The relationship between social and environmental factors and symptom severity in the seriously mentally ill population

Abstract

Background:

The goal of this article is to investigate the relationship of psychiatric symptom severity with internalised stigma, neighbourhood environment, and social support among individuals with serious mental illness.

Method:

Using a longitudinal study design we examined the relationship between psychiatric symptom severity with internalised stigma, neighbourhood environment, and social support among 271 adults with serious mental illness recruited from new admissions to two urban mental health clinics.

Results:

After controlling for demographics increased stigma levels predicted greater symptom severity, as measured by the Positive and Negative Syndrome Scale (PANSS) Positive, Negative, and General Psychopathology scales over a 4-year period (p < .05). In adjusted models, individuals who reported living in more disadvantaged neighbourhoods also reported higher PANSS Negative and General scores over time (p < .05). Social support from friends and relatives was not significantly related to PANSS Positive, Negative, or General Psychopathology scores among individuals with serious mental illness.

Conclusions:

Individuals with serious mental illness who experience internalised stigma and neighbourhood disadvantage experience greater symptom severity over time. Targeting stigma and housing during treatment could potentially impact symptom severity in this population.

Keywords

Serious mental illness stigma neighbourhood social support

Introduction

Stigma, social support, and neighbourhood environment have been found to impact the recovery of individuals with serious mental illness (Cullen et al., 2017; Kloos & Townley, 2011; Yanos et al., 2008), but research on these social and environmental factors’ relationships with psychiatric symptoms severity has had mixed results (Boyd et al., 2016; Lysaker et al., 2007; Yanos et al., 2008). Boyd et al. (2016) found that among homeless veterans with serious mental illness, higher levels of internalised stigma were associated with significant increases in psychotic and depressive symptoms as measured by Symptom Checklist-90-R. Another study of 36 individuals with schizophrenia reported no relationship between internalised stigma and the severity of negative symptoms as measured by the Positive and Negative Syndrome Scale (PANSS; Lysaker et al., 2007). Associations with positive symptoms as measured by the PANSS were only significant at the trend level (Lysaker et al., 2007). Similarly, Cullen et al. (2017) found that high PANSS scores reported by individuals with serious mental illness did not correlate with high stigma levels.

Social support and neighbourhood relations are frequently cited in the literature as predictors of well-being and recovery among individuals with serious mental illness (Kloos & Shah, 2009, Wright & Kloos, 2007). Many studies have examined social support and symptomology in first-episode psychosis (Bjornestad et al., 2017; Norman et al., 2005). One such study of patients with first-episode psychosis demonstrated that the frequency of interactions with friends, but not family, predicted symptom severity (Bjornestad et al., 2017). Another found that increased social support among people with first-episode psychosis was related to lower levels of positive symptoms (Norman et al., 2005). To the authors’ knowledge, no studies have evaluated how neighbourhood factors are related to PANSS scores in the seriously mentally ill population.

In this longitudinal study, we used data from a sample of 271 patients with serious mental disorders recruited from two inner city community psychiatry clinics to examine how stigma, neighbourhood factors, and social support predict PANSS scores over a 4-year period. We hypothesise that higher internalised stigma levels, greater levels of neighbourhood disadvantage, and few social supports will all be associated with more severe symptom levels in individuals with serious mental illness.

Method

Sample

The sample for this study included adults with serious mental illness who were patients at two urban community psychiatry clinics in Baltimore, Maryland. Participants were recruited from consecutive, new admissions from August 2008 through December 2012. Assessments in the study were conducted at annual intervals over the course of the study. This longitudinal cohort included four waves of data collection, including 1-, 2-, 3-year follow-up assessments from baseline (Wave 1, n = 271; Wave 2, n = 197; Wave 3, n = 164; Wave 4, n = 78). Clinical diagnoses were extracted from medical records. Adults with Diagnostic and Statistical Manual of Mental Disorders, 4th Edition (DSM-IV) primary diagnoses of schizophrenia, schizoaffective disorder, schizophreniform disorder, bipolar disorder (type I or type II), psychotic disorder not otherwise specified (NOS), or major depressive disorder with psychosis were eligible for the study. Participants were also required to speak English and live in the Baltimore area. All participants provided written consent. Previous studies have described the baseline sample and design of the study in more detail (Mojtabai et al., 2014, Nugent et al., 2015). Participants were followed up three additional times approximately 1 year apart.

The study was approved by the Institutional Review Board (IRB) at the Johns Hopkins Bloomberg School of Public Health.

