The socioeconomic patterning of alcohol misuse and mental disorder comorbidity

Text removed is struck through; text added is underlined. Unchanged paragraphs are summarised rather than repeated.

Objective for processing

It is estimated that 16.6% of adults in England drink hazardously, with [9 words unchanged] resulting in costs to the NHS, alcohol-related crime and loss in productivity. The alcohol harm paradox refers to the finding that lower Lower socioeconomic status (SES) groups have an increased risk of mortality resulting from alcohol attributable causes, yet report despite reporting lower levels of alcohol consumption. There are different patterns This phenomena is referred to as the alcohol harm paradox, which may be explained by their experience of alcohol use across SES groups with the prevalence of harmful alcohol use found to be higher in low SES groups but the reverse found for hazardous use. One explanation for these increased health harms in lower SES groups is that they experience multiple disadvantages, including an increased risk of poor mental health, whereby disadvantages. For example, research has found identified a higher prevalence of common mental disorder (CMD) in lower SES these groups. It is not known if The relationship between alcohol misuse causes worsening and mental health or if worsening is complex. Alcohol misuse exacerbates the symptoms of mental health leads illness. At the same time those suffering with psychological disorders often use alcohol to alcohol misuse. However, research relieve symptoms of distress. Research suggests that worsening mental health is likely to lead to an increase in using alcohol consumption, whereby alcohol is used as a maladaptive coping technique. technique explained by the self-medication theory. There is currently a lack of data as to whether the co-morbidity of alcohol misuse with mental disorder is more common in lower SES groups, but it groups. It may be hypothesised that individuals of lower SES with a mental disorder may be more likely to self-medicate due to limited resources to facilitate more effective coping, otherwise known as coping. Therefore, the self-medication theory. The aim of this research is to better understand the complex relationship between mental health and alcohol misuse across different socioeconomic groups using the Adult Psychiatric Morbidity Survey (APMS) uses robust measures (APMS), which is a nationally representative survey of England. This information may help to reduce socioeconomic inequalities by identifying how the risk factors (for example, poorer mental health) for alcohol use, mental health and SES, which enables people attributable disease in lower SES groups may differ as compared to be catergorised according to these measures and can be used to explore the social patterning of alcohol misuse and mental disorder. in high SES groups. This study is part of a PhD that is funded by the Society for the Study of Addiction; a charity that aims to add to and promote the scientific understanding of addiction, and advocate the use of evidence-base in policy and practice. The charity funds projects that align with these aims, therefore, they acknowledge the importance of this study and the likely benefits of reducing the implications of alcohol misuse and it’s associated harms that will come from understanding more about the socioeconomic patterning amongst alcohol misuse and mental disorders. The funders will be involved in providing financial support for this PhD but will not have access to the APMS dataset. The funders will receive yearly reports of the progress of the PhD, PhD thesis and any dissemination of analyses, such as publications and conference proceedings. The relationship between alcohol and mental health is further confounded as both heavy drinking and mental disorder increase the risk of developing a wide range of physical non-communicable diseases, with an established 10-year reduced life expectancy (Chang et al., 2011). Multimorbidity (MM) of alcohol use, mental disorder and physical non-communicable diseases places a substantial cost on the NHS and raises difficult questions about whether patients are best served by addressing the physical or mental symptoms first, or by attempting to change the underlying drinking behaviour. This form of MM is also likely to significantly contribute to socioeconomic inequalities in health, as levels of alcohol-related harm (Katikireddi et al., 2017), poor mental health and poor physical health (Foster et al., 2018) are all increased in lower SES groups. Inequalities in access to, and quality of, treatment will be important drivers of these effects. The APMS uses robust measures of alcohol use and mental health, and provides data on longstanding physical health conditions and SES. This enables the categorisation of people according to these measures and allows the exploration of the social patterning of alcohol misuse and mental disorder and MM with physical health conditions. The aims of this PhD are to i) advance the understanding of alcohol misuse and mental disorder co-morbidity, ii) study the socioeconomic patterning of this co-morbidity using national and regional databases, and iii) qualitatively examine the motivations and social drivers for alcohol use in individuals with a mental health problem. To meet the first two aims, University of Liverpool are requesting access to the Adult Psychiatric Morbidity Survey (released 2018). The University of Liverpool will be the data controller of this study. This work involves two linked programmes of work and this Agreement will permit the University of Liverpool to analyse the 2014 APMS data (released 2018) for Study 1 and Study 2. The University of Liverpool is the sole data controller and also processes the data for the purpose described in this Agreement. The first study aims to i) advance the understanding of alcohol misuse and mental disorder co-morbidity, ii) examine SES profiles of individuals with a mental disorder who are drinkers and non-drinkers, iii) examine