 Fy enw i'n edrych am ymddangos, mae Mary Abed-Ahad oedd ymwyloedd St Andrews, mae'n schyffordi gyda hwnnw i'r gwaith yn ystod yn cyflawni gyda hwnnw o'r University of St Andrews, a mae'n cyffiliadoddodd iddo'r hwnnw, a'r Max Planck Institute o'r Rhon Ùf yn Gyffredinolol yng Nghymru. Her work examines the effective air pollution and weather changes on health, mental health, well being, mortality and hospital admission in the general population and biathเอinnig subgroups. So today she's going to present some of that work and I'm going to hand over straight to Mary so that she can start her presentation, over to you. Thank you so much, Dar sensing for the introduction so I just came out here right now. Let's hope it works. So, gweithgwnaeth yng Nghymru, mae'n mynd i am hwnnw, ynddo i'r newid yma ar gyfer gyda'r gyfyniadau sy'n gynghwm ymwhiffydd Andrews, ac ydy'r gweithigol yma yn ymddangos cymdeithas ymddangos am ymgyrch ac ymwneud hynny'n gweithio y mynd yn gweithio ymgyrch ac mae hynny'n ddechrau eich gennyrch yn y Cymru, oherwydd mae'n gweithio'n ddynnal yma ar y Cymru, oherwydd mae'n gweithio'n ddynnal ymdano. First, I will give a general introduction about the topic. Mental health problems are noticeably rising globally, and this is causing serious socio-economic losses to the societies. According to Rigo et al in 2016, the global burden of mental diseases was estimated at 32% of years lived with disability, and 13% of disability adjusted life years. Mental health problems are a serious issue. Mental health disorders are mainly triggered by genetics and or by psychosocial risk factors. However, recent literature has been showing a relationship also between environmental factors including exposure to ambient air pollution and mental well-being. However, most of this evidence is still fragmented and inconclusive. So how does now air pollution affect mental well-being? So it can affect it either directly or indirectly. So it can affect it directly through biological mechanisms. So a particulate matter of small diameters, of small diameters such as PM1 or PM2.5 might initiate oxidative stress and lead to the formation of inflammatory cytokines that infiltrate the blood-brain barrier, causing neurodegeneration and neuroinflammation. Also particulate matter and nitrogen dioxide pollutants have been linked by observational research to increase traits of several depressive disorders and mental well-being disorders such as autism spectrum, schizophrenia, dementia, psychotic experiences, cognitive disabilities and anxiety. Also air pollution can affect mental well-being indirectly and this could be through the nuisance and individual coping behaviour to deal with this air pollution and nuisance caused by it. So some of the air pollutants can cause a static or odorous nuisance and this will inhibit psychological supporting outdoor activities due to air pollution and this in turn can lead to cognitive anxiety, stress, loneliness which leads to general fatigue and perceived symptoms of poor mental well-being. And in a recent systematic literature review of 178 published articles it was shown that air pollution is to degrees happiness and life satisfaction substantially and also to increase anxiety annoyance and mental problems and even suicide ideation and also it leads to coping approaches such as avoidance behaviour and migration to an area with less air pollutants. Also air pollution can lead to experiential anxiety emerging from worrying feelings about one's physical health and future. So because there is more conclusive research about the effect of air pollution on physical health including cardiopulmonary, immune system and cancer diseases, people who live in highly polluted areas might also experience stress and worrisome feelings of getting physical illness of this air pollution and this also will impact their mental well-being. So despite the establishment of linkages between air pollution and mental well-being in the literature, the results are still inconclusive. Most of the studies were cross-sectional or longitudinal studies that lack spatial temporal specificity and lengthy follow-up times. And to date no study has tried to address the association between long-term air pollution exposure and mental well-being using a within between longitudinal design which I will explain in a bit in the methods section. And published research also have not yet covered all population types and the potential moderating effect that demographic groups might have on this association between air pollution and mental well-being. For example, only age and gender social demographics are considered by most of the literature. However, ethnicity is not being considered in this topic. Therefore examining how the effect of air pollution on mental well-being varies by ethnic groups such as this can provide us with more conclusive results. So this study aims to assess longitudinality overall and the between within effects of long-term which is 11 years exposure to four air pollution nitrogen dioxide, sulfur dioxide and particulate matter of 10 and 2.5 diameters. In the UK on individuals reported mental well-being which was measured using the 12 items general health questioner scale. And also we aim to assess whether ethnic minorities such as Pakistani and Bangladeshi Indians, Black African Caribbean and other ethnicities and also non-UK born individuals if they suffer from a more pronounced risk for mental well-being with increasing concentrations of the four poll users. So methods we use data from the UK household longitudinal understanding the society data and we had a sample of 60,146 adult individuals who provided 349,748 repeated responses across 10 waves of data collection which is equivalent to 11 years from 1090 to 1019. And this individual level data was linked to yearly concentrations of the four poll using the local authority of residents for each individual. Later we took the air pollution