Measures

Dependent variable

Positive and Negative Syndrome Scale

Severity of symptoms was measured by the PANSS. The PANSS is a 33-item scale developed to measure the positive and negative symptoms and general severity of psychopathology in individuals with schizophrenia (Kay et al., 1987). It has three subscales that measure positive symptoms of schizophrenia, negative symptoms of schizophrenia, and general psychopathology.

Independent variables

Stigma

Internalised stigma towards mental illness was measured by selected items from the Internalised Stigma of Mental Illness (ISMI) Scale (Ritsher et al., 2003). These questions assessed negative attitudes towards mental illness in general and internalised stigma. A total of six items from the ISMI Alienation subscale were included in our survey. Examples of questions related to internalised stigma included the following: ‘I feel out of place in the world because I have a mental illness’, ‘I am disappointed in myself for having a mental illness’, ‘I feel inferior to others who don’t have a mental illness.’ Three items from the ISMI Stereotype Endorsement and Stigma Resistance subscales included general language towards mental illness. Examples of these general questions included the following: ‘Mentally ill people tend to be violent’, ‘People with mental illness need others to make decisions for them’, ‘Mentally ill people shouldn’t get married.’ One item was rephrased from the ‘cannot’ to ‘can’ so that it would be positively worded. Questions were rated on a scale from 1 to 4 such that higher scores indicated higher levels of internalised stigma (1 = Strongly Disagree; 2 = Disagree; 3 = Agree; 4 = Strongly Agree). The scores for this scale were computed by summing all items. Scores ranged from 12 to 48.

Neighbourhood environment

Neighbourhood characteristics were measured by the Neighbourhood Environment Scale (Elliott et al., 1985). This scale includes 18 items that assess experiences of neighbourhood disadvantage such as violent crime, safety, racism, prejudice, drug activity, poverty, and physical environment. Responses were scored on a scale of 1 to 6 (1 = Not at all; 2 = Very little; 3 = A little; 4 = Somewhat; 5 = Pretty much; 6 = Very much). Some questions were recoded such that higher scores reflected a more disadvantaged neighbourhood. Response scores were summed to create a final summary score for neighbourhood disadvantage, ranging from 18 to 108.

Social support

Perceived quality of support from friends and relatives was measured by a series of six questions. Questions related to perceived support from family included the following: ‘How much do your relatives really care about you?’ ‘How much can you rely on them for help if you have a serious problem?’ ‘How much can you relax and be yourself around them?’ ‘How often do you relatives make too many demands on you?’ ‘How often do they let you down when you are counting on them?’ ‘How often do they get on your nerves?’ The same six questions were also asked in the context of support from friends. These questions were only asked from individuals who reported that they had at least one relative or friend. Responses to questions were rated on a four-point ordinal scale (1 = Not at all; 2 = A little; 3 = Some; 4 = A lot). Responses were reverse coded such that higher scores indicated more supportive relationships. Responses were summed to create a final summary score, ranging from 6 to 24.

Statistical analyses

To include all four waves of data, we conducted longitudinal data analyses using a random-effects model. Our models accounted for repeated measurements of the PANSS at baseline, 1-, 2-, and 3-year follow up. Lagged variables were created for predictors, such that social factors at baseline were correlated with year 1 outcomes, controlling for year 1 social factors. The strengths of using a random-effects model for longitudinal data is that it accounts for follow up data at multiple time points and does not exclude participants from analyses based on missing data (Hedeker & Gibbons, 1997). First, stigma, neighbourhood disadvantage, and social network variables were placed in the regression model individually and their associations with PANSS scores were measured. Next, each independent variable was placed in a model, adjusting for age, sex, race, primary diagnosis.

All analyses were conducted using STATA 15 (StataCorp, 2017).

Results

A total of 874 patients met criteria for the study of whom 271 (31%) agreed to participate and completed the assessment. Of these, 197 (73%) participated in follow-up assessments 1 year later, 164 (60.05%) 2 years later, and 78 (29%) 3 years later.

Descriptive results

Baseline sample characteristics are presented in Table 1. The mean age of the sample was 41.8 (SD = 0.33). A majority of participants in the study were female (53%) and non-Hispanic Black (54%). A majority of participants had a diagnosis of schizophrenia (33%) or bipolar disorder (45%). About 15% of participants had a clinical diagnosis of major depressive disorder and 7% had a clinical diagnosis of psychosis NOS.

Table 1.