whether symptoms of a mental disorder are associated with alcohol use, and iv) study the socioeconomic patterning of this co-morbidity using national and regional databases. This study is part of a PhD that is funded by the Society for the Study of Addiction (SSA); a charity that aims to add to and promote the scientific understanding of addiction, and advocate the use of evidence-base in policy and practice. The charity funds projects that align with these aims, therefore, they acknowledge the importance of this study and the likely benefits of reducing the implications of alcohol misuse and its associated harms that will come from understanding more about the socioeconomic patterning amongst alcohol misuse and mental disorders. The funders will be involved in providing financial support for this PhD but will not have access to the APMS dataset. The funders will receive yearly reports of the progress of the PhD, PhD thesis and any dissemination of analyses, such as publications and conference proceedings. The second study will follow on from Study 1 which focused only on the comorbidity of different mental disorders and alcohol use in the APMS dataset, to also examine physical health conditions and determine if the MM (i.e. clustering) of physical conditions differs if someone has a mental disorder and/or is drinking harmfully. The second study has been funded by the Medical Research Council as part of a consolidator award. As with Study 1, the funders will not have access to the APMS dataset. This will allow the researcher to better understand how the health inequalities relating to alcohol attributable disease are greater for individuals with, compared to without, a mental disorder. This research (including Study 1 and 2) will meet Article 6 1(e) of the 2018 General Data Protection Regulations (GDPR) in analysing the data for ‘’public task’. This is justified because the data was originally collected through a population health survey in which the aim was to provide a better understanding of the health and wellbeing of the English population. Given the implications of alcohol misuse, poor health and the Government’s commitment to reducing alcohol harms (House of Commons, 2017), this research will identify groups of individuals at most risk of having a co-occurring alcohol and mental health problem which will help to tailor interventions more appropriately and reduce alcohol harms. This study will also determine whether the physical harms of alcohol are greater in individuals with a mental health problem and this information would be required to make recommendations to policy makers and commissioners that alcohol screening should be focused in individuals with a mental health problem. In addition, participants consented to take part in the data collection and for the future analyses of their data by researchers at academic and public institutions. The participants were aware when participating that the planned analyses would aim to provide a better understanding of the mental health of the general population in England. This research is also covered under Article 9 2(j) in using special category data (specifically on ethnic origin and data concerning health) for ‘reasons of substantial public interest’ under the substantial public interest condition point 13) ‘Journalism, academia, art and literature’. This work meets the criteria of substantial public interest because it aims to improve health and social care in the area of mental health care by analysing data that is representative of England. Specifically this work will identify which groups with a mental disorder are most likely to use alcohol harmfully in order to ensure that treatment and interventions around alcohol use can be targeted to the appropriate individuals and to ensure that there are not treatment gaps in current provision. [1 paragraph unchanged] The In Study 1 the APMS data will be utilised to i) identify the prevalence of non-drinking, hazardous and harmful alcohol use in individuals who meet the criteria for a mental health disorder, ii) understand how co-morbidity differs by SES, iii) investigate whether the SES profiles of groups explains the higher prevalence of non-drinking and harmful alcohol misuse in those in receipt of benefits is explained by having among individuals who meet criteria for a CMD, and mental disorder, iv) examine the SES predictors whether symptoms of alcohol and a mental disorder co-morbidity. are associated with alcohol use, and v) examine whether non-drinking and harmful drinking and meeting criteria for a mental disorder is more common in different ethnic groups. In Study 2 the APMS data will be utilised to understand i) how physical health conditions cluster with mental disorder and ii) to identify if the most common physical MM differ between individuals who meet criteria for a) a mental disorder, b) harmful alcohol use, c) comorbid mental disorder and harmful drinking or d) neither. [1 paragraph unchanged] The APMS collected interview data to monitor mental health and well-being in [28 words unchanged] mental health problems, and according to specific mental health problems (e.g. depression, post-traumatic posttraumatic stress disorder). Hazardous drinking is defined as consuming more than 14 units [21 words unchanged] is a pattern of alcohol consumption that causes mental or physical damage. Researchers will use self-report measures of alcohol use (Alcohol Use Disorder Identification Test (AUDIT)), mental health problems (e.g. Clinical Interview Schedule-Revised (CIS-R)) and SES (CIS-R)),SES (e.g. educational attainment and household income) and ethnicity (e.g. White British, Black British) collected by the APMS to categorise cases according to level of drinking (hazardous/harmful) (non-drinking/hazardous/harmful) and the presence or absence of a mental disorder (yes/no). For example, [21 words unchanged] to 4. Researchers will use individual’s total scores to categorise cases of non-drinking (score of 0) hazardous (score of 8-15) or harmful (16+) drinkers. drinkers as well as their scores on each AUDIT subscale (AUDIT consumption, AUDIT dependence and AUDIT harm). Another example is the use of the CIS-R to identify a CMD. [20 words unchanged] more defines a case of any CMD. Using cases of any CMD, Researchers researchers will then calculate the symptom score from the CIS-R to categorise individuals [17 words unchanged] for other mental health measures to identify cases of specific mental disorders. Researchers will also use psychiatric symptoms of a CMD measured by the CIS-R (e.g. somatic, irritability) to examine associations with AUDIT subscales. Based upon the calculation of cases of alcohol misuse and mental disorder, researchers will examine the associations between the prevalence of hazardous/harmful alcohol use (as categorised by total AUDIT score) in those with and without each of the specific mental disorders outcomes (as categorised by scores according to the specific mental disorder measure), for example post-traumatic stress disorder, CMD and personality disorder. The University of Liverpool will use approaches, such as logistic regression models, to examine mental health case status (yes/no) as a predictor of hazardous and harmful alcohol use separately. The University of Liverpool will also look at SES by developing a composite measure using indicators such as educational attainment, household income, employment, receipt of benefits and social class. Odds ratios and 95% confidence intervals will be reported for all analyses. The University of Liverpool will not link the APMS data. Study 1: This work is in the public interest as it aims to improve health and social care by analysing data that is representative of England to identify groups at most risk of misusing alcohol and having a mental disorder. This will have implications on how interventions can be tailored more appropriately to meet the needs of these groups and ensure that those in most need are targeted through interventions. Based upon the calculation of cases of alcohol misuse and mental disorder, researchers will examine the associations between the prevalence of non-drinking/hazardous/harmful alcohol use (as categorised by total AUDIT score) in those with and without each of the specific mental disorders outcomes (as categorised by scores according to the specific mental disorder measure), for example post-traumatic stress disorder, CMD and personality disorder. The University of Liverpool will use approaches, such as logistic regression models, to examine mental health case status (yes/no) as a predictor of non-drinking, hazardous and harmful alcohol use separately. The University of Liverpool will use advanced statistical methods to classify participants into SES categories using multiple indicators such as educational attainment, household income, employment, receipt of benefits and social class, the role of neighbourhood characteristics and social capital, and the subsequent associations of these SES categories with mental health and non-drinking, hazardous and harmful alcohol use. The University of Liverpool will also look at associations of symptoms of CMD (as categorised by scores for each symptom to meet criteria for a CMD) with alcohol use (as categorised by scores on each AUDIT subscale). Finally, the University of Liverpool will examine the associations between mental health status and non-drinking, hazardous and harmful alcohol use, across the different ethnic groups categorised in both the 2007 and 2014 dataset. Odds ratios and 95% confidence intervals will be reported for all analyses. The University of Liverpool will not link the APMS data. The University of Liverpool has determined that there are no moral or ethical issues from dissemination of data for this purpose. NHS Digital will pseudonymise the data before it is disseminated and the University of Liverpool will therefore be compliant with the ICO's "Anonymisation: managing data protection risk" code of practice. Study 2: The APMS includes data on a list of 22 physical health conditions that occurred since 16 years (such as high blood pressure, cancer and liver problems) including information on the timing of the onset of the disorder and whether the individual is receiving any medication for this disorder. This work will examine how common these different physical health conditions are and will additionally compare the prevalence of these 22 conditions across four groups, dependent on whether an individual meets criteria a) a mental disorder, b) harmful alcohol use, c) comorbid mental disorder and harmful drinking or d) neither (with all individuals assigned to a single category). The work will also examine how the 22 health conditions cluster using statistical techniques such as latent class analysis, looking for patterns of how these conditions co-occur within individuals and then whether these clusters may differ when individuals are categorised based upon their mental health and alcohol use comorbidity category. Further analyses will examine how self-reported health service use (defined in previous APMS work as i) no treatment, ii) medication only, iii) counselling or therapy only or iv) both medication and counselling) differs based upon their mental health and alcohol use comorbidity category and whether the self-reported impact of having a particular health condition on functioning differs based upon their mental health and alcohol use comorbidity category. Data summary work will include descriptive tests and group comparisons (e.g. chi-square) of the three defined groups. This will lead into latent class analysis to derive the most common clusters of conditions, stratifying individuals by the four categories. The University of Liverpool have considered the moral and ethical issues and believed that these are outweighed by the benefits of this research. NHS Digital will pseudonymise the data before it is disseminated and the University of Liverpool will therefore be compliant with the ICO's "Anonymisation: managing data protection risk" code of practice.