concentration and we decomposed it into between and within effects. So the between effect is the average 11 years of air pollution for each local authority. Whereas the within effect it is the annual air pollution deviation from the 11 years average for each local authority. And then we examined the association between air pollution and self-reported mental well-being in the general population and also by ethnic groups using three levels multi-level mixed effect linear models adjusting for important social demographics, cigarette smoking and gear dumps. Just to note that mental well-being was assessed using the general health questioner which has 12 items that assess that capture non-psychotic psychiatric illness and these 12 questions relate to mental well-being of the interviewed individuals in the past few weeks preceding the data collection and responses for each of these 12 questions are assessed on a four-point-like scale and then they are dichotomized and then they are summed up which results in a general score for the mental well-being that ranges from 0 to 12 with higher scores indicating poorer mental well-being. Now we come to results. So this table shows descriptive statistics for the first wave of the understanding society data and for the last wave. So we can see that the majority of the sample are females. They belong to the middle-aged group from 34 to 58 years old and the majority are British white and the other ethnicities, they are approximately 4%. So for example, Pakistan and Bangladesh, they are 3.5% in wave 1 and 4.5% in wave 10 of this data. The majority were born in the UK at 6% and around 14% were not born in UK. 53% were married. Around one-third of the sample had university degree and one-third had high school degree and around 60% of individuals said that they are living comfortably slash doing alright with respect to their financial situation. Majority also were non-smokers and only around 20% were smokers. This graph here also shows the air pollution levels across years from 2009 to 2019 and we can see fluctuations, some fluctuations in the air pollution terms. However, we can notice that in general air pollution degrees in recent years. So if we compare for example on 2009 to 2019 we can see for all the pollutants the level now is less than it used before. So now we come to the multi-level regression model results and here we are showing the overall pollution effect as well as the between effect which shows us the spatial effect of air pollution like the effect of living in more polluted local authorities versus not and the within pollution effect which is more temporary so it shows how the variation in air pollution across time affects the well-being within each local authority. So we can see that the overall pollution has positive effect on mental well-being which means as the concentration of all of these four pollutants increases the mental well-being will become poorer and we see a similar picture for the between spatial effect of air pollution however for the within effect which is more temporary we didn't see any significant associations. Now there's a graph here it shows how this association vary between ethnic groups and generally we don't see any conclusive evidence that the association between air pollution and mental well-being varies by ethnic groups. So it's just related to Pakistani slash Bangladeshi they are showing poorer mental well-being with increasing concentrations of particulate matter 2.5 and sulfur dioxide and only also for non-UK born individuals however it's not very conclusive and also when we assess the between and within effects of air pollution also we don't see any conclusive evidence. So concluding remarks using longitudinal individual level and contextual linked data this study highlighted the negative effect of air pollution on individuals mental well-being over time so as the concentrations of air pollution increases mental well-being will become poorer and significant overall and between effects like spatial between effects are spatial between local authorities we're shown for all the four pollutants so this shows that if you live in a more polluted local authority you will experience poorer mental well-being however like fluctuation of air pollution across time doesn't affect your mental well-being so much and analysis by ethnicity didn't show conclusive evidence it just showed elevated scores of poor mental well-being with increasing concentrations just of sulphur dioxide and particulate matter only for one ethnic group which is Pakistani and Bangladeshi and non-UK born individuals but not for the other ethnic groups so environmental policies to reduce air pollution emissions can eventually improve the mental well-being of people in the UK however we didn't find any conclusive evidence on the moderating effect of ethnicity for future research we recommend data that can be used to link air pollution to individual level data at a finer spatial resolution level because here we use the local authority this is what was available for us with this data set however for future we will be using another data set which is the Scottish Longitudinal Study which allows linkages at the postcode level of air pollution rather than using the local authority so we are hoping to see more interesting results there so thank you so much for listening for my presentation today and this is my email address in case anybody have any questions later and I'm looking forward for the discussion and questions okay we're going to go to our next presentation now so our next presenter today is Sophie Baker from Banga University so Sophie is in the final year of her PhD she's funded by the ESRC and she is based at Banga University so she's interested in the social determinants of psychosis in minority groups with a particular focus on linguistic minorities and her PhD is looking at the mental health of linguistic minorities in Wales using both