Loneliness and its Association with Health Behaviors in People with a Lived Experience of a Serious Mental Illness

Abstract

Objective:

To explore the association between loneliness and efficacy to engage in health behaviors that are known to reduce the risk of early mortality in people with serious mental illness (SMI).

Methods:

This secondary data analysis was based on a cross-sectional study of 113 participants with SMI residing in New Hampshire. Ordinary Least Squares regressions were used to examine bivariate relationships between variables of interest.

Results:

Participants had a primary mental health diagnosis of major depressive disorder (37.2%), schizophrenia spectrum disorder (28.3%), bipolar disorder (29.2%), or posttraumatic stress disorder (5.3%). High levels of loneliness were associated with low levels of self-efficacy to manage chronic diseases (p=0.0001), as well as low levels of self-efficacy to manage psychological well-being (R²=.31; F= 9.49, p=0.0001; RMSE =1.66).

Conclusion:

Loneliness may serve as a barrier to healthy behaviors, and thus, contribute to early mortality among people with SMI.

Early mortality in people with serious mental illness (i.e., SMI; defined as schizophrenia spectrum disorder, bipolar disorder, or major depressive disorder) is one of the greatest unrecognized health disparities. People with SMI die up to 32 years earlier than the general population.[1] Despite decades of research on the health of people with SMI, their life expectancy is decreasing.[2, 1] Early mortality in people with SMI is most commonly attributed to poor health behaviors[3]—however, limited knowledge exists on the influence of loneliness on health behaviors.

People with SMI experience loneliness at a rate 2.3 times higher than the general population. [4] Interventions designed to address early mortality in people with SMI target poor health behaviors (i.e., smoking)—not loneliness. [5] This study explored associations between loneliness and efficacy to engage in healthy behaviors that are known to reduce the risk of early mortality in people with SMI.

Methods

Participants

Data were collected from 113 participants enrolled in a randomized controlled trial. The trial aimed at evaluating remote monitoring of psychiatric symptoms to impact functioning. Participants were aged 18 years+ who were receiving treatment from a community mental health center. Mental health diagnoses were confirmed through a review of an electronic medical record. Participants were randomly assigned to receive usual care or telehealth for one-year. Data were collected at baseline and at three, six, and 12-months.

Following the receipt of [blinded for review] institutional review board approval in 2014, recruitment for this study began in February 2015 and continued through December 2018. Service users from the study site who met eligibility criteria were provided a description of the intervention and study procedures. Competence to provide informed consent was established by a Mini Mental Status Exam score ≥24. [5] After informed consent was obtained, a baseline interview was scheduled.

Measures

Loneliness.

The three-item version of the self-report UCLA Loneliness Scale [6] was used to measure loneliness. Each of the items was rated on a four-point scale, with higher values indicating greater loneliness (range of scores= 0–16). Items assess feelings of being left out and isolated from others, feeling understood by others, and the frequency of talking to people. A total score was created as the sum of the items. Higher scores represented higher levels of loneliness. The original UCLA Loneliness Scale demonstrates good validity and reliability. [7]

Katz Comorbidity Questionnaire.

This self-report questionnaire has 36-items designed to determine the presence of medical conditions, whether or not treatment is being received, and whether or not the condition impacts functioning. [8] Each of the 36-items is rated on a three-point scale (i.e., one point for the presence of the medical conditions, one point if they receive treatment for that medical conditions, and one point if the medical conditions impacts functioning). Possible scores range from 0–45. We examined the presence of a medical condition. This questionnaire is an efficient alternative to medical record review. [9]

Self-Rated Abilities for Health Practices Scale.

Self-efficacy to manage chronic diseases was measured with the Self-Rated Abilities for Health Practices Scale (SRAHPS). [7] The SRAHPS is a 28-item scale that assesses confidence to execute health practices and has demonstrated reliability and validity with adults with disabilities. [10] The SRAHPS includes four subscales: self-efficacy to manage exercise, nutrition, responsible health practice, and psychological well-being. Each subscale score represents the sum of the component items. Higher scores represent higher levels of self-efficacy.

