Higher Levels of Multiple Types of Stress Are Associated With Worse State Anxiety and Morning Fatigue Profiles in Patients Receiving Chemotherapy

Abstract

Background:

Anxiety and fatigue are common problems in patients receiving chemotherapy. Unrelieved stress is a potential cause for the co-occurrence of these symptoms.

Objectives:

Identify subgroups of patients with distinct state anxiety and morning fatigue profiles and evaluate for differences among these subgroups in demographic and clinical characteristics, as well as measures of global, cancer-specific, and cumulative life stress and resilience, and coping.

Methods:

Patients (n=1335) completed measures of state anxiety and morning fatigue six times over two cycles of chemotherapy. All of the other measures were completed prior to the second or third cycle of chemotherapy. Latent profile analysis was used to identify the state anxiety and morning fatigue profiles.

Results:

Three distinct joint profiles were identified: Low Anxiety and Low Morning Fatigue (59.0%); Moderate Anxiety and Moderate Morning Fatigue (33.4%); and High Anxiety and High Morning Fatigue class (7.6%). Patients in the two highest classes were younger, less likely to be married/partnered, and had a higher comorbidity burden. All of the stress scores demonstrated a dose response effect (i.e., as anxiety and morning fatigue profiles worsened, stress increased). Patients in two highest classes reported higher rates of emotional abuse, physical neglect, physical abuse, and sexual harassment.

Conclusions:

Over 40% of these patients experienced moderate to high levels of both anxiety and morning fatigue. Higher levels of all three types of stress were associated with the two highest profiles.

Implications for Practice:

Clinicians need to perform comprehensive evaluations of patients’ levels of stress and recommend referrals to psychosocial services.

Introduction

Fatigue occurs in 52%¹ to 85%² of oncology patients receiving active treatment. While most studies of oncology patients evaluated average fatigue, recent evidence suggests that diurnal variations exist in the severity of fatigue.^3–7 Equally important, a large amount of inter-individual variability exists in the severity of morning fatigue. For example, in our study that evaluated the severity of morning fatigue as a single symptom over two cycles of chemotherapy,⁴ four subgroups of patients with distinct morning fatigue profiles were identified (i.e., Very Low, Low, High, Very High).

Anxiety is an equally common symptom that occurs in 5% to 47% of oncology patients receiving chemotherapy.^8–12 Similar to morning fatigue, anxiety exhibits a large amount of inter-individual variability. In fact, in our previous study of anxiety as a single symptom,¹³ four distinct state anxiety severity profiles were identified (i.e., Low, Moderate, High, Very High).

The two studies of morning fatigue⁴ and state anxiety cited above¹³ used latent profile analysis (LPA) to identify the subgroups of patients with distinct symptom profiles. This analytic technique allows for an evaluation of demographic and clinical characteristics associated with a higher symptom burden. The common characteristics associated with membership in the Very High classes for both anxiety and morning fatigue included: younger age, having a higher comorbidity burden and having a poorer functional status.

Most of the studies that demonstrated positive associations between the occurrence or severity of average fatigue and anxiety used cross-sectional designs and correlational analyses.^{14, 15} For example, in a study of patients receiving chemotherapy,¹⁶ 83.0% and 84.7% reported anxiety and fatigue, respectively. Only one longitudinal study was identified that used latent transition analysis to evaluate for subgroups of oncology patients with distinct comorbidity patterns of depression, anxiety, and average fatigue.¹⁷ In this study, symptoms were assessed at the initiation and at 3 and 9 months after the initiation of psychosocial care. At the initiation of treatment, 67.7%, 54.3%, and 52.3% reported elevated levels of depression, anxiety, and fatigue, respectively. Three distinct classes of patients were identified across time (i.e., high levels of depression, anxiety, and fatigue; high levels of depression and anxiety and low levels of fatigue; low levels of all three symptoms). Poorer physical health was associated with membership in the class with persistent levels of all three symptoms.

Mechanistic hypothesis for the co-occurrence of fatigue and anxiety

From a mechanistic perspective, increases in inflammatory responses are associated with increases in anxiety,¹⁸ fatigue,^19,20 and stress.^21,22 Therefore, unrelieved stress is a potentially modifiable risk factor for both fatigue and anxiety in oncology patients receiving chemotherapy. While it is well documented that a diagnosis of cancer can be an extremely stressful event,²³ operational definitions of stress are inconsistent. For example, some studies focus on global perceived stress, which is how an individual tends to appraise stressful situations.^24,25 Other studies focus on the stressors themselves, including: childhood adversity,²⁶ cumulative life stressors,²⁷ non-cancer daily stressors faced by patients or survivors,²⁸ or the stress related to the cancer itself (referred to as cancer-related distress or traumatic stress symptoms).^29,30 Based on our previous findings of positive associations between stress and our worst anxiety¹³ and morning fatigue³¹ profiles, all three types of stress (i.e., global stress, cancer-related stress, and cumulative life stress) warrant evaluation in cancer patients. Equally important, an emerging body of evidence suggests that an individual’s level of resilience (i.e., one’s ability to respond to and/or adapt to stress),³² as well as the use of a variety of coping strategies,³³ may facilitate patients’ adaptation to stress and be associated with lower levels of anxiety and fatigue.

While cross-sectional and longitudinal studies of anxiety and fatigue as single symptoms provide useful information, none of them did a detailed characterization of inter-individual variability in the co-occurrence of these two symptoms over time and associated demographic and clinical characteristics. Equally important, while emerging evidence suggests that increased stress associated with cancer and its treatment contributes to increases in anxiety³⁴ and fatigue,³⁵ no studies have evaluated for associations between multiple types of stress and the co-occurrence of anxiety and morning fatigue. Therefore, the purposes of this study, using LPA, were to identify subgroups of oncology patients with distinct joint state anxiety and morning fatigue profiles and to evaluate for differences among these subgroups in demographic and clinical characteristics, as well as measures of three types of stress (i.e., global, cancer-specific, cumulative life stress), resilience, and coping. Based on our findings for the single symptoms,^4,13 we hypothesized that four joint state anxiety and morning fatigue profiles would be identified and that significant differences in the scores for all three types of the stress, as well as in resilience and coping, would be found among the classes.

Methods

Patients and Settings

This longitudinal study is part of a larger study that utilized the Theory of Symptom Management as its theoretical framework.³⁶ The patients’ symptom experience (i.e., state anxiety and morning fatigue) were the foci for this analysis. This experience was evaluated in the context of person (e.g., demographic characteristics) and health and illness (e.g., clinical characteristics, stress) contextual concepts.

