Gratitude, Optimism, and Satisfaction with Life and Patient-Reported Outcomes in Patients Undergoing Hematopoietic Stem Cell Transplantation

Abstract

Objective:

Associations between positive psychological well-being (PPWB) and patient-reported outcomes (PROs, e.g., quality of life [QOL]) have yet to be studied extensively in patients with hematologic malignancies who are allogeneic hematopoietic stem cell transplant (HSCT) survivors, despite substantial evidence that PPWB impacts PROs of other medical populations.

Methods:

We conducted a secondary analysis of cross-sectional data examining the association of PPWB and PROs at day 100 post-transplant among 158 allogeneic HSCT recipients. Optimism, gratitude, life satisfaction, and PROs (i.e., QOL, anxiety, depression, and PTSD symptoms) were assessed using the Life Orientation Test-Revised, Gratitude Questionnaire, Satisfaction with Life Scale, Functional Assessment of Cancer Therapy-Bone Marrow Transplant, Hospital Anxiety and Depression Scale, and Post-Traumatic Stress Disorder (PTSD) Checklist-Civilian Version, respectively. We used linear and multivariate regressions for all analyses and controlled for patient factors.

Results:

Optimism was associated with better QOL (β=1.46; p<0.001) and lower levels of anxiety (β=−0.28; p<0.001), depression (β=−0.31; p<0.001), and PTSD (β=−0.58; p<0.001). Gratitude was associated with better QOL (β=1.11; p<0.001) and lower levels of anxiety (β=−0.21; p=0.001), depression (β=−0.14; p=0.021), and PTSD (β=−0.32; p=0.032). Finally, satisfaction with life was associated with better QOL (β=1.26; p<0.001) and lower levels of anxiety (β=−0.18; p<0.001), depression (β=−0.21; p<0.001), and PTSD (β=−0.49; p<0.001).

Conclusion:

Optimism, gratitude, and satisfaction with life were all associated with better QOL and lower levels of psychological distress in allogeneic HSCT survivors. These data support studies to harness PPWB as a therapeutic intervention for this population throughout HSCT recovery.

Introduction

Hematopoietic stem cell transplantation (HSCT) offers a potential cure for many patients with malignant and non-malignant hematological conditions.¹ However, the treatment is prolonged and strenuous,² often resulting in lengthy hospitalizations, numerous physical symptoms (e.g., diarrhea, fatigue), high levels of psychological distress (e.g. anxiety, depression) and low levels of positive psychological well-being (PPWB).^3,4 Due to the debilitating effects of increased psychological distress on patients throughout the HSCT trajectory, most psychosocial resources designed to improve well-being in the HSCT population have focused on alleviating distress symptoms.⁵ However, high levels of PPWB (e.g., gratitude) have been associated with numerous outcomes in patients undergoing HSCT (e.g., fewer days to neutrophil engraftment and lower mortality),⁶ with unique impacts on physical and psychological outcomes independent of the effects of psychological distress.⁷ Notably, most studies describing PPWB associations with HSCT outcomes were conducted more than a decade ago and did not focus on the 100-day post-HSCT recovery period when isolation restrictions begin to lessen and patients start to consider ways to re-engage with life activities they were unable to engage in during the initial post-transplant period. The broaden-and-build theory of positive emotions asserts that PPWB enhances cognitive and emotional resources for coping.⁸ Hence, more studies describing their relationships with patient-reported outcomes (PROs, e.g., quality of life [QOL]) are needed to inform the incorporation of PPWB tools into psychosocial resources for HSCT recipients.⁹

As with psychological distress, a useful way to conceptualize different PPWB constructs in the context of the HSCT population is by how they fluctuate over time. Historically, state constructs (e.g., positive affect) were understood to likely to change over shorter periods of time,¹⁰ while trait (or dispositional) constructs (e.g., optimism, life satisfaction, and gratitude) were expected to remain relatively stable over time.¹¹ However, our recent work on positive psychology interventions that promote PPWB in the HSCT population has suggested that trait constructs like optimism could change over time with intervention.¹² In qualitative studies, patients undergoing HSCT report experiencing diverse PPWB constructs including optimism, gratitude, and satisfaction with life.^4,13,14 Yet, most quantitative studies have exclusively examined optimism.⁹ Compared to other medical populations (e.g., patients with cardiovascular disease, diabetes),^15-19 there is minimal to no literature on the association between PPWB constructs beyond optimism and outcomes in the HSCT population. Hence, for this study, we describe the association between gratitude, life satisfaction, and sociodemographic and clinical factors, as well as PROs in HSCT to highlight opportunities to promote PPWB that could have broader effects on well-being and health for this vulnerable population.