Baseline characteristics of patients with serious mental illness.

Variable (N = 271)	%
Gender (Female)	53.1%
Race
White	33.6%
Black	54.2%
Other	12.2%
Primary diagnosis
Schizophrenia	32.8%
Bipolar	45.4%
Major depression with psychotic features	14.8%
Psychosis NOS	7.0%
Marital status
Married	8.1%
Never married	46.1%
Divorced	18.1%
Other (separated, widowed, unmarried)	27.7%
Education
⩽8th grade	8.1%
<12th grade	24.4%
12th grade or GED	37.6%
Some college	19.2%
⩾4 years of college	10.7%
Employment
Currently working	11.5%
Not currently employed	24.4%
Disabled	62.2%
Other (student, retired)	1.8%
Number of relatives
0	9.6%
1	12.5%
2–4	44.3%
5–6	15.5%
>6	18.1%
Number of friends
0	23.3%
1	15.9%
2–4	34.4%
5–6	10.7%
>6	15.6%
Age (SD)	41.8 (0.33)
Stigma (SD)	29.5 (4.8)
Neighbourhood disadvantage (SD)	50.1 (20.7)
PANSS
PANSS positive (SD)	15.2 (6.3)
PANSS negative (SD)	12.3 (5.5)
PANSS general (SD)	31.1 (9.0)

The mean internalised stigma score for this population at baseline was 29.5 (SD = 4.8; range 12–48); the mean score for neighbourhood disadvantage was 50.1 (SD = 20.7; range 18–108); the mean friend support score was 18.0 (SD = 3.1; range 6–24), and the mean relative support score was 17.1 (SD = 3.6; range 6–24). PANSS General Psychopathology, PANSS Positive, and PANSS Negative, mean scores were 31.1 (SD = 9.0), 15.2 (SD = 6.3), and 12.3 (SD = 5.5), respectively. At baseline, nearly one in four participants (23%; n = 63) reported that they did not have any friends, neighbours, or coworkers with whom they keep in touch; 9.6% (n = 26) of participants reported that they did not have any family members or relatives at all. These individuals were not included in the relative/friend support score. Further, a majority of participants (50.3%) reported having one to four friends. Across assessment waves, drop out from the study was not consistently associated with stigma, social support, or neighbourhood environment, on the one hand, and with PANSS scores, on the other hand.

PANSS positive

After adjusting for age, sex, race and primary diagnosis, analyses showed that stigma was significantly related to PANSS Positive scores over time (Table 2). Participants whose levels of internalised stigma increased over time had higher PANSS Positive scores over time (β = 0.17; p = 0.02; SE = 0.08). Neither neighbourhood environment nor social support were found to be significantly related to PANSS Positive scores over time.

Table 2.

Relationship of stigma, neighbourhood and social support to PANSS scores.

Variable	PANSS positive^a			PANSS negative^a			PANSS general^a
Variable	Regression coefficient	CI	p	Regression coefficient	CI	p	Regression coefficient	CI	p
Stigma	0.17	0.02–0.32	.02	0.21	0.06–0.37	.006	0.36	0.12–0.59	.003
Neighbourhood	0.02	−0.01–0.06	.19	0.05	0.01–0.08	.01	0.06	0.004–0.12	.04
Social support
Friends	−0.14	−0.36–0.08	.20	−0.12	−0.34–0.10	.29	−0.24	−0.62–0.13	.21
Relatives	−0.09	−0.24–0.06	.25	−0.10	−0.25–0.04	.17	−0.20	−0.43–0.03	.09

Note. ^aRandom regression model adjusted for age, sex, race, and primary diagnosis.

PANSS negative

In adjusted models, individuals who reported increased stigma over time had higher PANSS negative scores (Table 2; β = 0.21; p = 0.006; SE = 0.08). Additionally, higher levels of neighbourhood disadvantage were significantly associated with higher PANSS Negative scores over time (Table 2; β = 0.05; p = 0.01; SE = 0.02). The degree of social support from family and friends was not significantly related to PANSS Negative scores over the course of the study.

PANNS general psychopathology

In adjusted analyses, higher internalised stigma scores were significantly associated with higher PANSS General Psychopathology scores over time (Table 2; β = 0.36; p = 0.003; SE = 0.12). Higher levels of neighbourhood disadvantage were significantly associated with higher PANSS General Psychopathology scores over time (Table 2; β = 0.06; p = 0.04; SE = 0.03). Social support was not significantly related to PANSS General Psychopathology scores in adjusted analyses.