Eligible patients were ≥18 years of age; had a diagnosis of breast, gastrointestinal, gynecological, or lung cancer; had received chemotherapy within the preceding four weeks; were scheduled to receive at least two additional cycles of chemotherapy; were able to read, write, and understand English; and provided written informed consent. Patients were recruited from two Comprehensive Cancer Centers, one Veteran’s Affairs hospital, and four community-based oncology programs. A total of 2234 patients were approached and 1343 consented to participate (60.1% response rate). The major reason for refusal was being overwhelmed with their cancer treatment. A total of 1335 patients who had sufficient data on the state anxiety and morning fatigue measures were included in the joint LPA.

Study Procedures

The study was approved by the Committee on Human Research at the University of California, San Francisco and by the Institutional Review Board at each of the study sites. Eligible patients were approached by a research staff member in the infusion unit, during their first or second cycle of chemotherapy, to discuss the study procedures and their interest in study participation. Written informed consent was obtained from all patients. Patients completed the anxiety and fatigue measures, in their homes using paper questionnaires, a total of six times over two cycles of chemotherapy (i.e., prior to chemotherapy (Assessments 1 and 4), a week following the administration of chemotherapy (Assessments 2 and 5), two weeks following the administration of chemotherapy (Assessments 3 and 6). The remaining questionnaires were completed at enrollment (i.e., prior to the second or third cycle of chemotherapy). Medical records were reviewed for disease and treatment information.

Instruments

Demographic and clinical characteristics

Patients completed a demographic questionnaire that obtained information on age, gender, ethnicity, marital status, living arrangements, education, employment status, and income. In addition, they completed the Karnofsky Performance Status (KPS) scale which is widely used to evaluate functional status in patients with cancer.³⁷ Patients rated their functional status using the KPS scale that ranged from 30 (I feel severely disabled and need to be hospitalized) to 100 (I feel normal; I have no complaints or symptoms).

The Self-Administered Comorbidity Questionnaire (SCQ) consists of 13 common medical conditions simplified into language that can be understood without prior medical knowledge.³⁸ Patients indicated if they had the condition; if they received treatment for it (proxy for disease severity) and if it limited their activities (indication of functional limitations). For each condition, the patient can receive a maximum of 3 points. The total SCQ score ranges from 0 to 39.

The Alcohol Use Disorders Identification Test (AUDIT) is a 10-item questionnaire that assesses alcohol consumption, alcohol dependence, and the consequences of alcohol abuse in the last 12 months. The AUDIT gives a total score that ranges between 0 and 40. Scores of 8 or more are defined as hazardous use and scores of 16 or more are defined as use of alcohol that is likely to be harmful to health.^39,40 In this study, its Cronbach’s alpha was 0.63.

The MAX-2 score was used to evaluate the toxicity of the various chemotherapy regimens based on the most toxic drug in the regimen.⁴¹ Scores can range from 0 to 1 with a higher score indicating a worse toxicity profile.

State Anxiety and Morning Fatigue Measures

The 20 items on the Spielberger State-Trait Anxiety Inventories (STAI-T and STAI-S) were rated from 1 to 4.⁴² The STAI-S measures a person’s temporary anxiety response to a specific situation or how anxious or tense a person is “right now” in a specific situation. The STAI-T measures a person’s predisposition to anxiety as part of one’s personality. A cut-off score of ≥32.2 indicates a high level of state anxiety. The Cronbach’s alpha for the STAI-S was 0.96. The six state anxiety scores were used in the joint LPA.

The Lee Fatigue Scale (LFS) was designed to assess physical fatigue and energy.⁴³ Total fatigue and energy scores were calculated as the mean of the 13 fatigue items and the 5 energy items, respectively. Higher scores indicate greater fatigue severity and higher levels of energy. Using separate questionnaires, patients rated each item based on how they felt within 30 minutes of awakening (i.e., morning fatigue) and prior to going to bed (i.e., evening fatigue). The LFS has established cut-off scores for clinically meaningful levels of fatigue (i.e., ≥3.2 for morning fatigue, ≥5.6 for evening fatigue).⁴⁴ The Cronbach’s alpha was 0.96 for morning fatigue. The six morning fatigue scores were used in the joint LPA.

Stress, Resilience, and Coping Measures

The 14-item Perceived Stress Scale (PSS) was used as a measure of global perceived stress according to the degree that life circumstances are appraised as stressful over the course of the previous week.⁴⁵ Each item (e.g., felt that you were unable to control the important things in your life) was rated on a 0 to 4 Likert scale (i.e., 0 = never, 1 = almost never, 2 = sometimes, 3 = fairly often, 4 = very often). In a probability sample drawn from the United States population,⁴⁶ scores of 18.8 and 20.2 were reported by male and female participants, respectively. Total PSS scores can range from 0 to 56. Its Cronbach’s alpha was 0.89.

The 22-item Impact of Event Scale-Revised (IES-R) was used to measure cancer-related distress.^47,48 Patients rated each item based on how distressing each potential difficulty was for them during the past week “with respect to their cancer and its treatment.” Each item was rated on a 0 (not at all) to 4 (extremely) Likert scale. Three subscales evaluate levels of intrusion, avoidance, and hyperarousal perceived by the patient. The total score can range from 0 to 88. Sum scores of ≥24 indicate clinically meaningful post-traumatic symptomatology and scores of ≥33 indicate probable post-traumatic stress disorder (PTSD).⁴⁹ The Cronbach’s alpha for the IES-R total score was 0.92.

The 30-item Life Stressor Checklist-Revised (LSC-R) is an index of lifetime trauma exposure (e.g., being mugged, sexual assault).⁵⁰ The LSC–R assesses whether each stressful event occurred; at what ages the events occurred; how many times each event occurred; how dangerous the event was; and whether the individual had an intense emotional reaction to the event(s). The total LSC–R score is obtained by summing the total number of events endorsed (range of 0 to 30). If the patient endorsed an event, the patient was asked to indicate how much that stressor affected their life in the past year from 1 (not at all) to 5 (extremely). These responses were summed to yield a total “affected” sum score. In addition, a PTSD sum score was created based on the number of positively endorsed items (out of 21) that reflect the DSM-IV PTSD Criteria A for having experienced a traumatic event.⁵¹

The 10-item Connor-Davidson Resilience Scale (CDRS) evaluates a patient’s personal ability to handle adversity (e.g., “I am able to adapt when changes occur”).⁵² Items are scored on a 5-point Likert scale (“not true at all” to “true nearly all of the time”). Total scores can range from 0 to 40, with higher scores indicating higher self-perceived resilience. The normative adult mean score in the United States is 31.8 (±5.4),⁵³ with an estimated minimal clinically important difference of 2.7.⁵⁴ Its Cronbach’s alpha was 0.90.