Processing activities

The APMS dataset will be received in a pseudo-anonymised pseudonymised form so there will be no storage of identifiable data at any point. Data will be stored securely on personal password protected and encrypted drives on the secure IT server at the University of Liverpool, which can be accessed on campus and externally through a VPN. The personal drives on the server can only be accessed only by authorised personnel. the study team, who are employed by the University of Liverpool and have been appropriately trained in data protection and confidentiality. Data will be backed up on the University of Liverpool’s Active Data [20 words unchanged] will be no requirement nor attempt to reidentify individuals from the data. For study one, alcohol use will be derived from participants’ AUDIT score and categorised as non-drinkers, low risk, hazardous drinkers and harmful drinkers while mental disorders will be derived from screening questions used to assess each mental disorder. The researchers will use approaches, such as regression models, to examine mental health case status (yes/no) as a predictor of alcohol use. Researchers will then use advanced statistical methods to classify participants into SES categories using multiple indicators of SES, such as educational attainment and employment, the role of neighbourhood characteristics, social support and social capital to examine the subsequent associations with mental health and alcohol use. The researchers will also look at associations of symptoms of CMD (characterised by scores of each symptom) with alcohol use. Finally, the researchers at the University of Liverpool will examine associations between alcohol use and mental health status across different ethnic groups using both the 2007 and 2014 APMS dataset. Odds ratios and 95% confidence intervals will be reported for all analyses. This will allow the researcher to answer the aims for Study 1 in understanding how hazardous and harmful alcohol use may be more common in individuals with specific mental health problems and additionally whether these associations differ across SES groups. For Study 2 the researchers at the University of Liverpool will examine how the 22 self-reported physical health conditions cluster using statistical techniques such as latent class analysis, looking for patterns of how these conditions co-occur within individuals and then whether these clusters may differ when individuals are categorised based upon their mental health and alcohol use comorbidity category. Data summary work will include descriptive tests and group comparisons (e.g. chi-square) of the three defined groups. This will lead into latent class analysis to derive the most common clusters of conditions, stratifying individuals by the four categories. This will allow researchers to answer the aims for Study 2 in determining whether alcohol attributable diseases are more common in individuals with specific mental health problems and if individuals who have comorbid alcohol and mental health problems are more likely to report a greater number of physical health conditions which are attributable to alcohol. [1 paragraph unchanged] The 2014 APMS dataset is held on behalf of NHS Digital by [36 words unchanged] will be able to download the dataset from UKDS for the period specific specified within the Data Sharing Agreement (DSA) and they must securely destroy the [23 words unchanged] years in accordance with the University of Liverpool's Research Data Management Policy. [1 paragraph unchanged] All organisations party to this agreement must comply with the Data Sharing [11 words unchanged] that use) by “Personnel” (as defined within the Data Sharing Framework Contract ie: - i.e. employees, agents and contractors of the Data Recipient who may have access to that data). [2 paragraphs unchanged]

