriage) are a burden for individuals living with schizophrenia in India, but also internal factors, e.g. low self-esteem and feelings of inferiority.

Worldwide, millions of people immigrate from poorer into richer countries, and rates of severe mental illness persist to be higher among second-generation immigrants (Bourque et al. Reference Bourque, van der Ven and Malla2011). Thus the findings summarised by Semrau and colleagues are relevant for our understanding of mental illness stigma not only in LAMICs, but equally so in high-income countries (Rüsch et al. Reference Rüsch, Todd, Bodenhausen and Corrigan2010). In Camberwell, South London, where Maya Semrau and colleagues wrote their paper, mental health service providers are likely to be more effective clinicians if they are well accustomed with attitudes towards people with mental illness in African and Caribbean countries (or with attitudes towards people with mental illness having migrated from or with their cultural background in those countries); otherwise they may have difficulties to understand patients and their families. Given the size of current immigration movements, the same likely applies throughout Western Europe.

Whether or not to disclose one's mental illness is often a tough choice for people with mental illness. Non-disclosure or secrecy can protect against discrimination, but may be stressful on the long run with harmful consequences such as unemployment and social isolation (Link et al. Reference Link, Mirotznik and Cullen1991). Disclosure on the other hand may reduce self-stigma as well as public stigma but may go along with other short- and long-term risks, e.g. labour market discrimination (Corrigan et al. Reference Corrigan, Kosyluk and Rüsch2013). Therefore recent research has examined the efficacy of interventions that support people with mental illness in their disclosure decisions (Corrigan et al. submitted; Henderson et al. Reference Henderson, Brohan, Clement, Williams, Lassman, Schauman, Dockery, Farrelly, Murray, Murphy, Slade and Thornicroft2013; Rüsch et al. Reference Rüsch, Abbruzzese, Hagedorn, Hartenhauer, Kaufmann, Curschellas, Ventling, Zuaboni, Bridler, Olschewski, Kawohl, Rössler, Kleim and Corrigan2014a ). Previous studies examined cultural differences with respect to disclosure of a mental illness between Western and non-Western countries (Weiss et al. Reference Weiss, Jadhav, Raguram, Vaunatsou and Littlewood2001), and in India the discomfort to disclose seems to be particularly strong in the context of marriage (Koschorke et al. Reference Koschorke, Padmavati, Kumar, Cohen, Weiss, Chatterjee, Pereira, Naik, John, Dabholkar, Balaji, Chavan, Varghese, Thara, Thornicroft and Patel2014). Therefore it appears to be an open question what interventions to support people with mental illness with respect to partnership/marriage decisions should look like in LAMICs (while there is little doubt that adequate solutions will vary across LAMICs).

Finally, the authors discuss a few intriguing problems about the measurement of stigma. This is not a technical side issue: in order to tackle stigma effectively, we first need to understand its quantity and quality. There are at least two problems here: the need to adapt measures and their content to the local culture; and the issue of perceived legitimacy of discrimination. Regarding the adaptation of stigma measures we agree with the authors and the ‘what matters most’ approach (Yang et al. Reference Yang, Thornicroft, Alvarado, Vega and Link2014) that this adaptation is helpful. However there may be a downside, at least for quantitative studies, that findings will be more difficult to compare across settings, including the efficacy of anti-stigma interventions.

The second measurement issue is thoughtfully discussed by Mirja Koschorke and her colleagues (Koschorke et al., Reference Koschorke, Padmavati, Kumar, Cohen, Weiss, Chatterjee, Pereira, Naik, John, Dabholkar, Balaji, Chavan, Varghese, Thara, Thornicroft and Patel2014). If participants with mental illness respond in an interview or questionnaire that they have not experienced unfair discrimination, this may not indicate the absence of discrimination but possibly just their internalised view that they do not deserve any better treatment. High perceived legitimacy of discrimination is associated with self-stigma (Rüsch et al. Reference Rüsch, Lieb, Bohus and Corrigan2006b ), and studies using the framework of system justification theory provide strong evidence that members of disadvantaged groups tend to perceive the social order as fair, including their stigmatised status (Jost et al. Reference Jost, Banaji and Nosek2004), which in turn undermines their willingness to fight stigma (Jost et al. Reference Jost, Chaikalis-Petritsis, Abrams, Sidanius, van der Toorn and Bratt2012). This has at least two consequences for stigma research. First, results based on responses of people with mental illness whether they were treated unfairly may seriously underestimate the frequency and severity of discrimination. Second, we should strive to measure real-world behaviour in order to estimate discrimination, e.g. whether or not employers decide to hire people with mental illness.

In summary, the paper by Maya Semrau and her colleagues stimulates the discussion in the field and challenges our thinking about anti-stigma work, both locally and internationally. Following previous calls for research on stigma from an international perspective (Weiss & Ramakrishna, Reference Weiss and Ramakrishna2006) the authors make a convincing case that we need to get better at developing and evaluating anti-stigma interventions both in high-income settings and in LAMICs. As shown by Maya Semrau and colleagues there remains a lot to be done.