The 28-item Brief COPE was used to assess patients’ use of 14 coping strategies.⁵⁵ Patients rated their use of each coping strategy “since beginning chemotherapy.” Use of each coping strategy was evaluated using 2 items. Each item was rated on a 4-point Likert scale that ranged from 1 (“I haven’t been doing this at all”) to 4 (“I have been doing this a lot”). Scores for each coping strategy can range from 2 to 8, with higher scores indicating greater use of each strategy. Engagement coping strategies and their associated Cronbach’s alphas include active coping (0.75), planning (0.74), positive reframing (0.79), acceptance (0.68), humor (0.83), religion (0.92), emotional support (0.77), and instrumental support (0.77). Disengagement coping strategies and their associated Cronbach’s alphas include self-distraction (0.46), denial (0.72), venting (0.65), substance use (0.87), behavioral disengagement (0.57), and self-blame (0.73).

Data Analysis

LPA was used to identify subgroups of patients with distinct state anxiety and morning fatigue profiles. Using Mplus version 8.4.,⁵⁶ this LPA was done with the combined set of variables over time (i.e., using the STAI-S [state anxiety] and morning LFS [morning fatigue] scores obtained during the six assessments in a single LPA). This approach provides a profile description of these two symptoms with parallel profiles over time.

To incorporate expected correlations among the repeated measures of the same variable and cross-correlations of the series of the two variables (i.e., STAI-S and morning LFS scores), we included covariance parameters among measures at the same occasion and those that were one or two occasions apart. Covariances of each variable with the other at the same assessments were included in the model and autoregressive covariances were estimated with a lag of two with the same measures and with a lag of one for each variable’s series with the other variable. We limited the covariance structure to a lag of two to accommodate the expected reduction in the correlations that would be introduced by two chemotherapy cycles within each set of three measurement occasions and to reduce model complexity.⁵⁷ Model fit was evaluated to identify the solution that best characterized the observed latent class structure with the Bayesian Information Criterion (BIC), Vuong-Lo-Mendell-Rubin likelihood ratio test (VLMR), entropy, and latent class percentages that were large enough to be reliable.⁵⁸ Missing data were accommodated for with the use of the Expectation-Maximization algorithm.⁵⁹

Additional data were analyzed using SPSS version 28 (IBM Corporation, Armonk, NY). Differences among the state anxiety and morning fatigue classes in demographic and clinical characteristics, as well as stress, resilience, and coping scores were evaluated using parametric and nonparametric tests. Bonferroni corrected p-value of < .017 (i.e., .05/3 possible pairwise contrasts) was considered statistically significant for the pairwise contrasts.

Results

Latent Class Solution

The 3-class solution was selected because the BIC for that solution was lower than the BIC for the 2-class solution. In addition, the VLMR was significant for the 3-class solution, indicating that three classes fit the data better than two classes. Although the BIC was smaller for the 4-class than for the 3-class solution, the VLMR was not significant for the 4-class solution, indicating that too many classes were extracted. The three classes were named based on clinically meaningful levels of state anxiety and morning fatigue.

As shown in the Figure, 59.0% of the patients were in the Low Anxiety and Low Morning Fatigue class (Both Low); 33.4% were in the Moderate Anxiety and Moderate Morning Fatigue class (Both Moderate); and 7.6% were in the High Anxiety and High Morning Fatigue class (Both High). For the Both Low and Both Moderate classes, while anxiety scores remained relatively stable across the six assessments, morning fatigue scores increased slightly at assessments 2 and 5 (i.e., week following the administration of chemotherapy). For the Both High class, both symptom scores remained relatively stable over time.

Figure.

Trajectories of Joint Anxiety and Morning Fatigue Profiles

Demographic and Clinical Characteristics

Compared to the Both Low class, the other two classes were younger; more likely to be Hispanic, Mixed, or Other self-reported ethnicity; less likely to be married/partnered; more likely to have a lower annual household income; more likely to report back pain; and more likely to have received a neurokinin-1 receptor antagonist and two other antiemetics (Table 2). Compared to the Both Low class, the Both Moderate class was more likely to be female; more likely to live alone; less likely to be employed; and had a higher MAX 2 score. Compared to the Both Low class, the Both High class was less likely to exercise on a regular basis and more likely to self-report having lung disease, ulcer or stomach disease, or anemia. Differences among the latent classes in KPS score (i.e., Both Low > Both Moderate > Both High) and number of comorbid conditions, SCQ scores, and self-reports of depression (i.e., Both Low < Both Moderate < Both High) followed similar patterns.

Table 2.

Differences in Demographic and Clinical Characteristics Among the Anxiety and Morning Fatigue Latent Classes at Enrollment