Methods

Study Procedure

We completed a cross-sectional secondary analysis of baseline data collected at day 100 post-transplant from two studies assessing patients undergoing HSCT at the Dana-Farber Cancer Institute, Boston, MA from November 2021 to May 2023 – a prospective study assessing treatment adherence and a randomized controlled trial (RCT) of a psychological intervention (NCT05147311; data was collected at baseline prior to randomization). Both studies were approved by the Dana-Farber/Harvard Cancer Center Institutional Review Board (21-252 and 21-362, respectively), and all participants provided informed consent.

Participants

Inclusion criteria included English-speaking adult (≥18 years of age) patients with hematologic malignancies who had undergone HSCT approximately 100 days prior and could comprehend and complete surveys with little assistance. We excluded patients whose transplant oncologist believed a history of serious mental illness or comorbid disease would interfere with the patient’s ability to effectively complete informed consent and study procedures.

Sociodemographic and Clinical Data

Patients self-reported demographic information, including age, sex, race, ethnicity, relationship status, religious beliefs, education, and income at enrollment. We recorded disease, treatment, and complications (e.g., graft-versus-host disease [GVHD]) information from the electronic health record.

Positive Psychological Well-Being

We measured trait/dispositional optimism with the 10-item Life Orientation Test-Revised (LOT-R; Cronbach’s alpha=0.87); higher scores indicate greater optimism.²⁰ We measured trait/dispositional gratitude with the 6-item Gratitude Questionnaire (GQ-6; Cronbach’s alpha=0.81); higher scores indicate that an individual is more prone to experience gratitude in daily life.²¹ We measured trait satisfaction with life with the 5-item Satisfaction with Life Scale (SWLS; Cronbach’s alpha=0.88); higher scores indicate greater satisfaction with life.²²

Quality of Life

We used the 47-item Functional Assessment of Cancer Therapy-Bone Marrow Transplant (FACT-BMT; Cronbach’s alpha=0.92) to measure QOL.²³ The FACT-BMT consists of five subscales assessing well-being across physical, functional, emotional, social, and bone marrow transplant symptom domains. Higher scores signify better QOL.

Psychological Distress

We used the 14-item Hospital Anxiety and Depression Scale (HADS; Cronbach’s alpha=0.85) to measure anxiety and depression symptoms.²⁴ The HADS consists of two 7-item subscales which measure anxiety and depression symptoms, respectively; higher scores indicate worse mood symptoms. We used the 17-item Post-Traumatic Stress Disorder Checklist-Civilian Version (PCL-C; Cronbach’s alpha=0.90) to assess PTSD symptoms; higher scores indicate worse clinically significant PTSD symptoms.²⁵

Social Support

We used the 26-item Social Support Effectiveness Questionnaire (SSEQ; Cronbach’s alpha=0.94) to measure patients’ perception of social support.²⁶ Higher scores indicate higher levels of perceived social support.

Statistical Analysis

We used descriptive statistics (e.g., mean, standard deviation) for continuous data and proportions for categorical data to describe participant characteristics. Linear regression models were used to determine the unadjusted associations between PPWB constructs and sociodemographic (e.g., age) and disease (e.g., cancer type) factors. We then used multivariate regression models to assess the relationship between PPWB constructs and PROs, adjusting for patient, disease, and transplant factors that were found to be associated with PPWB in our unadjusted linear regression models to account for any potential confounders.^27,28 We used STATA 17.0 (StataCorp, College Station, TX) to complete all statistical analyses. A two-sided p-value < 0.05 was considered statistically significant.

