Does sexual identity and religious practice have implications for individual’s subjective health and wellbeing? Secondary data analysis of the Community Life Survey

ABSTRACT

The health and wellbeing of Lesbian, Gay, Bisexual and Transgender (LGBT) individual's has gained attention in recent years, particularly the unique stressors associated with health. Few studies have explored implications of sexual identity and religious status on mental health and wellbeing. A secondary data analysis was performed on the Community Life Survey. A multivariate interaction was found between age, religious practice and sexual identity when considering wellbeing. An ANOVA of wellbeing scores revealed a significant difference between sexual identity groups, LGBT group scoring lowest, with a significant interaction between religion and sexual identity. General health generated a significant difference between groups for religious practice. Understanding and challenging cultural norms in service settings is key for policy and practice. LGBT individuals' are at greater risk of lower wellbeing and increase dissatisfaction.

KEYWORDS:

• Sexuality
• religion
• mental health
• wellbeing

The social, emotional and psychological needs of Lesbian, Gay, Bisexual and Transgender (LGBT) individuals has gained attention in recent times (Foster et al., 2015; Vaughan et al., 2014; Wilkinson & Johnson, 2020), including exploration of the intersecting of identities (Foster et al., 2015; Herek et al., 2010; Meyer & Northridge, 2007; Zinnbauer et al., 1997). LGBT individuals of religious affiliation or faith experience unique stressors (Page et al., 2013) as they negotiate the realities of these identity positions.

Sexuality and sexual identity remain the focus of religious and faith-based debate and conflict (Hunter, 2013). An established scholarly literature documents the relationship between religious beliefs and negative public attitudes towards homosexuality and same-sex marriage (Olson et al., 2006). Homosexual behaviours are labelled “deviant” and “wrong” by some western-religious cultures (Heermann et al., 2007). Sexual acts or behaviours that are conducted in same-sex relationships are deemed sinful by many traditional Jews (Mahl, 2008); require punishable church discipline within some Mormon communities (Heermann et al., 2007) and are punishable by the death penalty in some Islamic cultures (Hamdi et al., 2017). Some religious and faith communities have adopted acceptance towards homosexual individuals by separating identity from behaviour (Rosik et al., 2007). This is in keeping with the account of sexuality having three dimensions (Geary et al., 2018), including sexual attraction (or interest), sexual behaviour and sexual identity (Geary et al., 2018). Sexual identity encompasses how an individual wishes to self-define themselves; sexual interest or attraction related to what an individual wants to do regardless of whether they do it; and sexual behaviour as what an individual does regardless of their sexual interest or sexual identity. Sexual orientation, therefore, describes a distinct type of intense sexual interest (Moser, 2016).

The existence of two sexual orientations (homosexuality and heterosexuality) has been accepted widely, however, some would argue these are identities (Moser, 2016) and therefore, other sexual orientations groups have been widely debated (Moser, 2016). Elsewhere, it is argued, that sexual orientation is more fluid and less “fixed” (Epstein et al., 2012; Ross et al., 2012).

The experience of rejection by others (including family and friends), relating to sexual behaviour (Rosik et al., 2007), regardless of feeling accepted as a person, might have a more detrimental impact on a person if they try to change their sexual orientation to feel “fully accepted” (Itzhaky & Kissil, 2015; Page et al., 2013). Experiences like these have been practiced by some religious and faith groups who advocate conversion therapy (Dehlin et al., 2015). A recent report claims 51% of LGBT individuals received conversion therapy by faith groups (GEO, 2019). Despite the support for reorientation therapies (e.g., conversion therapy) by some religious groups, mental health organisations have argued they cause more harm than good and should be avoided by mental health practitioners (APA, 2009).

Religious teaching and discourse have fed into the formation of a heteronormative social and political order (Bowers et al., 2010), marginalising LGBT communities. Homonegativity, in some communities, has been attributed to a heterosexist interpretation of the religious text rather than the religious text itself (Yip, 2007). The lack of belonging to groups and wider society heightens marginalisation and the risk of dissonance between two critical parts of one’s identity: the religious and the sexual (Rodriguez & Ouellette, 2000), which can lead to emotional distress, depression and suicidality (Evans & Barker, 2010). Irrespective of UK government efforts to foster equality for the LGBT community (e.g., same-sex marriage), barriers, especially within faith and religious communities, continues to exist. For example, protests at faith-based schools in the UK for introducing “no outsider programmes” which educates about LGBT relationships (Parveen & Weale, 2019).