Characteristic	Low Anxiety and Low Morning Fatigue (0) 59.0% (n = 788)	Moderate Anxiety and Moderate Morning Fatigue (1) 33.4% (n = 446)	High Anxiety and High Morning Fatigue (2) 7.6% (n = 101)	Statistics
	Mean (SD)	Mean (SD)	Mean (SD)
Age (years)	58.8 (11.9)	55.0 (12.9)	54.1 (11.4)	F = 17.35, p <.001 0 > 1 and 2
Education (years)	16.3 (3.0)	16.0 (3.0)	16.1 (3.3)	F = 1.51, p = .221
Body mass index (kg/m2)	26.0 (5.3)	26.3 (6.1)	26.7 (6.1)	F = 0.68, p = .508
Alcohol Use Disorders Identification Test score	2.9 (2.2)	3.1 (2.8)	2.9 (3.2)	F = 0.51, p = .600
Karnofsky Performance Status score	83.3 (11.6)	76.1 (12.2)	71.3 (11.3)	F = 78.26, p <.001 0 > 1 > 2
Number of comorbid conditions	2.2 (1.3)	2.6 (1.5)	3.2 (1.6)	F = 28.12, p <.001 0 < 1 < 2
Self-administered Comorbidity Questionnaire score	4.9 (2.8)	6.0 (3.4)	7.5 (3.9)	F = 42.21, p <.001 0 < 1 < 2
Time since diagnosis (years)	2.0 (3.7)	2.2 (4.4)	1.2 (2.0)	KW = 2.82, p = .244
Time since diagnosis (median, years)	0.42	0.42	0.44
Number of prior cancer treatments	1.6 (1.5)	1.7 (1.5)	1.6 (1.4)	F = 0.74, p = .479
Number of metastatic sites including lymph node involvement	1.2 (1.2)	1.3 (1.3)	1.2 (1.1)	F = 0.07, p = .931
Number of metastatic sites excluding lymph node involvement	0.8 (1.0)	0.8 (1.1)	0.7 (1.0)	F = 0.40, p = .674
MAX2 score	0.17 (0.07)	0.18 (0.08)	0.17 (0.08)	F= 4.60, p = .010 0 < 1
	% (n)	% (n)	% (n)
Female (% yes)	74.0 (582)	84.3 (376)	80.2 (81)	X2 = 18.06, p <.001 0 < 1
Self-reported ethnicity				X2 = 13.91, p = .031
White	71.3 (555)	68.6 (301)	60.4 (61)	NS
Asian or Pacific Islander	13.0 (101)	11.4 (50)	13.9 (14)	NS
Black	7.3 (57)	6.8 (30)	7.9 (8)	NS
Hispanic, Mixed, or Other	8.4 (65)	13.2 (58)	17.8 (18)	0 < 1 and 2
Married or partnered (% yes)	69.4 (539)	58.7 (257)	50.5 (51)	X2 = 23.14, p < .001 0 > 1 and 2
Lives alone (% yes)	18.0 (140)	26.4 (116)	27.7 (28)	X2 = 13.95, p < .001 0 < 1
Currently employed (% yes)	39.8 (310)	27.6 (122)	31.7 (32)	X2 = 19.11, p < .001 0 > 1
Annual household income				KW = 37.15, p <.001 0 > 1 and 2
Less than $30,000b	12.2 (85)	24.8 (101)	37.0 (34)
$30,000 to $70,000	20.5 (143)	23.1 (94)	16.3 (15)
$70,000 to $100,000	18.7 (130)	15.2 (62)	10.9 (10)
Greater than $100,000	48.6 (338)	36.9 (150)	35.9 (33)
Childcare responsibilities (% yes)	19.5 (151)	25.1 (108)	29.7 (30)	X2 = 8.59, p = .014 no significant pairwise contrasts
Elder care responsibilities (% yes)	6.6 (47)	10.6 (42)	6.3 (6)	X2 = 6.03, p = .049 no significant pairwise contrasts
Past or current history of smoking (% yes)	33.7 (262)	36.4 (159)	43.4 (43)	X2 = 3.94, p = .139
Exercise on a regular basis (% yes)	73.2 (568)	69.2 (299)	58.8 (57)	X2 = 9.48, p = .009 0 > 2
Specific comorbid conditions (% yes)
Heart disease	5.7 (45)	6.5 (29)	2.0 (2)	X2 = 3.14, p = .208
High blood pressure	30.6 (241)	28.7 (128)	33.7 (34)	X2 = 1.11, p = .575
Lung disease	9.9 (78)	12.3 (55)	17.8 (18)	X2 = 6.30, p = .043 0 < 2
Diabetes	8.2 (65)	9.0 (40)	14.9 (15)	X2 = 4.77, p = .092
Ulcer or stomach disease	3.7 (29)	5.4 (24)	11.9 (12)	X2 = 13.38, p = .001 0 < 2
Kidney disease	1.0 (8)	1.6 (7)	4.0 (4)	X2 = 5.64, p = .060
Liver disease	6.7 (53)	6.7 (30)	3.0 (3)	X2 = 2.19, p = .335
Anemia or blood disease	10.3 (81)	14.1 (63)	18.8 (19)	X2 = 8.37, p = 015 0 < 2
Depression	9.1 (72)	29.8 (133)	51.5 (52)	X2 = 151.42, p < .001 0 < 1 < 2
Osteoarthritis	12.2 (96)	11.7 (52)	12.9 (13)	X2 = 0.14, p = .932
Back pain	19.7 (155)	32.5 (145)	43.6 (44)	X2 = 42.64, p < .001 0 < 1 and 2
Rheumatoid arthritis	2.9 (23)	3.4 (15)	4.0 (4)	X2 = 0.42, p = .810
Cancer diagnosis				X2 = 9.34, p = .155
Breast	39.1 (308)	43.0 (192)	38.6 (39)
Gastrointestinal	32.7 (258)	26.0 (116)	31.7 (32)
Lung	10.5 (83)	13.0 (58)	15.8 (16)
Gynecological	17.6 (139)	17.9 (80)	13.9 (14)
Prior cancer treatment				X2 = 7.15, p = .308
No prior treatment	26 (199)	23.2 (101)	24.0 (24)
Only surgery, CTX, or RT	41.2 (315)	43.4 (189)	43.0 (43)
Surgery and CTX, or surgery and RT, or CTX and RT	21.3 (163)	17.7 (77)	18.0 (18)
Surgery and CTX and RT	11.5 (88)	15.6 (68)	15.0 (15)
Metastatic sites				X2 = 4.24, p = .645
No metastasis	32.1 (250)	33.6 (147)	30.0 (30)
Only lymph node metastasis	21.3 (166)	21.9 (96)	28.0 (28)
Only metastatic disease in other sites	22.6 (176)	19.2 (84)	19.0 (19)
Metastatic disease in lymph nodes and other sites	24.1 (188)	25.3 (111)	23.0 (23)
Received targeted therapy (% yes)	30.7 (238)	28.5 (124)	29.3 (29)	X2 = 0.69, p = .709
Cycle length				KW = 3.87, p = .144
14-day cycle	44.3 (347)	37.8 (167)	42.9 (42)
21-day cycle	48.3 (378)	55.0 (243)	51.0 (50)
28-day cycle	7.4 (58)	7.2 (32)	6.1 (6)
Emetogenicity of the CTX regimen				KW = 1.46, p = .481
Minimal/low	19.1 (150)	19.9 (88)	20.4 (20)
Moderate	62.8 (492)	57.5 (254)	63.3 (62)
High	18.1 (142)	22.6 (100)	16.3 (16)
Antiemetic regimen				X2 = 16.94, p =.009
None	7.9 (61)	6.1 (26)	5.2 (5)	NS
Steroid alone or serotonin receptor antagonist alone	20.3 (156)	21.5 (92)	17.5 (17)	NS
Serotonin receptor antagonist and steroid	50.6 (389)	43.6 (186)	42.3 (41)	NS
NK-1 receptor antagonist and two other antiemetics	21.2 (163)	28.8 (123)	35.1 (34)	0 < 1 and 2

Abbreviations: CTX, chemotherapy; kg, kilograms; KW, Kruskal Wallis; m², meters squared; n/a, not applicable; NK-1, neurokinin-1; NS, not significant; pw, pairwise; RT, radiation therapy; SD, standard deviation.

^aTotal number of metastatic sites evaluated was 9.

^bReference group.