qualitative and quantitative methods so over to you Sophie good afternoon everyone my name is Sophie and I'm an ESRC funded PhD student at Banga University working under the supervision of Dr Chris Savill and Dr Mike Jackson so today I'm just going to take you through some preliminary findings of one of my PhD studies which uses National Survey for Wales data to examine group density associations for mental health in language, national identity and ethnic groups in Wales so I'll just begin with a bit of background so mental illness is not equally distributed throughout the population rather our risk of developing a mental illness varies substantially by social group and geographic area so this is not a novel finding a seminal work by Farrison Dunham and Odegard as far back as the 1930s found higher rates of psychiatric admissions in migrants to the United States two conflicting explanations stem from these findings Farrison Dunham proposed a social causation hypothesis suggesting that living in socially adverse neighbourhoods is what drives the increased risk while Odegard thought the higher rates in migrants were due to selective migration in other words individuals who are more vulnerable to developing a mental illness are more likely to migrate or drift into these more deprived areas before the onset of their illness so since these early studies there have been significant statistical advancements which makes it easier to disentangle the individual and area level variables associated with mental illness one well replicated epidemiological finding is that some ethnic minority and migrant groups have poorer mental health than their majority group counterparts however the extent of this risk is somewhat contingent on the immediate area in which the minority group individual lives for example ethnic or group density associations have found that minority group individuals have better mental health when they live in areas where their group is well represented compared to minorities living in areas with a low proportion of their own group there is some suggestion that low own group density areas do not pose the same risk across different minority groups and risk is particularly marked in black individuals and to date most group density studies have examined these kinds of associations in ethnic minority and migrant groups so there are in terms of mechanisms there are material and psychological reasons that might help to explain the poor mental health observed in some ethnic minority and migrant groups material processes refer to factors restricting an individual's access to resources networks and a voice to kind of gain control over their life and personal circumstances for example discriminatory behaviour or policy directed towards ethnic minorities that restricts their access to opportunities and resources and sustains an equal power balance between the minority and majority group psychological processes refer more to the mental consequences of belonging to a group that one perceives as lower status it is thought that ethnic minority in itself is not a risk factor for mental health rather belonging to a disempowered or marginalised group the evidence base is limited but some studies have found that these density relationships extend to other types of minorities such as those classified by political affiliation lower social class and sexual minority status to identify mechanisms involved in the group density effect it is therefore theoretically interesting to examine group differences in the effect including in minorities classified by other types of social characteristics language is quite interesting in this regard because language is different in terms of power and status and it's quite a salient marker of in group identity language can of course also present other more practical challenges for individuals who are not able to communicate in a widely spoken language in their local area which might pose problems we have accessed in social capital for example national identity is of interest because I guess it's kind of a classic us versus them variable it's also quite interesting in the sense that it's more of a flexible identity compared to ethnic group and language status so in short examining associations in these different types of groups can help provide clues as to the important social process involved in the risk of lower own group density areas Wales is a good social context within which to examine these group density effects so Welsh speaking ability is very socially salient in Wales Welsh is an official language of Wales but is only spoken by about 90% of the population however if we consider this at a smaller neighbourhood level Welsh speakers actually comprise the majority in many areas so most Welsh speakers are bilingual Welsh and English speakers so English speakers who do not speak Welsh are more likely to be linguistically isolated in their area in the sense of them not being able to communicate in a widely spoken language where they live these maps show units of geography at the lower super output level at the LSOA level which are units of geography comprising about 1500 people the map on the left shows more Welsh speaking areas in the darker blue colours so as you can see most Welsh speaking areas are situated in the west of the country but there is high variation in Welsh speaking when we look at this small neighbourhood level the second map indicates more Welsh identifying areas the proportions people selected that they have a Welsh national identity in the census so as you can see there's a lot of highly Welsh identifying areas and this is particularly high in the south of the country in Cardiff and in the south Wales valleys the third map shows the geographical distribution of non-white British ethnic minorities and the fourth map indicates area deprivation as measured by the Welsh index of deprivation so the darker colours on the map indicate higher levels