The existing empirical and theoretical evidence about LGBT individuals experience of religion (and or spirituality) and their wellbeing (and or psychological health) suggests that it is complex and can come with an array of challenges (Meyer, 2016). However, there are mix of experiences reported with regards to religion and/or spirituality acting as a “protective factor” or a “risk factor” in regard to wellbeing. Religion has been associated with psychological and social resources, for example, support communities (Weber & Pargament, 2014) when coping with stress or associated conditions (Koenig, 2009) and resilience building (Foster et al., 2015; Yip, 2007). Consequently, religion/spirituality might be considered a “protective factor” providing a supportive resource when experiencing depression, suicide or anxiety (Weber & Pargament, 2014). The extent of the protective, as opposed to the detrimental, nature, is debated (Meltzer et al., 2011). Non-heterosexual youths have described feeling rejected by religious groups (Page et al., 2013; Hamblin & Gross, 2013) therefore experiencing the opposite of support, often leading to elevated levels of stress. This is an example of minority stress (Meyer, 2016). Moreover, LGBT individuals, who mature in a religious context, have described the increased risk of experiencing internalised homophobia and consequently, increased suicidal thoughts and behaviours (Gibbs & Goldbach, 2015). In some cases, however, LGBT individuals turn to religion to persevere against the challenges brought about by the oppression and social injustice that religious indoctrination brings to their lives (Foster et al., 2015; Jeffries et al., 2008; McCarthy & das Nair, 2018).

The existing evidence is limited by sampling biases and representation of groups. The majority of research in this field has derived from datasets population by United States citizens, few have been conducted in Europe. Most of the datasets have recruited specified samples, for example colleges or university campuses that assume a religion based on the ethos of the college or university. Some studies have recruited through religious or LGBT-based charities. Consequently, many studies focus on individuals of a particular religious belief and generally have low representation of individuals who identify as bisexual or transgender. However, based on this data, taking into consideration the limitations, there is a consistency of evidence from qualitative and quantitative research approaches (Wilkinson & Johnson, forthcoming; Wilkinson & Johnson, 2020)

This unique study explores differences between individuals with and without a religious practice, for heterosexual compared to sexual minority individuals’ (LGBT), with regards to subjective health and wellbeing using a dataset based on English citizens; the Community Life Survey 2016–2017 (Department for Culture, Media and Sport, 2019). This was the latest dataset available at the time of data analysis, in special licence version, allowing access to sexual identity data. This study focuses on sexual minority status; LGBT as this focus reflects that which has been captured in the sexual identity question of the Community Life Survey. I fully acknowledge the need to better understand the experiences of the transgender community, with regards to their experiences with religion or faith, and implications for physical health, mental health and wellbeing. On this occasion, the data to support this kind of analysis, for transgender, was not provided in the current dataset. However, given the unique experience of transgender individuals, this kind of analysis requires a study of its own.

Research question:

How do sexual identity and religious practice have implications for individual’s subjective health and wellbeing?

How does sexual identity and religious practice interact?

Hypotheses:

• (1). Subjective wellbeing and self-disclosed general health will be higher for older age categories compared to younger age categories; practicing religion compared to non-practicing religion; heterosexual compared to non-heterosexual (LGBT)

• (2). There will be a relationship between age group, religious practice and sexual identity when considering individuals subjective wellbeing scores.

Method

Design

This study was based on secondary data analysis, using 3 × 4 MANOVA and 3 × 4 ANOVA design, making use of the following selected variables: Religious Practice (combining Relstat); Sexual Identity (using Sid2); General Health (using GHealth) and Wellbeing (using WellB1, WellB2, WellB3 and WellB1-4) from the Community Life Survey 2016–2017 [special licence] (Department for Culture, Media and Sport, 2019).