Stress and Resilience Characteristics

As shown in Table 3, differences among the classes in PSS, IES-R intrusion, avoidance, arousal, and total scores, as well as LSC-R affected sum and PTSD sum scores followed a similar pattern (i.e., Both Low < Both Moderate < Both High). Compared to the Both Low class, the other two classes had higher LSC-R total scores. Differences among the classes in CDRS scores were as follows: Both Low > Both Moderate > Both High.

Table 3.

Differences in Stress and Resilience Measures Among the Anxiety and Morning Fatigue Latent Classes at Enrollment

Measuresa	Low Anxiety and Low Morning Fatigue (0) 59.0% (n = 788	Moderate Anxiety and Moderate Morning Fatigue (1) 33.4% (n = 446)	High Anxiety and High Morning Fatigue (2) 7.6% (n = 101)	Statistics
	Mean (SD)	Mean (SD)	Mean (SD)
PSS total score (range 0 to 56)	14.5 (6.2)	22.9 (6.4)	30.4 (7.1)	F = 429.69, p <.001 0 < 1 < 2
IES-R total score	13.6 (9.2)	23.8 (12.2)	37.7 (17.2)	F = 269.59, p <.001 0 < 1 < 2
(≥24 – clinically meaningful PTSD symptomatology)
(≥33 – probable PTSD)
IES-R intrusion	0.6 (0.5)	1.2 (0.7)	1.9 (0.9)	F = 249.11, p <.001 0 < 1 < 2
IES-R avoidance	0.8 (0.6)	1.1 (0.7)	1.5 (0.9)	F = 78.36, p <.001 0 < 1 < 2
IES-R hyperarousal	0.4 (0.4)	0.9 (0.6)	1.7 (1.0)	F = 315.92, p <.001 0 < 1 < 2
LSC-R total score (range 0–30)	5.6 (3.5)	6.6 (4.5)	7.7 (4.5)	F = 13.79, p<.001 0 < 1 and 2; 1 < 2
LSC-R affected sum (range 0–150)	9.9 (8.5)	14.1 (12.9)	19.6 (14.2)	F = 38.97, p <.001 0 < 1 < 2
LSC-R PTSD sum (range 0–21)	2.7 (2.7)	3.6 (3.4)	4.7 (3.6)	F = 22.58, p <.001 0 < 1 < 2
CDRS total score  (31.8 (±5.4) – normative range for the United States population)	32.3 (5.2)	27.5 (6.5)	24.2 (6.4)	F = 151.37, p <.001 0 > 1 > 2

Abbreviations: CDRS, Connor Davidson Resilience Scale; IES-R, Impact of Event Scale – Revised; LSC-R, Life Stressor Checklist-Revised; PSS, Perceived Stress Scale; PTSD, post traumatic stress disorder; SD, standard deviation.

^aClinically meaningful cutoff scores or range of scores.

Differences in Occurrence of Life Stressors

As shown in Table 4, compared to the Both Low class, the other two classes reported higher occurrence rates for emotional abuse, physical neglect, sexual harassment, physical abuse at ≥16 years, forced touching at ≤16 years, and having a serious physical or mental illness not related to cancer. Compared to the Both Low class, the Both Moderate class reported higher occurrence rates for forced sex at <16 years and having seen a robbery/mugging. Compared to the Low class, the Both High class reported higher occurrence rates for family violence in childhood, physical abuse at <16 years, and forced touching at ≥16 years. Significant differences among the classes in the occurrence of serious money problems were as follows: Both Low < Both Moderate < Both High.

Table 4.

Differences among the Anxiety and Morning Fatigue Latent Classes in the Percentage of Patients Exposed to Specific Stressors

Stressful Life Event	Low Anxiety and Low Morning Fatigue (0) 59.0% (n = 788)	Moderate Anxiety and Moderate Morning Fatigue (1) 33.4% (n = 446)	High Anxiety and High Morning Fatigue (2) 7.6% (n = 101)	Statistics
	% (n)	% (n)	% (n)
Interpersonal Violence, Abuse, and Neglect Stressors
Family violence in childhood	21.0 (134)	25.7 (81)	38.9 (28)	X2 = 12.38, p = .002 0 < 2
Emotional abuse	17.3 (111)	26.6 (85)	38.4 (28)	X2 = 23.77, p <.001 0 < 1 and 2
Physical neglect	3.1 (20)	6.6 (21)	12.3 (9)	X2 = 15.29, p <.001 0 < 1 and 2
Sexual harassment	15.3 (97)	21.8 (69)	28.2 (20)	X2 = 11.27, p = .004 0 < 1 and 2
Physical abuse - <16 years	12.6 (80)	15.5 (49)	24.7 (18)	X2 = 8.31 p = .016 0 < 2
Physical abuse - ≥16 years	10.7 (68)	17.4 (55)	21.4 (15)	X2 = 12.32, p = .002 0 < 1 and 2
Forced to touch - <16 years	8.8 (56)	15.6 (49)	19.2 (14)	X2 = 13.74, p = .001 0 < 1 and 2
Forced to touch - ≥16 years	4.4 (28)	7.6 (24)	13.9 (10)	X2 = 12.10, p = .002 0 < 2
Forced sex - <16 years	3.0 (19)	7.0 (22)	5.4 (4)	X2 = 8.15, p = .017 0 < 1
Forced sex - ≥16 years	5.4 (34)	7.6 (24)	10.8 (8)	X2 = 4.29, p = .117
Other Stressors
Been in a serious disaster	41.8 (266)	39.2 (125)	40.0 (30)	X2 = 0.64, p = .727
Seen serious accident	34.6 (220)	28.6 (92)	34.7 (26)	X2 = 3.66, p = .161
Had serious accident or injury	23.9 (151)	24.1 (76)	28.4 (21)	X2 = 0.75, p = .688
Jail (family member)	18.9 (121)	21.8 (69)	29.7 (22)	X2 = 5.12, p = .077
Jail (self)	5.3 (34)	8.8 (28)	10.8 (8)	X2 = 6.06, p = .048 no significant pairwise contrasts
Foster care or put up for adoption	2.2 (14)	2.5 (8)	4.1 (3)	X2 = 1.01, p = .603
Separated/divorced (parents)	19.8 (127)	24.7 (79)	25.7 (19)	X2 = 3.71, p = .157
Separated/divorced (self)	35.2 (226)	36.8 (118)	41.1 (30)	X2 = 1.08, p = .584
Serious money problems	15.0 (96)	24.9 (80)	40.5 (30)	X2 = 34.36, p <.001 0 < 1 < 2
Had serious physical or mental illness (not cancer)	15.9 (102)	23.0 (74)	28.0 (21)	X2 = 11.32, p = .003 0 < 1 and 2
Abortion or miscarriage	44.9 (218)	43.9 (118)	40.0 (24)	X2 = 0.55, p = .759
Separated from child	1.8 (11)	2.3 (7)	4.3 (3)	X2 = 2.01, p = .366
Care for child with handicap	4.0 (25)	3.9 (12)	2.8 (2)	X2 = 0.25, p = .882
Care for someone with severe physical or mental handicap	22.4 (141)	26.1 (82)	34.2 (25)	X2 = 5.68, p = .059
Death of someone close (sudden)	50.0 (318)	46.6 (145)	55.6 (40)	X2 = 2.14, p = .343
Death of someone close (not sudden)	79.7 (500)	79.1 (246)	72.6 (53)	X2 = 2.01, p = .365
Seen robbery/mugging	18.9 (121)	28.0 (89)	23.3 (17)	X2 = 10.28, p = .006 0 < 1
Been robbed/mugged	25.1 (160)	29.4 (92)	28.4 (21)	X2 = 2.13, p = .346