Analyses

We used descriptive statistics to characterize the sample at baseline. The regression analyses included data collected at 12-months when the UCLA Loneliness Scale was included. The three-items from the UCLA Loneliness Scale were combined into a subscale because the internal reliability was adequate (Cronbach’s alpha =.76). [10] Ordinary Least Squares regressions were used to model associations with loneliness score (the dependent variable). Only variables with p values less than or equal to .09 in the bivariates were used in the final model. We evaluated the final model with several goodness of fit indices: R², F-statistic and p-value, and RMSE of the model. Analyses were conducted with SAS, v.9.4. Missing data were addressed using list-wise deletion in the regression models.

Results

Participants in the sample were 52.2% female (n=59) and 51.3% never married (n=58). Most participants identified as white/Caucasian (63.7%, n=72) followed by African American (16.8%, n=19). Participants had a primary mental health diagnosis of: major depressive disorder (37.2%), schizophrenia spectrum disorder (28.3%), bipolar disorder (29.2%), or post traumatic stress disorder (5.3%). Chronic comorbid health conditions were common. Based on UCLA Loneliness Scale scores, most participants in this sample experienced loneliness “some of the time” or “often” (M=6.5, SD=1.95) (see Table 1).

Table 1.

Sociodemographic Characteristics of the Sample and Loneliness and Physical Health Conditions and Self-Efficacy to Manage Chronic Diseases in People with SMI

	Bivarates with Loneliness					Final Model
Sociodemographic	N (%) or M (SD)	Range	Beta	SE	p	Beta	SE	p
Sociodemographic
Age *include N % range and SD	44.3 (12.6)	(21–69)	0.00	0.01	0.76
Gender *N (%)
Male	54 (47.8)		0.02	0.37	0.97
Female	59 (52.2)
Race/Ethnicity
White	72 (63.7)		0.51	0.38	0.19
Black	19 (16.8)
Asian	2 (1.77)
Other/More than one race	20 (17.7)
Hispanic	19 (16.8)
Marital Status
Ever Married	58 (51.3)		0.20	0.37	0.59
Never Married	55 (48.7)
Education
Less than high school	25 (22.1)
High School or More	88 (77.9)		0.47	0.44	0.29
Primary Mental Health Diagnosis
Schizophrenia Spectrum Disorders	32 (28.3)		−0. 59		0.41	0.15
Bipolar Disorder	33 (29.2)		−0.02	0.41	0.97
Major Depressive Disorder	42 (37.2)		0.72	0.38	0.06	0.46	0.33	0.17
Posttraumatic Stress Disorder	6 (5.3)
Employed
No	90 (79.7)
Yes	23 (20.3)		−0. 12	0.46	0.80
Housing
Independent	91(8 0.5)		0.09	0.47	0.85
Supported or Supervised	22 (19.5)
Selected Self-Report Health Condition
Heart attack	9 (7.96)
Heart failure	9 (7.96)
Heart disease	1 (0.88)
COPD	17 (15.04)
Diabetes	16 (14.15)
Pain/Arthritis	15 (13.27)
% with any chronic health conditions	47 (41.6)		0.25	0.38	0.51
SRAHPS
Nutrition	20.4 (5.3)	(3–28)	−0.09	0.03	0.01	−0.03	0.04	0.40
Exercise	16.2 (7.1)	(0–28)	−0.10	0.02	<.0001	0.01	0.03	0.88
Psychological Well-being	15.8(6.6)	(1–28)	−0.16	0.02	<.0001	−0.15	0.03	<.0001
Responsible Health Practice	21.9 (5.1)	1–28	−0.10	0.04	0.01	0.00	0.04	0.94
Dependent Variable
UCLA Loneliness Scale	6.5 (1.95)	(3–9)

Note: SRAHPS=Self-Rated Abilities for Health Practices Scale; COPD=Chronic Obstructive Pulmonary Disorder

Bivariate analyses demonstrated that the SRAHPS subscales of self-efficacy to manage exercise, nutrition, responsible health practice, and psychological well-being were negatively related to the UCLA Loneliness Scale (see Table 1). There was a relationship between diagnosis of major depressive disorder and higher levels of loneliness. The presence of a medical condition was not associated with loneliness.

The independent variables in the final model included the four SRAHPS subscales and diagnosis of major depressive disorder. In the final model, self-efficacy to manage chronic disease and psychological well-being was negatively related to loneliness. The final model fit the data well, as illustrated by total variance explained of R²=.31 (F= 9.49, p=0.0001; RMSE =1.66). Results can be seen in Table 1.