Expected output

[1 paragraph unchanged] An academic paper describing the socioeconomic patterning of the comorbidity of hazardous and harmful alcohol with mental disorders will be submitted to Addiction in 2019. An initial academic paper describing the socioeconomic patterning of the comorbidity of hazardous and harmful alcohol with mental disorders will be ready for submission to the peer-reviewed high impact journal, for example Drug and Alcohol Dependence, in 2020 and as part of a press release at the University of Liverpool, with additional academic papers published from Study 1 across 2020/21 (journals to be selected). Submission was originally scheduled for 2019, however this was delayed due to data cleaning, analysis and interpretation of findings taking longer than expected. Dependant on the results of the research, iIt is anticipated that these findings will also presented at an academic conference focusing on the use of big data for epidemiology. Further academic papers will be produced and published in peer-reviewed journals based upon the questions from study 2 on the MM of harmful alcohol use with mental disorders and physical health conditions (for example, to the journal Addiction) to be published in 2021. Booklets providing a summary of the research and the implications will be distributed to University of Liverpool, NHS England, Department of Health and third sector organisations through existing University of Liverpool links. The research project will be presented at the SSA Annual Conference 2020 (5-6 November, 2020; online). Dependant on the results of the research, it is anticipated that these findings will also be presented at an academic conference focusing on the use of big data for epidemiology in the future (for example, at the EPA Section in Epidemiology and Social Psychiatry in 2021). The University of Liverpool will hold a public engagement event towards the end of the project to share findings with relevant charities and NHS services, a lay booklet presenting these findings will also be distributed at this event. Booklets providing a summary of the research projects and the implications will be distributed to University of Liverpool, NHS England, Department of Health and third sector organisations through existing University of Liverpool links. At each 12 months, a report of the projects progress and findings will be compiled and passed onto the funders, the Society for the Study of Addiction, including towards the end of the project where findings from this study will form part of a PhD thesis (October 2019, October 2020, October 2021). The University of Liverpool will hold public engagement events towards the end of the project to share findings with relevant charities and NHS services, a lay booklet presenting these findings will also be distributed at this event. At each 12 months, a report of the projects progress and findings will be compiled and passed onto the funders of study 1 (SSA), including towards the end of the project where findings from this study will form part of a PhD thesis. This report will not be published as its aim is to determine whether the PhD is progressing at a reasonable level. The research findings of Study 2 may be publicly disseminated by the Medical Research Council on their website and in lay summaries in 2021. The research findings of both studies will also be used to support future grant applications in the area of alcohol and mental health comorbidity in 2021 onwards.