of deprivation so many group density studies are conducted in large urban cities often in England and the Netherlands so Wales is quite a different study setting it's more rural and for example if we were to look at these kinds of density effects in London for example ethnic minority position and linguistic minority status perhaps more conflated so given that it seems plausible that mechanisms involved in the ethnic density effect might extend to other types of identities we expect to see group density associations for all three identity types we expect to see a group density relationship for ethnic minorities in line with many previous studies but we also expect these effects to extend to language and national identity so for our data set we obtained a licence from the Welsh Government to obtain the LSOAs for individual respondents this was merged with area level data from the 2011 UK census we combined the data from the 2012 to 13 13 to 14 and 14 to 15 waves of survey so we had a sample size of 28,248 people we derived a continuous mental health dependent variable from a factor analysis of four mental health related outcomes in the surveys so as an example for the anxiety variable respondents were asked to rate how anxious they felt yesterday on a 10 point scale for our analysis we used the GLM TMB package in R we fitted mixed effects models to test the presence of group density associations for each of three identity types so the first model comprised of the interaction between the individual and the area level variable as fixed effects, for example the interaction between Welsh speaking status and the LSOA level proportion of Welsh speakers the basic model also included sampling weights from the survey and a random intercept to account for nesting within the different LSOAs the full models are adjusted for all individual and area level covariance that are outlined here so in terms of our results at the individual level proficient Welsh speakers were defined as individuals who specified that they were proficient in Welsh or could speak a fair amount of Welsh and everyone else was coded as a normal speaker there were 4361 proficient Welsh speakers in the sample and the rest were non Welsh speakers so just on the left on the regression plot here you can see proficient Welsh speakers are the blue line and non Welsh speakers are the red line so as you can see the more Welsh speaking areas were associated with better mental health for both groups which isn't expected because we didn't expect to see this in the non Welsh speakers in more Welsh speaking areas the forest plot on the right is a visualisation of the fully adjusted mix effects model so the purple indicates variables with a protective relationship of mental health and the green indicates detrimental associations so as you can see the fully adjusted model indicated a significant main effect of area level Welsh speaking such that the more Welsh speaking areas were associated with better mental health however inconsistent with our hypothesis we found no evidence of an interaction between individual level Welsh speaking status and the area level proportion of Welsh speaking speakers after adjustment for covariates that's just circled at the bottom here as for national identity nearly 18,000 people indicated that they had Welsh national identity and no Welsh national identity so again on the plot on the left people with a Welsh national identity are the blue line and no Welsh national identity is the red line so more Welsh identifying areas were associated with poor mental health however after adjusting for the covariates in the full model there was no main effect of area level national identity and no evidence of an interaction between individual national identity so again inconsistent with our hypothesis we found no indication of a group density association for national identity for these preliminary analysis we conducted quite a crude group density analysis for ethnic groups in Wales so we examined the association between ethnic group and area level proportion of non-white British ethnic minorities just to say you've got two minutes left so the ethnic minority samples were small so there was the biggest group was white although with 588 people then there was 277 Asian individuals 150 with a mixed ethnic background 90 from another ethnic background group and 77 black individuals so as you can see mental health sharply decreases as ethnic density increases in the white British group in the ethnic minority groups there seems to be a protective association in black individuals but the relationships seem to vary more for the other groups again on the right we can see the fully adjusted model so more highly ethnic density areas were associated with poor mental health but then when we looked at the association between individual level ethnic minority status and area level ethnic density for all ethnic groups and significantly so for white of the black and Asian groups which is consistent with our hypothesis so we found no evidence of a group density association for language and national identity but we did find evidence of an association for ethnic groups so just in terms of some strengths and limitations overall we had a large sample size and these associations in language is novel as far as I know other studies have looked at this in terms of limitations some of our variables potentially have poor reliability and validity for example our mental health variable also there's small ethnic minority sample sizes and the ethnic density analysis was quite crude but these are preliminary analyses and will conduct more granular analysis in future so just to conclude in these preliminary findings we found an effect for ethnic minorities but not for language or national identity in Wales I think these are quite interesting no findings because we did expect that associations would extend to these groups so like I said the studies in its