Data Sharing

Please contact hermioni_amonoo@dfci.harvard.edu for original data.

Results

Participant Characteristics

Table 1 summarizes participant characteristics. The mean (SD) age of the 158 participants was 58.7 (13.3) years. Most participants were White (n=146; 92%), non-Hispanic (n=148; 94%), married or living with a partner (n=118; 75%), had received at least a college degree (n=86; 54%), had leukemia (n=98; 62%), and had received reduced intensity conditioning (n=103; 65%). Almost a tenth of participants (8.9%) reported an income of less than $25,000.

Table 1.

Participant Characteristics

	Total N=158
Age in years, mean (standard deviation [SD])	58.7 (13.3)
Sex, n (%)
Female	74 (46.8%)
Male	84 (53.2%)
Race, n (%)
Asian/Asian Background	5 (3.2%)
Black	2 (1.3%)
Native American/Indigenous	2 (1.3%)
White	146 (92.4%)
Other	1 (0.6%)
Missing	2 (1.3%)
Ethnicity, n (%)
Not Hispanic or Latinx	148 (93.7%)
Hispanic or Latinx	7 (4.4%)
Missing	3 (1.9%)
Relationship Status, n (%)
Married/Living Together	118 (74.7%)
Single	19 (12.0%)
Relationship/Not living together	7 (4.4%)
Separated/Divorced	6 (3.8%)
Widowed/Loss of Partner	6 (3.8%)
Missing	2 (1.3%)
Religion, n (%)
Agnostic	12 (7.6%)
Atheist	6 (3.8%)
Buddhist	4 (2.5%)
Catholic Christian	63 (39.9%)
Other Christian	42 (26.6%)
Hindu	1 (0.6%)
Jewish	11 (7.0%)
Other	16 (10.1%)
Missing	3 (1.9%)
Education, n (%)
< High School Diploma	3 (1.9%)
High School Diploma (GED)	21 (13.3%)
Some College	45 (28.5%)
College Degree	37 (23.4%)
Some Postgraduate/Professional Education	12 (7.6%)
Post-Graduate/Professional	37 (23.4%)
Missing	3 (1.9%)
Employment, n (%)
Employed	30 (19.0%)
Homemaker	2 (1.3%)
On Disability	68 (43.0%)
Retired	52 (32.9%)
Student	1 (0.6%)
Other	2 (1.3%)
Missing	3 (1.9%)
Income
<$25,000	14(8.9%)
$25,000-$49,000	18 (11.4%)
$50,000-$74,000	44 (27.8%)
$75,000-$99,000	28 (17.7%)
$100,000-$149,000	47 (29.4%)
Missing	7 (4.4%)
Disease Indication for Transplant, n (%)
Leukemia	98 (62.0%)
Lymphoma	15 (9.5%)
Myeloproliferative/Myelodysplastic Neoplasms	38 (24.1%)
Other	7 (4.4%)
Conditioning Regimen, n (%)
Myeloablative Conditioning	55 (34.8%)
Reduced Intensity Conditioning	103 (65.2%)
Total Body Irradiation, n (%)
No	134 (84.8%)
Yes	24 (15.2%)
Graft-versus-host Disease Prophylaxis, n (%)
Tacrolimus-Based	106 (67.1%)
Cyclophosphamide-Based	51 (32.3%)
None	1 (0.6%)
Transplant Length of Stay in days, mean (SD)	21.4 (7.7)
Acute Graft-versus-host Disease, n (%)
No	131 (82.9%)
Yes	27 (17.1%)

Patient Factors Associated with Optimism, Gratitude, and Life Satisfaction

We identified unadjusted associations between patient factors, optimism, gratitude, and life satisfaction (Supplemental Table 1). Older age was positively associated with optimism (β=0.076; p=0.022), gratitude (β=0.078; p=0.012), and life satisfaction (β=0.103; p=0.016). Having a disability status precluding employment was negatively associated with life satisfaction (β=−2.561; p=0.026). Level of perceived social support was positively associated with gratitude (β=0.137; p<0.001) and life satisfaction (β=0.198; p<0.001). Being agnostic or atheist was negatively associated with gratitude (β=−2.964; p=0.040). Having acute GVHD was negatively associated with gratitude (β=−2.202; p=0.042).