Participants/respondents

A total number of 10,117 individuals responded to the survey, 6951 respondents of the survey answered both the sexual identity and religions status question. Of those individuals, 6649 were heterosexual and 302 were non-heterosexual (LGBT). Table 1 below presents the actual figures of survey respondents broken down by age and sex. Table 2 below presents the actual figures displayed by ethnicity. Table 3 presents a breakdown of religious status. A new variable Religious Practice was created by combining all of the “practising” responses together and the same for “non-practising”. No religion remained its own optional response.

Table 1. Survey respondents age and sex.

Age	16–24	25–34	35–44	45–54	55–64	65–74	75–84	85+	Total
Male	382	647	806	774	777	810	370	78	4644
Female	576	908	946	887	899	793	365	99	5473
Total	958	1555	1752	1661	1676	1603	735	177	10117

Table 2. Ethnicity of respondents.

Ethnicity	Total
White	8022
Asian – Indian	455
Asian – Pakistani	292
Asian – Bangladeshi	109
Asian – Chinese	113
Asian – Other	142
Black – African	226
Black – Caribbean	117
Black – other	17
Mixed	467
Other	163

Table 3. Frequency of religious status (Relstat) to demonstrate a range of religious preferences and practice status in the sample.

Practice status for each Religion	N
Practising Christian	1324
Non-practising Christian	2518
Practising Buddhist	31
Non-practising Buddhist	37
Practising Hindu	140
Non-practising Hindu	98
Practising Muslim	454
Non-practising Muslim	138
Practising Sikh	42
Non-practising Sikh	23
Practising Other incl Jewish	83
Non-practising Other incl Jewish	110
No Religion	2274

Materials and data collection procedure

This unique study made use of a big dataset from England, United Kingdom, using the Department of Culture, Media and Sport (2019) Community Life Survey 2016–2017 [special licence]. The Community Life Survey is held annually, to track trends and developments in areas that encourage social action and to empower communities. It is a household, self-completion, survey of adults aged 16 + in England. The survey is a key source of evidence on social cohesion, community engagement and social action and therefore contains some key variables important to this area of study. The 2016–2017 dataset was the latest version available in [special licence] at the time of this study. Special licence was required in order to access data on personal characteristics such as sexual identity.

The following variable data were utilised for the analysis of this study:

Age was analysed using variable rage 9 which captured age responses by participants of the survey through eight categories: 16–24; 25–34; 35–44; 45–54; 55–64; 65–74; 75–84; 85+. The categorical recording of age allowed for a MANOVA analysis combined with the following variables.

The created variable Religious Practice combined all religious practise and non-practise responses from the Relstat variable which provided information about respondent’s religious status, which religion they identified with, as well as practising or non-practising. This variable included data following the question “what is your religious status (practising or non-practising)”? Possible responses were: not classified; not applicable; Practising Christian; Non-practising Christian; Practising buddhist; Non-practising Buddhist; Practising Hindu; Non-practising Hindu; Practising Muslim; Non-practising Muslim; Practising Sikh; Non-practising Sikh; Practising other including Jewish; Non-practising other including Jewish; No Religion. Combining religious practice and non-practice responses from different religions avoided cells with small numbers for a more robust analysis and was in keeping with the theoretical basis and rationale for the implications of religion.

Sexual identity was captured in the Sid2 variable of the survey with the question worded as “which of the follow best describes how you think of yourself?” with possible responses as Heterosexual/Straight; Gay/Lesbian/Bisexual; No reponse. It is not unusual for studies to combine LGBT responses due to data representation and small sample sizes relating to power for statistical analysis. Therefore, this pre-combined variable by the database owners was deemed appropriate for this study.

Subjective wellbeing and self-rated general health were captured through a number of relevant variables. Measured on a scale of 0–10, WellB1 – how satisfied are you with life nowadays? WellB2 – overall how happy did you feel yesterday? WellB3 – overall how anxious did you feel yesterday? WellB4 – overall to what extent do you feel the things you do in your life are worthwhile? and self-rated general health GHealth – How is your health? was rated as either “very good, good, fair, bad, very bad”. A reliability analysis of the wellbeing variables produced a Cronbach’s alpha score of .8, suggesting a high level of internal consistency between the four variables.

Ethics

Permission to access the Community Life Survey 2016–2017 [Special Licence] was granted via the UK Data Service archive. This study was ethically recorded via the Institute of Police, Research Ethics Committe at the University of Chester.