early stages so I'm keen to get into that these analyses were exploratory so in future we'll be conducting analyses on more recent ways of the national survey for Wales and these data include a clinical mental health variable that we can use and finally with survey data of course you're limited by the questions asked in the survey so another of my PhD studies involves a qualitatively interviewing individuals with psychosis in low own linguistic density areas and I also hope to conduct these analyses using a more psychosis related outcome because we might see that these kinds of associations are more likely to be observed for psychosis because there is a degree of specificity to psychosis compared to more common mental health problems okay and that's everything thank you very much for listening and does anybody have any questions okay everyone we're going to go to the final presentation from today from today's panel so this is a recording presentation and it is by Ji Yong Siw from University College London Ji Yong moved to the UK in 2017 and she's going to graduate from her undergraduate degree in population health at UCL so congratulations to Ji Yong she's interested in equality, inclusion and diversity and social everyday reality and is keen to study further emerging health issues caused by poverty and inequality in public health and international development and she is going to start studying for an MSc from this autumn hello I'm Ji Yong Siw and today I'm going to introduce my dissertation project to examine whether there is an association between changing neighbourhood ethnic density and individual mental health in the UK this presentation is constructed with four main sections first I'll explain the ethnic density which is a fundamental CC service that we should look at before listening to my project I'll explain the UK household longitudinal study UK HLS and study methods that are selected for this project then we will look at the findings and interpretation of the outcomes then finally I will discuss some strengths and limitations of the studies and the points that future studies need to focus on so I will start with what is ethnic density and its related CCs ethnic density is defined differently depending on the research but we use the general term which shows the proportion of ethnic minority groups in a neighbourhood ethnic minority groups in the UK are more likely to be exposed to greater risk of health including both physical and mental health but this ethnic density suggested interesting outcomes that greater density in a neighbourhood can provide protective or positive impact on health outcomes this is quite a new concept but some of the UK-based studies have tested and supported the positive impacts of greater density on mental health particularly in those from Indian and Pakistani backgrounds so this study tests whether the positive impact are also applied when we look at the change in ethnic density between two different time points as a predictor variable and based on the excessive research the mechanism was developed as ethnic density in a neighbourhood might provide ethnic minority to have stronger social cohesion and support or the sense of community by sharing their cultural, religious and linguistic values and this kind of environment might also reduce or prevent the racial discrimination or low status stigma in this neighbourhood against ethnic minority compared to neighbourhood with low proportion of ethnic minority which also have positive or protective impacts on mental health and their wellbeing the study was designed to test full main hypothesis the first hypothesis was increasing ethnic density in a neighbourhood is associated with better mental health the second hypothesis is the positive impacts of increased density are greater for non-white groups than white ethnic groups this will be because ethnic minority has stronger social cohesion with the same ethnic groups or other ethnic groups as found by previous studies and this is tested in the third hypothesis which is the greater impacts of increased density for non-white groups are moderated by better social cohesion the last hypothesis was the positive impact of increased density is greater for living in less deprived neighbourhood and this test whether the magnitude of positive impact are different depending on neighbourhood deprivation levels which is another environmental condition that might affect mental health this study mainly used the UK household longitudinal study UKHLS which is an annual household survey started in 2009 to 2010 using over 40,000 UK households this provides high quality multidisciplinary data on health, socio-economic status and social life and the UKHLS households are selected through a two-stage cluster random sampling and all household members become a part of the samples the study data was taken from Wave 9 which was collected in 2017-19 from more than 24,000 households and a special licence was applied to access the UKHLS data and access to data was under restrictive conditions this data was combined with 2001-2011 census data which were used to calculate the acid density and neighbourhood deprivation and these were combined using the lower level super-upper area codes the outcome variable in the study was mental health and this was measured by general health questionnaire 12 GHQ12 was a validated screening tool for minor psychiatric immobility and a good proxy measure for depressive disorder and general population and non-clinical setting I used the total score of GHQ12 where higher value represents worse mental health the predictor variable was to change the acid density between 2001 and 2011 and as mentioned before this was measured using the census data on a lower level covariates are the variables that are also able to affect mental health so I included neighbourhood deprivation and this was measured by the Townsend deprivation index which was widely used in