Differences in the Effects of Various Stressors

As shown in Table 5, compared to the Both Low class, the other two classes had higher effect scores for forced to touch at <16 years, being separated or divorced, having serious money problems, having had an abortion/miscarriage, caring for someone with a severe physical or mental handicap, the sudden death of someone close, and the death of someone close that was not sudden. Compared to the Both Low class, the Both Moderate class had higher effect scores for physical abuse at <16 years, having separated/divorced parents, and having a serious physical or mental illness other than cancer. Compared to the Both Low class, the Both High class reported higher effect scores for family violence in childhood, seeing a serious accident, and having a serious accident or injury. Compared to the other two classes, the Both High class had higher effect scores for emotional abuse and caring for someone with a severe physical or mental handicap.

Table 5.

Differences Among the Anxiety and Morning Fatigue Latent Classes in the Effect of Stressors on Life in the Past Year

Stressful Life Eventa	Low Anxiety and Low Morning Fatigue (0) 59.0% (n = 788)	Moderate Anxiety and Moderate Morning Fatigue (1) 33.4% (n = 446)	High Anxiety and High Morning Fatigue (2) 7.6% (n = 101)	Statistics
	Mean (SD)	Mean (SD)	Mean (SD)
Interpersonal Violence, Abuse, and Neglect Stressors
Family violence in childhood	1.3 (1.0)	2.1 (1.3)	2.4 (1.2)	KW = 11.09. p = .004 0 < 2
Emotional abuse	2.3 (1.4)	2.7 (1.3)	3.5 (1.0)	KW = 21.48, p <.001 0 and 1 < 2
Physical neglect	2.4 (1.5)	3.0 (1.2)	3.1 (1.4)	KW = 2.17, p = .337
Sexual harassment	1.4 (0.9)	1.6 (1.0)	1.7 (1.0)	KW = 4.29, p = .117
Physical abuse - <16 years	1.7 (1.1)	2.3 (1.4)	2.1 (1.1)	KW = 6.67, p = 036 0 < 1
Physical abuse - ≥16 years	1.7 (1.2)	1.9 (1.2)	2.4 (1.4)	KW = 3.98, p = .137
Forced to touch - <16 years	1.6 (1.1)	2.3 (1.4)	2.7 (1.7)	KW = 11.08, p = .004 0 < 1 and 2
Forced to touch - ≥16 years	1.6 (0.8)	2.4 (1.5)	1.7 (1.1)	KW = 3.91, p = .142
Forced sex - <16 years	1.7 (1.2)	2.2 (1.3)	2.0 (1.5)	KW = 1.86, p = .395
Forced sex - ≥16 years	1.6 (1.0)	2.0 (1.4)	1.9 (1.4)	KW = 1.36, p = .508
Other Stressors
Been in a serious disaster	1.3 (0.7)	1.5 (0.9)	1.8 (1.0)	KW = 25.31, p <.001 0 < 1 < 2
Seen serious accident	1.4 (0.8)	1.5 (0.9)	2.0 (1.2)	KW = 11.23, p = .004 0 < 2
Had serious accident or injury	1.5 (0.9)	1.7 (1.1)	2.1 (1.2)	KW = 9.18, p = .010 0 < 2
Jail (family member)	1.7 (1.3)	2.0 (1.4)	2.5 (1.6)	KW = 7.19, p = .027 no significant pairwise contrasts
Jail (self)	1.6 (1.1)	2.1 (1.4)	1.0 (0.0)	KW = 5.96, p = .051
Foster care or put up for adoption	2.2 (1.4)	2.9 (1.7)	1.7 (1.2)	KW = 1.63, p = .443
Separated/divorced (parents)	1.6 (1.0)	2.0 (1.2)	2.0 (1.5)	KW = 9.18, p = .010 0 < 1
Separated/divorced (self)	1.9 (1.3)	2.3 (1.4)	2.8 (1.5)	KW = 18.44, p <.001 0 < 1 and 2
Serious money problems	2.2 (1.5)	3.0 (1.7)	3.5 (1.5)	KW = 20.27, p <.001 0 < 1 and 2
Had serious physical or mental illness (not cancer)	2.2 (1.3)	2.7 (1.5)	2.6 (1.2)	KW = 6.62, p = .037 0 < 1
Abortion or miscarriage	1.4 (0.9)	1.7 (1.0)	2.1 (1.4)	KW = 12.55, p = .002 0 < 1 and 2
Separated from child	2.3 (1.6)	3.3 (1.4)	4.0 (1.7)	KW = 3.94, p = .139
Care for child with handicap	3.3 (1.4)	3.0 (1.3)	4.5 (0.7)	KW = 2.22, p = .329
Care for someone with severe physical or mental handicap	2.4 (1.4)	2.6 (1.5)	3.5 (1.4)	KW = 11.91, p = .003 0 and 1 < 2
Death of someone close (sudden)	2.0 (1.3)	2.3 (1.4)	2.9 (1.6)	KW = 18.31, p 001 0 < 1 and 2
Death of someone close (not sudden)	1.9 (1.2)	2.5 (1.4)	3.0 (1.5)	KW = 42.62, p <.001 0 < 1 and 2
Seen robbery/mugging	1.4 (0.9)	1.6 (1.1)	1.9 (1.3)	KW = 4.58, p = .101
Been robbed/mugged	1.5 (1.0)	1.7 (1.2)	2.1 (1.4)	KW = 2.60, p = .273

Abbreviation: SD, standard deviation.

^*Range = 1 “not at all” to 5 “extremely”

^aThese data are reported for those patients who reported the occurrence of the stressor (see Table 4).