Analysis / Results

Wellbeing data: analysis strategy

MANOVA was identified as the appropriate method of analysis as the dependent variables were measured using 11-point (0-10) “impact” scales, demonstrated to be more effective than four-, five- and six-point likert scales (Awang et al., 2016; Leung, 2011), and can be treated as continuous/interval data (Harpe, 2015; Leung, 2011; Wu & Leung, 2017). Parametric testing of rating scales is appropriate so long as certain conditions are met. MANOVA manages the issues associated with multiple significance testing being applied to the same data (Howitt & Cramer, 2011).

MANOVA was conducted on the independent variables age (rage9), sexual identity (Sid2) and religious practise (created combining all practicing religious groups and non-practising religious groups – see material section above for more information) and dependent variables wellbeing (WellB1, WellB2, WellB3, and WellB4). As cell sizes were unequal and Box’s test of equality of covariances was significant indicating violations in many cases, Allen and Bennett (2008) recommend that MANOVA is robust on samples over 30 to continue. Tabachnick and Fidell (2012) suggest that where samples are uneven, if the larger samples produce greater variance and covariance, then the probability values will be conservative and if in reverse significant findings should be treated with caution. MANOVA was performed and results were interpreted using Pillais trace statistics as a more conservative measure.

Hypothesis:

• (1). Subjective wellbeing and self-disclosed general health will be higher for older age categories compared to younger age categories; practicing religion compared to non-practicing religion; heterosexual compared to non-heterosexual (LGBT)

• (2). There will be a relationship between age group, religious practice and sexual identity when considering individuals subjective wellbeing scores.

See Table 4 below for descriptive statists of sexual identity and wellbeing (WellB1, WellB2, WellB3 and WellB4) variables, Tables 5 and 6 for religious status and wellbeing (WellB1, WellB2, WellB3 and WellB4), and Table 7 for age and wellbeing.

Table 4. Descriptive statistics for sexual identity and wellbeing (1–4).

	WellB1		WellB2		WellB3		WellB4
	M	SD	M	SD	M	SD	M	SD	N
Heterosexual	7.08	2.02	7.08	2.17	3.42	2.78	7.26	2.10	6682
gay/lesbian/bisexual	6.56	2.09	6.44	2.43	4.19	2.87	6.51	2.36	306
Non-response	7.03	2.34	6.99	2.73	3.60	3.01	7.31	2.31	2777
Mean Total	7.05	2.11	7.04	2.28	3.50	2.85	7.25	2.18	9766

Table 5. Frequency of religious status (Relstat) to demonstrate a range of religious preferences and practice status in the sample.

Practice status for each Religion (N)	Christian	Buddist	Hindu	Muslim	Sikh	Jewish	No Religion
Practising	1324	31	140	454	42	83	2274
Non-practising	2518	37	98	138	23	110

Table 6. Descriptive statistics for religious practice (new combined variable) and wellbeing (1–4)*.

Practice status	WellB1		WellB2		WellB3		WellB4
	M	SD	M	SD	M	SD	M	SD	N
Practising	7.29	1.98	7.31	2.08	3.56	2.85	7.63	2.02	2064
Non-practising	7.01	2.07	7.03	2.21	3.44	2.77	7.13	2.14	2915
Non-religious	6.84	2.06	6.83	2.27	3.48	2.75	6.93	2.20	2265
Non-responded	7.08	2.31	7.00	2.47	3.53	3.02	7.38	2.27	2522

Table 7. Descriptive statistics for age groups and wellbeing (1–4)*.

Age group	WellB1		WellB2		WellB3		WellB4
	M	SD	M	SD	M	SD	M	SD
16–24	6.87	2.07	6.76	2.31	3.99	2.84	6.83	2.28
25–34	7.0	1.99	7.02	2.20	3.77	2.79	7.12	2.22
35–44	6.92	2.02	6.90	2.19	3.76	2.78	7.25	2.10
45–54	6.7	2.24	6.70	2.36	3.61	2.79	7.11	2.26
55–64	6.89	2.19	6.91	2.34	3.48	2.90	7.19	2.14
65–74	7.62	1.96	7.61	2.11	2.77	2.76	7.75	1.91
75–84	7.68	2.02	7.62	2.13	2.96	2.92	7.64	2.09
85+	6.85	2.71	6.86	2.61	3.44	3.12	6.79	2.88