Discussion

A high level of loneliness was associated with a low level of self-efficacy to manage chronic diseases and a low level of self-efficacy to manage psychological well-being. Interventions designed to promote self-efficacy to change health behaviors in people with SMI may not be enough to address the early mortality gap. The experience of loneliness in the general population is due to maladaptive social cognitions (e.g., dysfunctional and irrational beliefs, self-defeating thoughts), and interventions such as cognitive behavioral therapy for the general population target maladaptive social cognitions to reduce loneliness; [11] however, for people with SMI, loneliness is often caused by a combination of maladaptive social cognitions, impaired social skills, lack of opportunities to participate in social activities, and functional challenges. [12] Potentially, existing interventions designed to address early mortality through changing health behaviors should include social-cognitive, meta-cognitive, and other psychological intervention components that address maladaptive social cognitions surrounding loneliness.

This study is not without limitations. First, it is a secondary analysis of data from a study that was not focused on loneliness, and in which objective measurements of health were limited and chronic diseases were self-reported. Second, this study includes cross-sectional data; thus, examining causal relationships between loneliness and self-efficacy is not possible. Third, the small sample size did not allow for an examination of diagnostic differences in loneliness. Finally, the UCLA-R loneliness scale is a measure of chronic loneliness, thus it quantifies trait-level loneliness, not state-specific loneliness.

Public Health Implications

Similar to other social dimensions of health, loneliness is associated with a decline in self-efficacy to manage chronic disease and psychological well-being. Given that people with SMI experience more than twice the rate of loneliness compared to the general population, [4] policymakers and researchers should target loneliness as a mechanism to address early mortality in this population.

Policy implications:

The growing body of literature that demonstrates the importance of addressing loneliness in people with SMI should stimulate policymakers and researchers to target loneliness as a mechanism to address early mortality in people with SMI.

Biography

Karen Fortuna, PhD, MSW. Dr. Fortuna holds a doctorate in Social Welfare and a master’s degree in Social Work. Dr. Fortuna is an Assistant Professor of Psychiatry in the Geisel School of Medicine at Dartmouth College. Dr. Fortuna is using community-engaged research methods to develop and implement a peer-supported mobile health intervention. Dr. Fortuna was awarded an NIMH K01 award, a NARSAD Young Investigator Grants from the Brain and Behavior Foundation, the Alvin R. Tarlov & John E. Ware Jr. Award in Patient Reported Outcomes for her work, and the National Association of Gerontological Education in Social Work’s Faculty Achievement Award. Dr. Fortuna served on the International Standards Advisory Committee to develop the first-ever international accreditation standards for behavioral health care for older adults. Dr. Fortuna’s work can be seen in numerous book chapters on digital peer support, in Nature, Psychiatric Services, and Forbes Magazine. She currently serves on the International Editorial Board for the British Journal of Social Work, she is an editor of JMIR: Journal of Participatory Medicine, and she was recently invited to serve as a member of the American Psychiatric Association’s Smartphone App Expert Advisory Panel.

Joelle Ferron, PhD, MSW. Dr. Ferron is an Adjunct Professor of Psychiatry at the Geisel School of Medicine at Dartmouth. A social worker by training, Dr. Ferron’s focus has been developing and evaluating smoking cessation interventions for people with serious mental illness.

Cynthia L. Bianco, MS. Mrs. Santos has her Masters in Clinical Psychology and is a Senior Research Coordinator in the Department of Psychiatry at the Geisel School of Medicine at Dartmouth.

Meghan M. Santos, LISW. Mrs. Santos is a clinical social worker and is a Senior Research Coordinator in the Department of Psychiatry at the Geisel School of Medicine at Dartmouth.

Ashley Williams, BA. Ms. Williams is 2^nd Year Masters in Social Work Student at the University of New Hampshire in the School of Social Work.

Michael Williams, MPA. Mr. Williams has his Masters in Public Administration. Mr. Williams currently works in Washington State Health Authority.

George Mois, LISW. Mr. Mois is PhD student at University of Georgia in the School of Social Work. A social worker by training, Mr. Mois is studying assistive technologies for older adults.

Sarah I. Pratt, PhD. Sarah Pratt, PhD is an Associate Professor of Psychiatry at the Geisel School of Medicine at Dartmouth. A clinical psychologist by training, Dr. Pratt has devoted her career to developing and evaluating interventions to enhance the health and functioning of people with serious mental illness, including models of integrated care, fitness promotion, including the In SHAPE program, and smoking cessation.