Differences in Coping Strategies

In terms of the use of engagement coping strategies (Table 6), compared to the Both Low class, the other two classes had lower scores for active coping, positive reframing, and acceptance. In terms of the disengagement coping strategies, differences among the latent classes in the use of denial, venting, behavioral disengagement, and self-blame followed a similar pattern (i.e., Both Low < Both Moderate < Both High).

Table 6.

Differences Among the Anxiety and Morning Fatigue Latent Classes in the Brief COPE Subscale Scores

Subscalea	Low Anxiety and Low Morning Fatigue (0) 59.0% (n = 788)	Moderate Anxiety and Moderate Morning Fatigue (1) 33.4% (n = 446)	High Anxiety and High Morning Fatigue (2) 7.6% (n = 101)	Statistics
	Mean (SD)	Mean (SD)	Mean (SD)
Engagement coping strategies
Active coping	6.2 (1.6)	5.8 (1.7)	5.4 (1.6)	F = 15.40, p <.001 0 > 1 and 2
Planning	5.2 (1.9)	5.4 (1.7)	5.5 (1.7)	F = 1.31, p = .269
Positive reframing	5.6 (2.0)	5.2 (1.9)	4.8 (1.9)	F = 9.62, p <.001 0 > 1 and 2
Acceptance	6.9 (1.2)	6.4 (1.4)	6.2 (1.5)	F= 28.93, p <.001 0 > 1 and 2
Humor	4.4 (2.0)	4.2 (1.9)	4.2 (2.0)	F = 1.12, p = .328
Religion	5.0 (2.4)	4.9 (2.2)	5.1 (2.3)	F= .38, p = .686
Using emotional support	6.4 (1.7)	6.2 (1.6)	5.9 (1.8)	F = 4.60, p = .010 0 > 2
Using instrumental support	5.2 (1.8)	5.5 (1.7)	5.3 (1.8)	F = 3.16, p = .043 0 < 1
Disengagement coping strategies
Self-distraction	5.4 (1.8)	5.7 (1.5)	5.6 (1.5)	F = 5.00, p = .007 0 < 1
Denial	2.3 (0.8)	2.7 (1.3)	3.2 (1.7)	F = 46.38, p <.001 0 < 1 < 2
Venting	3.7 (1.6)	4.3 (1.6)	4.9 (1.6)	F = 42.63, p <.001 0 < 1 < 2
Substance use	2.2 (0.7)	2.3 (0.8)	2.3 (0.9)	F = 4.13, p = 0.16 0 < 1
Behavioral disengagement	2.1 (0.5)	2.3 (0.9)	2.8 (1.2)	F = 44.67, p <.001 0 < 1 < 2
Self-blame	2.5 (0.8)	3.2 (1.3)	4.3 (1.9)	F = 153.96, p <.001 0 < 1 < 2

Abbreviation: SD, standard deviation.

^aEach item was rated on a 4-point Likert scale that ranged from 1 (“I haven’t been doing this at all”) to 4 (“I have been doing this a lot”). Each coping strategy is evaluated using 2 items. Scores can range from 2 to 8 with higher scores indicating greater use of each of the coping strategies.

Discussion

This study extends our previous evaluation of inter-individual differences in the single symptoms of state anxiety¹³ and morning fatigue⁴ to identify distinct joint profiles for the two symptoms. While both analyses of the single symptoms identified four classes, when the two symptoms were combined, only three classes were identified. A comparison of the profiles across the three analyses suggests that for both state anxiety and morning fatigue, the single symptom LPAs identified a very small class with a more severe symptom profile (i.e., Very High state anxiety class (4.5%); Very High morning fatigue class (10.6%); (data not shown). Reasons for these different class solutions require additional investigation.

As illustrated in the Figure, while the anxiety scores for the three classes remained relatively stable over two cycles of chemotherapy, consistent with previous research,⁶⁰ for the Both Low and Both Moderate classes, morning fatigue scores increased in the week following the administration of chemotherapy (i.e., Assessments 2 and 5). One of the reasons that anxiety scores may remain relatively stable is that some patients perceive cancer as a constant threat to their survival.⁶¹ In contrast, the reason why patients in the Both High class had relatively stable morning fatigue scores over the two cycles of chemotherapy may be related to the development of a response shift in how they evaluated the severity of fatigue in relationship to previous cycles.⁶² Given the increases in morning fatigue, during the week following the administration of chemotherapy, clinicians need to educate patients to conserve energy during this period of treatment.

The occurrence rates for mild (9.5%), moderate (3.4%) and high (2.7%) anxiety among the general population⁶³ are lower than the 5% to 48% reported for oncology patients receiving chemotherapy.^8–12 In addition, while no studies evaluated for the prevalence of morning fatigue, 7% to 45% of the general population of the United States^64,65 and 52%¹ to 85%² of oncology patients report significant levels of average fatigue. Therefore, the prevalence rate of 41.0% for the co-occurrence of moderate to severe anxiety and morning fatigue is higher than the general population but congruent with current estimates for the single symptoms in patients with cancer. The remainder of the Discussion describes the risk factors associated with a worse anxiety and morning fatigue profile that clinicians can use to identify high risk patients (see Table 7).

Table 7.

Characteristics Associated with Membership in the Moderate and High Anxiety and Morning Fatigue Latent Classes

Characteristica	Moderate Anxiety and Moderate Morning Fatigue	High Anxiety and High Morning Fatigue
Demographic Characteristics
Younger age	■	■
More likely to be female	■
More likely to be Hispanic, Mixed, or other ethnicity	■	■
Less likely to be married or partnered	■	■
More likely to live alone	■
Less likely to be employed	■
More likely to have a lower annual household income	■	■
Less likely to exercise on a regular basis		■
Clinical Characteristics
Lower functional status	■	■
Higher number of comorbidities	■	■
Higher comorbidity burden	■	■
Higher MAX2 score	■
More likely to self-report lung disease		■
More likely to self-report ulcer or stomach disease		■
More likely to self-report anemia or blood disease		■
More likely to self-report depression	■	■
More likely to self-report back pain	■	■
More likely to receive an NK-1 receptor antagonist and two other antiemetics	■	■
Stress Characteristics
Higher Perceived Stress Scale score	■	■
Higher Impact of Event Scale-Revised total score	■	■
Higher Impact of Event Scale-Revised intrusion score	■	■
Higher Impact of Event Scale-Revised avoidance score	■	■
Higher Impact of Event Scale-Revised hyperarousal score	■	■
Higher Life Stressor Checklist-Revised total score	■	■
Higher Life Stressor Checklist-Revised affected sum score	■	■
Higher Life Stressor Checklist-Revised PTSD sum score	■	■
Lower Connor Davidson Resilience Scale total score	■	■
Higher Occurrence of Life Stressors
Family violence in childhood		■
Emotional abuse	■	■
Physical neglect	■	■
Sexual harassment	■	■
Physical abuse - <16 years		■
Physical abuse - ≥16 years	■	■
Forced touch – <16 years	■	■
Forced to touch - ≥16 years		■
Forced sex - <16 years	■
Serious money problems	■	■
Had serious physical or mental illness (not cancer)	■	■
Seen robbery/mugging	■
Higher Effect of Life Stressors
Family violence in childhood		■
Emotional abuse		■
Physical abuse- <16 years	■
Forced to touch - <16 years	■	■
Been in a serious disaster	■	■
Seen serious accident		■
Had serious accident or injury		■
Separated/divorced (parents)	■
Separated/divorced (self)	■	■
Had serious physical or mental illness (not cancer)	■
Abortion or miscarriage	■	■
Caring for someone with severe physical or mental handicap		■
Death of someone close (sudden)	■	■
Death of someone close (not sudden)	■	■
Use of Coping Strategies
Lower use of active coping	■	■
Lower use of positive reframing	■	■
Lower use of acceptance	■	■
Lower use of emotional support		■
Higher use of instrumental support	■
Higher use of self-distraction	■
Higher use of denial		■
Higher use of venting	■	■
Higher use of substance use	■
Higher use of behavioral disengagement	■	■
Higher use of self-blame	■	■