The MANOVA analysis highlighted significant multivariate effects when considering variables sexual identity (Sid2) and religious practice by the four measures of wellbeing (WellB1, WellB2, WellB3 and WellB4), Pillai’s trace F (8, 14340)    1.84, p < .05, partial η²= .001. Also, a multivariate effect between sexual identity (Sid2) and age (rage9), Pillai’s trace F (52, 28,688)    1.84, p < .02, partial η²= .003. However, there were non-significant multivariate effects between religious practice (relprac) and age (rage9), Pillai’s trace F (28, 28,688)    1.37, p = .09, partial η²= .001, and overall a non significant effect between sexual identity (Sid2), religious practice (relprac) and age (rage9), Pillai’s trace F (48, 28,688)    1.13, p = .25, partial η²= .002.

The same data were analysed using ANOVA, combining variables WellB1, WellB2, WellB3, and WellB4 (with reverse-scored WellB3) to form one overall dependent variable of general wellbeing. See Table 6 below for descriptive data. The ANOVA analysis suggested significant difference between groups for sexual identity, F(2)    9.87, p = .00, but religious practice was non-significant F (3,)    1.36, p > .05, however, an interaction affect was found between these variables (p < .05) (Table 8).

Table 8. Descriptive statistics of combined wellbeing scores for sexual identity and religion.

Sexual identity	Religious practice	M	SD	N
Heterosexual	Practising	28.92	6.88	1904
	Non-practising	27.84	7.33	2729
	Non-religious	27.37	7.36	1997
	Non-response	27.21	6.72	53
Gay/Lesbian/Bisexual	Practising	26.13	7.49	45
	Non-practising	25.68	8.12	97
	Non-religious	24.98	7.62	160
	Non-response	20.75	6.20	4
Non-response	Practising	25.66	7.74	115
	Non-practising	26.35	7.50	89
	Non-religious	25.63	8.30	108
	Non-response	27.96	7.92	2465

General health data

Table 9 presents the data for the dependent variable general health (GHealth) measure on a 1–5 scale.

Table 9. Descriptive statistics of religious practice, sexual identity and general health.

Sexual Identity	Religious Practice	M	SD	N
Heterosexuality	practising	2.04	.86	1910
	Non-practising	2.15	.84	2735
	Non-religious	1.97	.85	2002
	Non responded	2.35	.91	55
Gay/Lesbian/Bisexual	Practising	2.40	1.03	45
	Non-practising	2.22	.89	97
	Non-religious	2.02	.83	160
	Non responded	2.50	.58	4
Non-responded	Practising	2.24	.98	115
	Non-practising	2.31	.96	91
	Non-religious	2.10	.79	112
	Non responded	2.26	.73	19
Total	Practising	2.06	.88	2070
	Non-practising	2.16	.85	2923
	Non-religious	1.98	.85	2274
	Non responded	2.33	.85	78

A 4 × 3 ANOVA was conducted with the independent variables sexual identity (Sid2) and religious practice and the dependent variable general health (GHealth). The analysis suggested non-significant differences between groups for sexual identity, F(2)    1.80, p > .05, but religious practice was significant F (3)    6.36, p =  .000, however, an interaction affect was not found between these variables (p > .05).

Discussion

This study analysed data from the Community Life Survey uncovering patterns in the dataset representing age, sexual identity and religion when considering implications on individuals’ perceptions of wellbeing and general health. It was expected that subjective wellbeing and self-disclosed general health would be better for older age groups compared to younger age categories; those individuals practicing a religion compared to non-practicing religion; and those who identify as heterosexual compared to non-heterosexual (LGBT). Generally speaking, the participants of the survey data demonstrated better wellbeing in the older age categories compared to the younger age categories – although this declined with the top age group (i.e., 85+). Specifically, the 16–24 age group demonstrated higher levels of stress (WellB3) and lower levels of wellbeing (WellB1, WellB2 & WellB4) which is consistent with the theories capturing the iGen data (Twenge, 2017). Those identifying as heterosexual had better wellbeing that those identifying as non-heterosexual (LGBT). However, there was no difference between those individuals practising a religion compared to non-practicing for wellbeing.