Expected measurable benefits

University of Liverpool aim to advance the understanding of alcohol misuse and mental disorder co-morbidity. Mental disorder has been found to be highly co-morbid with alcohol misuse, however, it is not known how this co-occurs across the spectrum of mental disorders. The direction of the association between alcohol misuse and mental disorders is also not well known, however, research suggests that worsening mental health is likely to lead to an increase in alcohol consumption, whereby alcohol is used as a maladaptive coping technique. This research will analyse data that is representative of England to identify mental disorders that are more likely to co-occur with alcohol misuse. Through this research, the University of Liverpool will aim to advance the understanding of alcohol misuse and mental disorder co-morbidity and inform the development of Study 2. This research will provide evidence on a number of gaps in the current literature on alcohol misuse and mental disorder comorbidity and will outline the benefits to health and social care. The research for Study 1 will also contribute to PhD work and so will support the academic development of the PhD student. This study also aims to identify those most vulnerable to alcohol harms. Those from lower socioeconomic status (SES) groups have been found to experience more alcohol-attributable harms and have a common mental disorder. There is currently a lack of data as to whether the co-morbidity of alcohol misuse with mental disorder is more common in lower SES groups, but it may be hypothesised that individuals of lower SES with a mental disorder may be more likely to self-medicate due to limited resources to facilitate more effective coping. This research will use data representative of England to identify groups at risk of experiencing alcohol-attributable harms and worsen symptoms of a mental disorder. The specific research aims and the linked benefits will now be outlined: A project steering group will be established using existing links. This will be formed of representatives from NHS England, Department of Health and mental health charities in the North West of England. The group will be regularly briefed on the progress and asked to provide guidance to ensure that the approach is sensible and relevant. 1) To analyse data that is representative of England to identify which mental disorders are more likely to co-occur with alcohol misuse. These findings can be used by health professionals providing care to individuals with a mental disorder to understand which of their patients are most likely to experience alcohol harms and to provide better and more tailored alcohol screening and referrals as part of their care. This work is close to completion in 2020 and once published the academic findings will be publicised through a press release and the paper will be circulated to relevant organisations, such as NHS England, who can act on the evidence, for example in informing future changes to the Quality and Outcomes Framework. This will benefit the wider population by helping to reduce the £21 billion cost of alcohol to society and the rates of hazardous drinking in England which is currently estimated at 16.6% through impacting on policy decisions around alcohol screening for individuals with a mental disorder and highlighting a need to implement more tailored interventions. 2) To identify which groups with a mental disorder are most vulnerable to alcohol harms. There is currently a lack of data as to whether the co-morbidity of alcohol misuse with mental disorder is more common in lower SES groups and in particular ethnic groups. Though it is known that some ethnic groups are more likely to experience alcohol harms and poor mental health, compared to those of White ethnicity. For example, 10.2% of Black ethnic groups drink harmfully compared to 8.2% of White British groups. It may be hypothesised that individuals of lower SES, and of Black, Asian and Minority ethnic groups, with a mental disorder may be more likely to self-medicate due to their experiences of disadvantage and discrimination resulting in limited resources to facilitate more effective coping. This research will use data representative of England to identify groups at risk of experiencing alcohol-attributable harms and through dissemination to organisations such as NHS England and the Mental Health Foundation this can inform targeted alcohol screening and provision of intervention to those groups most in need and reduce the current rates of co-occurring alcohol and mental disorders. 3) This work will explore the physical harms in individuals with alcohol and mental health problems. Individuals with alcohol and mental health MM do not currently receive the quality of healthcare they should with comorbidity often acting as a barrier to treatment. Therefore, as a result of the proposed work around MM the researchers will highlight the need to better integrate mental health and alcohol services in the UK by demonstrating that alcohol attributable conditions are more common in individuals with a mental disorder. The researchers will additionally identify the most commonly present physical health conditions based on an individual’s previous mental health and alcohol use and what conditions their GP should be screening for in order to both provide a more preventive approach to avoid the onset of these conditions through preventive treatments (for example, statins, hypertensive medication) and to additionally ensure that diagnosis and treatment is provided in a timely manner. These changes will result in cost savings to the NHS by reducing the prevalence of alcohol attributable diseases in individuals with a mental health problem. A project steering group for both studies 1 and 2 will be established using existing links. This will be formed of representatives from NHS England, Department of Health and mental health charities in the North West of England. The group will be regularly briefed on the progress and asked to provide guidance to ensure that the approach is sensible and relevant. These organisations will be the most appropriate to act on the research recommendations to implement changes in policy to improve health and social care for individuals with comorbid alcohol misuse and mental disorder.

Benefits reported

Yielded Benefits is not a requirement for new applications. Since 2018, the University of Liverpool has preliminarily identified that those with a mental disorder are more likely to be either non-drinkers, hazardous or harmful drinkers compared to those without a mental disorder. Researchers have presented their findings to an associated project steering group, including organisations from Alcohol Change UK. In addition, researchers have presented the preliminary findings at the Society for the Study of Addiction conference on 7th November 2019 as part of a poster presentation. Study 1 findings were also presented to the Medical Research Council in justifying the public need for Study 2.