Associations between Optimism, Gratitude, and Life Satisfaction and QOL and Psychological Distress Symptoms

Tables 2-4 summarize multivariate analysis results for optimism, gratitude, and life satisfaction, and QOL and psychological distress symptoms. After adjusting for age, sex, race, religion, relationship status, transplant length of stay, social support, and study type, optimism was significantly associated with enhanced QOL (β=1.463; p<0.001) and reduced anxiety (β=−0.280; p<0.001), depression (β=−0.314; p<0.001), and symptoms of PTSD (β=−0.576; p<0.001). After adjusting for age, sex, race, religion, relationship status, employment, transplant length of stay, social support, and presence of GVHD at enrollment, gratitude was significantly associated with enhanced QOL (β=1.111; p<0.001) and reduced anxiety (β=−0.205; p=0.001), depression (β=−0.136; p=0.021), and symptoms of PTSD (β=−0.319; p=0.032). After adjusting for age, sex, race, religion, relationship status, employment, transplant length of stay, and social support, life satisfaction was significantly associated with enhanced QOL (β=1.258; p<0.001) and reduced anxiety (β=−0.177; p<0.001), depression (β=−0.209; p<0.001), and symptoms of PTSD (β=−0.485; p<0.001).

Table 2.

Multivariate Associations between Optimism, Quality of Life, and Psychological Distress Symptoms

	Optimism
Outcome	β	95% Confidence Interval		P-Value
Quality of Life	1.463	0.955	1.970	<0.001***
Post-Traumatic Stress Disorder	−0.576	−0.839	−0.313	<0.001***
Anxiety	−0.280	−0.376	−0.183	<0.001***
Depression	−0.314	−0.407	−0.221	<0.001***

Note. Multivariate regression models controlled for age, sex, race, religion, relationship status, transplant length of stay, social support, and study type. Each outcome was modeled separately. ***p<0.001

Table 4.

Multivariate Associations between Satisfaction with Life, Quality of Life, and Psychological Distress Symptoms

	Satisfaction with Life
Outcome	β	95% Confidence Interval		P-Value
Quality of Life	1.258	0.865	1.651	<0.001***
Post-Traumatic Stress Disorder	−0.485	−0.686	−0.284	<0.001***
Anxiety	−0.177	−0.261	−0.093	<0.001***
Depression	−0.209	−0.287	−0.130	<0.001***

Note. Multivariate regression models controlled for age, sex, race, religion, relationship status, employment, transplant length of stay, and social support. Each outcome was modeled separately. ***p<0.001

Discussion

This secondary data analysis from HSCT survivors who were 100 days post-allogeneic HSCT identified associations between sociodemographic factors and PPWB constructs. We also highlight significant associations between patients’ optimism, gratitude, and life satisfaction and greater levels of QOL and lower levels of anxiety, depression, and symptoms of PTSD. As one of the first studies to describe the relationships between optimism, gratitude, life satisfaction, and PROs at a critical timepoint (i.e., 100 days post-HSCT) in HSCT recovery, our findings highlight possible benefits of measuring and intentionally enhancing these PPWB constructs in the recovery trajectory of HSCT starting at the vulnerable 100-day post-HSCT period.

We found that age, religious affiliation, disability status, and perceived social support were associated with certain PPWB constructs but not others. For example, while older age was associated with greater levels of all three PPWB constructs, being on disability was associated with lower levels of life satisfaction but not optimism and gratitude. Little is known about the mechanisms by which sociodemographic factors like age impact optimism, gratitude, or life satisfaction in the HSCT population.²⁹ One possible explanation is that sociodemographic factors mediate the association between PPWB constructs, subjective well-being, and health.^30,31 For example, gratitude has been shown to be higher in older adults compared to middle age and younger adults, but at all ages, gratitude is associated with greater subjective well-being.³⁰ Further, individuals who identify as religious demonstrate higher levels of gratitude but comparable levels of subjective well-being as those without religious affiliation.³¹ Notably, some sociodemographic factors may also be influenced by other factors. For example, religious affiliation is influenced by culture, familial traditions, personal experiences, and geographic location.³² Since age and other sociodemographic factors also impact clinical outcomes (e.g., QOL, mortality), larger prospective studies are needed to fully characterize the relationships between sociodemographic factors, PPWB, and PROs over the course of the HSCT trajectory.