Sexual identity, religious preference and practice alongside age are complex and challenging concepts to analyse and understand, especially when they are considered in the context of their relationship to wellbeing. The data included in this study suggests a relationship between sexual identity and religion, also between age and sexual identity but the data did not support that the three variables interact with each other. However, might infer from the descriptive data, that certain sexual orientation groups, alongside certain practicing and non-practicing groups, and certain age groups are having better experiences than others leading to more positive wellbeing and general health compared to other groups.

Whilst the results of the MANOVA of wellbeing data should be interpreted with caution, there appear to be some emerging patterns. It is not clear how the combination of religion and faith alongside sexual identity relates to wellbeing and general health, there are clearly some groups that are at higher risk of lower levels of wellbeing and general health. However, the MANOVA analysis highlighted significant multivariate effects of wellbeing on religious practice and sexual identity as well as sexual identity and age, suggesting that these concepts each interact with one another. In other words, sexual identity and religious practice, independently, have implications for individual’s wellbeing, as does sexual identity and age. The descriptive statistics highlight that the positive wellbeing scores (i.e., WellB1, WellB2 and WellB4) were higher in heterosexuals compared to LGBT and the wellbeing score for level of anxiety (WellB3) were lower in heterosexuals compared to LGBT individuals. In other words, LGBT individuals have higher levels of anxiety and lower levels of satisfaction, happiness and worthwhile based on mean average scores. The descriptive data for religion and wellbeing suggests that practising individuals score highest for wellbeing (WellB1, WellB2, WellB3 and WellB4), followed by non-response, non-practising then non-religious. This trend is also true for the anxiety score (WellB3). The ANOVA analysis of the combined wellbeing variable scores suggested an interaction affect between sexual identity and religious status, however, there were significant differences between groups independently for sexual identity groups but not for religious groups.

The self-reported general health variable (GHealth) revealed interesting results. In this case, there were no significant differences in terms of general health scores for the different sexual identity groups, however, there was a difference for the different religious groups (practicing vs non practicing vs non-religious). The non-religious group have the lowest mean score for general health.

These findings, to some extent, support and align with previous studies of LGBT individuals’ subjective experiences of their wellbeing and mental health, particularly in the context of those following a religion or faith (Wilkinson & Johnson, 2020). Whilst religious groups can, in some cases, provided valuable support networks (Weber & Pargament, 2014) and where individuals successfully navigate their sexual and religious identity, they can build good levels of resilience (Foster et al., 2015), leading to a positive outcome. These experiences seem to be dependent on “gay-positive” or affirming religious groups (e.g., Dehlin et al., 2015; Rodriguez & Ouellette, 2000). Example of these types of groups and developments are Dignity for LGBT Catholics, where amendments or evolutions of an existing group have been developed (Lease et al., 2005) or developing entirely new faith groups (e.g., the Metropolitan Community Church) (Lease et al., 2005). Such groups provide good levels of support rather than contributing to experiences of rejection for sexual minority groups. This allows the individual to remodel, rework or re-author (e.g., Bozard & Sanders, 2011; Minwalla et al., 2005) themselves, their beliefs and their lifestyles. However, the evidence suggests that the occurrence of these kinds of experiences are rare, and usually sexual minority individuals experience rejection and homonegativity from religious groups and peers (Page et al., 2013; Hamblin & Gross, 2013), and in some cases, this can lead to severe mental health concerns and internalised homophobia or homonegativity (Gibbs & Goldbach, 2015; Wilkinson & Johnson, 2020, 2021; Yip, 2007).

The differences that can be observed in mean scores between heterosexual and LGBT individual’s wellbeing and general health is interesting in this set of data. Meyer’s (2015) Minority Stress Theory proposes that sexual minority groups, but possibly other minority groups too, experience unique stressors which increases the chances of stress and consequential physical health concerns. Meyer (2015) maintains that the unique stressors may results in activation of bodily responses, suggesting that the unique underlying stressors manifest in mental and physical health concerns. The data deriving from the general health question in this study could map onto physical health concerns. However, the descriptive data does not support, in this case, the negative general health for LGBT individuals in comparison to the heterosexual individuals. This is of particularly relevance when considering the levels of stress induced by the experience of rejection by religious groups, peers, friends and family for sexual minorities individuals and therefore is implied in the data relating to religious status (Meyer, 2019; Wilkinson & Johnson, 2020, 2021).