health research and calculated by aggregating standardised values of full deprivation indicators I also added social cohesion which was measured by instrument and this compromises of attraction to the neighbourhood neighbouring and psychological sense of community then I included age, sex, ethnicity monthly gross income, highest qualification and living areas multi-level modelling was used because individuals in the UK are nested within household and neighbourhood but household level was excluded in this analysis because around half of households represent only one individual and there was statistical evidence of no significant difference on mental health between individual and household levels I will now proceed to present the findings from this project Between 2001 and 2011 we can find evidence of England and Wales have become more ethnically diverse and dense 92% of neighbourhoods experience an increase in the proportion of ethnic minority and the average increase across all neighbourhood was 6.4% The previous findings show that ethnic minority groups concentrated in more deprived neighbourhood but as you can see in this box plot the greatest increase was also reported in the most deprived neighbourhood which is coloured in purple in this plot This might mean that ethnic minority density was already high in the most deprived neighbourhood and it increased the most over the 10 years This table compared the outcomes of 5 multi-level regression models and please post a recording here if you would like to take a look at the results and I will go over them on the next slide Through the project we can find that increased ethnic density was initially associated with better mental health and non-white groups as we hypothesised but we found worse mental health in white ethnic groups but when the model considered neighbourhood deprivation variable the positive impacts for non-white groups were no longer statistically significant This might be because non-white groups concentrated in more deprived neighbourhood so controlling neighbourhood deprivation variable might result in weak statistical power to detect an effect across different neighbourhood deprivation levels or another explanation might be that people in deprived neighbourhood have exposed the greater density before the research as we are looking at the change in ethnic density on mental health rather than just higher density at one point When we added the interaction between change in density and social cohesion the increase in ethnic density was associated with better mental health for white ethnic groups but not for non-white groups So a marginal effects analysis we can find that white ethnic groups have positive impacts when they keep low or moderate social cohesion with neighbours but there are adverse impacts when social cohesion is higher than a certain point These findings might reflect the difficulty of disentangling the density effects as another study showed but it might also reflect the fact that the increase in ethnic density of non-white groups might not be homogeneous but maybe in fact the variety of ethnic minority groups entering the area who could have weak social interaction between themselves Moving on from the results I will briefly touch on the strengths and limitations of the project One of the main strengths of this project was that there are two nationally representative sources UK, Chalas and Senses These have large sample size across different age, socio-economic status and ethnic minority groups and this is expected to produce valid and reliable results The project also used a change in ethnic density over time to show stronger association of ethnic density effects and mental health and to my knowledge there has been no project considered relevant neighbourhood characteristics including residential mobility or ethnic density change over time This helps overcome the temporality of neighbourhood effects on health by measuring exposures over a longer duration Lastly, this project used less distinctive categories of ethnic groups when calculating density change because we used the proportion of ethnic minority rather than focusing on a specific ethnic group I believe this increased validity of the results to a broader set of ethnic minorities than focusing on a specific group and also reflects the changing view of general population regarding the ethnicity There were some limitations in this study Firstly, we didn't consider the initial level of ethnic density levels but focus on the change in ethnic density Secondly, it was also difficult to detect the direction of causality because the data was still coming from cross-sectional study Lastly, although this is a minor problem as UKHLS has a large sample size but some data was excluded when the data was processed especially because I used the complicated cases In conclusion, the increased ethnic density and neighbourhood deprivation were not important because there were many mental health using existing UKHLS data In this study, social cohesion was known as a potential mediator that partially linked increased density and mental health for white ethnic groups but this still did not directly explain the density change effect because high level of social cohesion was found to have poor mental health This study was also unable to find the association and any meaningful mediator or moderator for non-white groups So, a future study needs to verify the findings in this study using other study data or measures and we might need to test other confounding factors including exposure to diverse culture and reduce low status stigma But I believe that this kind of study will have a fundamental impact on health especially in implementing more cost-effective and efficient health interventions because it examines the neighbourhood effect which is often overlooked in favour of individual risk Thank you very much for your time and please let me know if you have any questions