Prior work describing the association between PPWB and PROs in the HSCT population has focused primarily on optimism.^9,33,34 However, this study newly found that higher levels of gratitude and life satisfaction are also associated with higher levels of QOL and lower levels of psychological distress symptoms (i.e., symptoms of anxiety, depression, and PTSD). Although prior work in medical and non-medical populations have shown that gratitude or “giving thanks” and optimism are inversely associated with PTSD symptoms,^35,36 our study is among the first to show the association between PPWB and PTSD in the HSCT population. Prior work in this population has mostly focused on the association between PPWB constructs and anxiety and depression symptoms.^37,38 Since psychological distress symptoms persist throughout the HSCT care continuum, future work that examines the association between PPWB and psychological distress longitudinally may provide further insights on how best to leverage positive psychology interventions (i.e., interventions which cultivate PPWB through simple exercises like writing gratitude letters and performing small acts of kindness) to buffer against distress and enhance PPWB throughout the recovery of HSCT.³⁹

Clinical Implications

Understanding the relationships between patient factors and PPWB may assist in identifying patients who are resilient versus those who are at high risk for low PPWB to further tailor supportive interventions for these high-risk patients. Though some PPWB constructs, especially optimism, are considered stable, there is evidence that they may be modified through interventions.¹² Additionally, incorporating PPWB assessments into routine psychosocial assessments for HSCT recipients may unveil patients’ inherent traits that inform how they engage with treatment and decision-making throughout their care trajectory and recovery. Hence, assessing and enhancing PPWB may lead to important clinical benefits such as those seen in patients with chronic medical conditions like cardiovascular disease.¹⁵

Study Limitations

This study has several limitations worthy of consideration. First, this study took place at a tertiary academic cancer center with a primarily non-Hispanic White, married, and highly educated sample, limiting generalizability to ethnic minority or less educated populations. Second, we did not assess the effects of potential moderators, such as self-efficacy, on PPWB and QOL or psychological distress symptoms. Third, given the cross-sectional study design, we did not assess changes in optimism, gratitude, and life satisfaction over time or the chronicity of the illness experience from diagnosis through the transplant and recovery. Therefore, we could not draw conclusions about the causality or directionality of relationships. These PPWB traits, while typically viewed as stable over time, can change with time or situation,⁴⁰ such as over the course of the HSCT trajectory. Measures of state optimism,⁴⁰ gratitude, and life satisfaction, while not yet extensively validated, may prove to be useful in the HSCT population. Future longitudinal studies are warranted to assess the relationships between trait PPWB, state optimism or gratitude as distinct constructs from trait optimism or gratitude, and outcomes over the course of HSCT treatment and recovery.

Conclusions

In conclusion, this study adds to a growing body of evidence that diverse PPWB constructs are associated with PROs. Future prospective studies describing PPWB over the course of the HSCT trajectory from pre-transplant to long-term survivorship will inform interventions to best leverage the potential protective attributes of PPWB to promote overall well-being in the HSCT population.

Table 3.

Multivariate Associations between Gratitude, Quality of Life, and Psychological Distress Symptoms

	Gratitude
Outcome	β	95% Confidence Interval		P-Value
Quality of Life	1.111	0.532	1.689	<0.001***
Post-Traumatic Stress Disorder	−0.319	−0.610	−0.029	0.032*
Anxiety	−0.205	−0.319	−0.090	0.001**
Depression	−0.136	−0.252	−0.021	0.021*

Note. Multivariate regression models controlled for age, sex, race, religion, relationship status, employment, transplant length of stay, social support, and presence of graft-versus-host disease at enrollment. Each outcome was modeled separately. *p<0.05; **p<0.01; ***p<0.001