Discussion

Study findings shed more light on the relationship between internalised stigma, neighbourhood factors, and social support versus psychiatric symptom severity. Across all three PANSS scales internalised stigma was significantly associated with symptom severity over time. Although other studies did not find the same relationship for all three PANSS scales (Lysaker et al., 2007), these results fit well with studies that have noted the impact of stigma on seeking treatment, adhering to treatment, and overall recovery (Chronister, 2013; Link et al., 1997, 2001; Sirey et al., 2001). These studies have shown that stigma is associated with decreased adherence to treatment (Sirey et al., 2001) and a lower a likelihood of recovery (Chronister, 2013). While all those involved in this study were in treatment, it is possible that the presence of internalised stigma had an adverse effect on medication and appointment adherence with a resultant increase in their symptom levels across all domains.

Our findings suggest that neighbourhood environment also plays a role in symptom severity in this population. For individuals with serious mental illness, living in disadvantaged neighbourhoods was strongly associated with higher levels of negative symptoms and general psychopathology over time. While no prior studies have reported on these particular associations, social and structural living conditions are known to be critical factors in recovery among individuals with serious mental illness (Kloos & Shah, 2009). Additionally, previous studies have suggested that neighbourhood social climate can play a role in the experience of psychological distress among individuals with serious mental illness as measured by the Brief Symptom Inventory (Kloos & Townley, 2011, Wright & Kloos, 2007). This would support our findings of increased symptomology among those who are living in a disadvantaged neighbourhood.

Contrary to hypotheses, we did not find that social support from family and friends to be associated with symptom levels on any of the PANSS scales. This is in keeping with prior work where we found that while the perceived support from relatives and friends impacted internalised stigma levels such that the greater the network and support the lower the stigma level, these results were independent of PANSS levels (Cullen et al., 2017). A recent meta-analysis by Degnan et al. (2018) looked at the strength of associations between social network size and clinical outcomes in schizophrenia. Here they found no association between network size and positive symptoms, but they did find an association between negative symptoms and network size (Degnan et al., 2018). Our study focused on characteristics of social support rather than network size and because 33% of our sample reported having no relatives or friends, they did not complete the social support scale, so data was not available for analyses. It is possible that our overall sample size may have made it difficult to detect differences in this area.

The strength of this study lies in the rigorous assessments and its longitudinal nature. Limitations of this study include small sample sizes and a large proportion of participants lost to follow-up. Also, this study was conducted in a community psychiatry program in an urban setting and may not be generalisable to other contexts.

Conclusion

The findings of this study highlight how psychosocial and environmental factors can play an important role in the experience of psychiatric symptoms by patients with serious mental illness. Both individual interventions, such as cognitive behavioural therapy for internalised stigma (Morrison et al., 2016), and population-based interventions, such as public policies that impact neighbourhood environments (Compton et al., 2020), when combined with conventional treatment, may be beneficial to this population and reduce their burden of psychiatric symptoms and associated disability.

Footnotes

Conflict of interest

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: An anonymous donor provided funds for the collection of this data. Katrina M. Rodriguez is supported by the National Institute of Mental Health’s Psychiatric Epidemiology Training Program (5T32MH014592-39; PI: Volk, Heather).

ORCID iD

Tara Von Mach

Katrina Rodriguez

References

Bjornestad

ten Velden Hegelstad

Joa

Davidson

Larsen

T. K.

Melle

Veseth

Johannessen

J. O.

Bronnick

(2017). “With a little help from my friends” social predictors of clinical recovery in first-episode psychosis. Psychiatry Research, 255, 209–214.

Boyd

J. E.

Bassett

E. D.

Hoff

(2016). Internalized stigma of mental illness and depressive and psychotic symptoms in homeless veterans over 6 months. Psychiatry Research, 240, 253–259.

Chronister

Chou

C. C.

Liao

H. Y.

(2013). The role of stigma coping and social support in mediating the effect of societal stigma on internalized stigma, mental health recovery, and quality of life among people with serious mental illness. Journal of Community Psychology, 41(5), 582–600.

Compton

M. T.

Bakeman

Capulong

Pauselli

Alolayan

Crisafio

King

Reed

Broussard

Shim

(2020). Associations between two domains of social adversity and recovery among persons with serious mental illnesses being treated in community mental health Centers. Community Mental Health Journal, 56(1), 22–31.