It is crucial that the implications of the findings from this study are considered in policy and practice. If individuals are at greater risk of lower levels of wellbeing and higher levels of dissatisfaction as a result of both religious status and sexual identity independently, then we should consider how and where individuals are supported, especially giving the evidence of religious context for individuals of sexual minority status. Understanding and challenging the cultural and social norms in service and religious settings and support providers is pivotal to supporting the mental health and wellbeing of sexual minority individuals.

This dataset, and consequently the analysis as part of this study, did not capture individuals who have retired from or left their religious or faith group as a result of conflict between their beliefs and sexual identity. It could be interesting to consider the difference in levels of wellbeing (WellB1, WellB2, WellB3 & WellB4) and self-perceived general health for those individuals that have left their religion/religious group, given the evidence on “disaffiliation” affects (Fenelon & Danielsen, 2016), and particularly, as a result of their sexuality (also highlighted in Gibbs & Goldbach, 2015; Wilkinson & Johnson, 2020) and to consider whether this is perceived unresolved for those individuals or whether this decision/action is a way of resolving especially given the evidence that some individuals “re-authoring” to reconcile (Parker et al., 2019).

There are a number of limitations that accompany the use of large datasets, particularly for secondary data analysis where the question order, wording and process were set by another individual or team with potentially different intentions for the data. The wellbeing items from the Community Lifestyles survey specifically asked about satisfaction, happiness, anxiety and life worthwhile. These capture some important elements of wellbeing but are not necessarily a complete measure in themselves. However, subjective measures of wellbeing are arguably more useful than “objective” indicators of, for example, quality of life (Costanza et al., 2007) and in the case of the measures provided by the Community Life Survey, high levels of internal consistency were found between these items.

A particularly limiting factor of this dataset was that LGBT individuals were merged into one group, therefore, not allowing for comparison of these three differing group's characteristics with regards to both their interaction with religion and outcomes for wellbeing scores. Given the notoriously lower representation of LGBT people in research (Binson et al., 2007; Meyer & Wilson, 2009), especially for public health research (Binson et al., 2007) it is likely that these groups would have needed to be merged in order to do any analysis of the data set. However, based on the merged grouping of LGBT this dataset had a wealthy sample comparison of sexual identity groups with fairly good representation given general population representation vs sample representation. A related but equally limiting factor of this study was that the sexual identity question limited responses to LGBT, therefore some sexual identities did not have representation in the data and could not be considered in this analysis. A more comprehensive measure of sexual identity would be more informative and inclusive of diversity in future studies and analyses. In addition to the above, the capture of religious affiliation and practice was limited as it did not capture options such as atheist, agnostic or other, which could have allowed participants of the survey to specify. Therefore, the analysis has been limited to affiliation and practices of mainstream religions.

Future analysis might consider age implications by collecting age on a continuum rather than in categorical groups. Studies have previously consider the experiences of sexual minority youth (e.g., Dahl & Galliher, 2009; Page et al., 2013), however, it is thought that sexual minorities of younger generations, for a number of reasons, will be exposed to fewer stressors in their environments during early and adolescent years and therefore, it has been hypothesised that they will be exposed to greater risk factors and fewer protective factors in later years (Meyer, 2016; Twenge, 2017). This coupled with a reduction in religious and faith-based activities or affiliations in iGen (Twenge, 2017) might mean differences in this dynamic for the iGen.

Implications and conclusions

The analysis of data from the Community Life Survey supports that some groups of individuals are experiencing lower levels of wellbeing and general health and exploration for how these individuals can be supported better by services should be continued to be explored by health care services as well as specific services. Individuals who identify as non-heterosexuals are at greater risk of lower levels of wellbeing and increased levels of stress. Sexual identity and religious practice contribute to these experiences, and equally sexual identity and age both contribute to these experiences. In particular, there is a need to ensure that the wellbeing and general health of the younger generation are being supported and should be a focus of further work and review.

Acknowledgements

Thanks go to the organisers of Community Life Survey [special licence] (Department for Culture, Media and Sport, 2019) and to the UK Data Service.

Disclosure statement

No potential conflict of interest was reported by the author(s).