Cullen

B. A.

Mojtabai

Bordbar

Everett

Nugent

K. L.

Eaton

W. W.

(2017). Social network, recovery attitudes and internal stigma among those with serious mental illness. International Journal of Social Psychiatry, 63(5), 448–458.

Degnan

Berry

Sweet

Abel

Crossley

Edge

(2018). Social networks and symptomatic and functional outcomes in schizophrenia: A systematic review and meta-analysis. Social Psychiatry and Psychiatric Epidemiology, 53(9), 873–888.

Elliott

D. S.

Huizinga

Ageton

S. S.

(1985). Explaining delinquency and drug use. SAGE Publications.

Hedeker

Gibbons

R. D.

(1997). Application of random-effects pattern-mixture models for missing data in longitudinal studies. Psychological Methods, 2(1), 64.

Kay

S. R.

Fiszbein

Opler

L. A.

(1987). The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophrenia Bulletin, 13(2), 261–276.

10.

Kloos

Shah

(2009). A social ecological approach to investigating relationships between housing and adaptive functioning for persons with serious mental illness. American Journal of Community Psychology, 44(3–4), 316–326.

11.

Kloos

Townley

(2011). Investigating the relationship between neighborhood experiences and psychiatric distress for individuals with serious mental illness. Administration and Policy in Mental Health and Mental Health Services Research, 38(2), 105–116.

12.

Link

B. G.

Struening

E. L.

Neese-Todd

Asmussen

Phelan

J. C.

(2001). Stigma as a barrier to recovery: The consequences of stigma for the self-esteem of people with mental illnesses. Psychiatric Services, 52(12), 1621–1626.

13.

Link

B. G.

Struening

E. L.

Rahav

Phelan

J. C.

Nuttbrock

(1997). On stigma and its consequences: Evidence from a longitudinal study of men with dual diagnoses of mental illness and substance abuse. Journal of Health and Social Behavior, 38(2), 177–190.

14.

Lysaker

P. H.

Davis

L. W.

Warman

D. M.

Strasburger

Beattie

(2007). Stigma, social function and symptoms in schizophrenia and schizoaffective disorder: Associations across 6 months. Psychiatry Research, 149(1–3), 89–95.

15.

Mojtabai

Cullen

Everett

Nugent

K. L.

Sawa

Sharifi

Takayanagi

Toroney

J. S.

Eaton

W. W.

(2014). Reasons for not seeking general medical care among individuals with serious mental illness. Psychiatric Services, 65(6), 818–821.

16.

Morrison

A. P.

Burke

Murphy

Pyle

Bowe

Varese

Dunn

Chapman

Hutton

Welford

Wood

L. J.

(2016). Cognitive therapy for internalised stigma in people experiencing psychosis: A pilot randomised controlled trial. Psychiatry Research, 240, 96–102.

17.

Norman

R. M.

Malla

A. K.

Manchanda

Harricharan

Takhar

Northcott

(2005). Social support and three-year symptom and admission outcomes for first episode psychosis. Schizophrenia Research, 80(2–3), 227–234.

18.

Nugent

K. L.

Spahr

Toroney

Mojtabai

Nettles

Turner

L. W.

Fenton

Spivak

Cullen

B. A.

Everett

Eaton

W. W.

(2015). Interrater reliability of the modified Monitoring of Side Effects Scale for assessment of adverse effects of psychiatric medication in clinical and research settings. Journal of Clinical Psychopharmacology, 35(3), 324–328.

19.

Ritsher

J. B.

Otilingam

P. G.

Grajales

(2003). Internalized stigma of mental illness: Psychometric properties of a new measure. Psychiatry Research, 121(1), 31–49.

20.

Sirey

J. A.

Bruce

M. L.

Alexopoulos

G. S.

Perlick

D. A.

Friedman

S. J.

Meyers

B. S.

(2001). Stigma as a barrier to recovery: Perceived stigma and patient-rated severity of illness as predictors of antidepressant drug adherence. Psychiatric Services, 52(12), 1615–1620.

21.

StataCorp. 2017. Stata statistical software: Release 15. College Station, TX: StataCorp LLC.

22.

Wright

P. A.

Kloos

(2007). Housing environment and mental health outcomes: A levels of analysis perspective. Journal of Environmental Psychology, 27(1), 79–89.

23.

Yanos

P. T.

Roe

Markus

Lysaker

P. H.

(2008). Pathways between internalized stigma and outcomes related to recovery in schizophrenia spectrum disorders. Psychiatric Services, 59(12), 1437–1442.