Abbreviations: NK-1, neurokinin 1; PTSD, post-traumatic stress disorder.

^aComparisons done with the Low Anxiety and Low Morning Fatigue class.

^{■ =}Indicates the presence of the risk factor compared to the Low Anxiety and Low Morning Fatigue class.

Global and Cancer-Specific Stress

Across the three classes, all of the scores for the measures of global (PSS) and cancer-specific stress (IES-R) demonstrated a dose response effect (i.e., as the anxiety and morning fatigue profiles worsened, stress scores increased). While no clinically meaningful cutoff score is evaluable for the PSS, the Both Moderate and Both High classes’ scores were higher than the average of 21.8 reported by patients with advanced gastrointestinal cancer⁶⁶ and the scores of 18.8 and 20.2 reported by male and female participants, respectively, in a probability sample drawn from the United States population.⁴⁶ In terms of cancer-specific stress, the IES-R scores for the Both Moderate class suggest clinically meaningful PTSD symptomatology and for the Both High class probable PTSD. Our results are consistent with a recent review that noted positive associations between anxiety and stress in patients with cancer and that 10% to 20% of patients meet the criteria for subsyndromal PTSD and 7.3% to 13.8% met the criteria for PTSD.²³

While no studies were found that evaluated for associations between stress and morning fatigue, in a recent review and meta-analysis of the effects of “forest bathing” (i.e., nature walks that involve immersive exposure in a forest or green spaces),⁶⁷ the authors noted that about 55% of the world’s population live in fast-paced urban settings and face challenges with the management of stress. They noted that this prolonged stress and associated increases in anxiety and fatigue result in physical and mental health concerns. Given that the results of this meta-analysis suggested that this intervention results in decreases in anxiety (NB – fatigue was not evaluated), it may be a low-cost intervention for oncology patients with high levels of anxiety, morning fatigue, and stress.⁶⁷ Mindfulness-based interventions may be equally effective.⁶⁸

Stressful Life Events (SLEs)

Previous reports found positive associations between CRF and/or anxiety and a variety of SLEs including the death of a loved one, having a serious illness or injury, experiencing a financial crisis, and having a history of childhood trauma.^69–72 In the current study, of the ten interpersonal violence, abuse and neglect stressors listed in Table 4, occurrence rates for six and eight of them were significantly higher in the Both Moderate and the Both High classes, respectively. Specifically, a number of adverse childhood experiences (ACEs) occurred more often and had greater effects in the Both Moderate and Both High classes. These associations are consistent with studies of women with breast cancer that found that patients who experienced a history of childhood abuse and neglect reported higher levels of fatigue and stress.^25,73

Additional support for associations between the co-occurrence of anxiety and CRF with ACEs comes from studies of patients with chronic fatigue syndrome (CFS). For example, in a population-based study,⁷⁴ compared to non-fatigued controls, individuals with CFS reported higher childhood trauma scores. In addition, exposure to childhood trauma was associated with a 3-to 8-fold increased risk for CFS, as well as higher levels of state anxiety and symptoms associated with PTSD. Given the high IES-R scores and the high occurrence rates for a number of ACEs in the current study, clinicians may want to initiate screening programs. The Pediatric ACEs and Related Events Screener (PEARLS),⁷⁵ that can be completed by adults, can be used to identify patients for childhood maltreatment so that appropriate referrals can be made to psychosocial services.

Resilience

The CDRS scores for the Both Moderate and Both High classes were below the normative score for the population of the United States. This finding suggests that patients with moderate to high levels of both anxiety and morning fatigue have a decrease in their ability to cope with stress.⁷⁶ As noted in one review,⁷⁶ greater resilience results in less psychological distress and better adjustment to cancer.

However, the relationships between various ACEs and resilience warrants additional investigation. For example, and consistent with our findings, in a study of healthy adults,⁷⁷ higher levels of childhood personal trauma were associated with lower levels of resilience. In contrast, in the study of cancer survivors,⁷⁸ higher levels of lifetime trauma exposure were associated with higher levels of resilience. The authors hypothesized that exposure to lifetime trauma may modify how individuals are affected by recent adverse events.

Coping

Patients in the Both Moderate and Both High classes were more likely to use the disengagement coping strategies of denial, venting, behavioral disengagement, and self-blame. The use of these coping strategies is not surprising given that in a recent review of the efficacy of coping skills interventions,⁷⁹ problem-focused, as opposed to emotion-focused, coping interventions resulted in significant reductions in anxiety. In terms of fatigue, in a study of older oncology patients,⁸⁰ higher levels of average fatigue were associated with the use of disengagement coping strategies. This association may be related to higher levels of anxiety, stress, feelings of helplessness, and increased awareness of physical limitations.⁸¹

Demographic and Clinical Characteristics

Two of the clinical characteristics associated with membership in the Both Moderate and Both High classes were a lower functional status and a higher comorbidity burden. Of note, both classes had KPS scores in the 70s, which equates with the “patients can care for themselves but are unable to carry out normal activities or perform active work”.⁸² This finding correlates with the relatively low percentages of patients in these two classes who reported that they exercised on a regular basis. While higher levels of anxiety and stress were associated with lower levels of physical activity,⁸³ active participation in programs that are designed to increase physical activity resulted in decreases in anxiety in oncology patients.^84,85 While not evaluated in terms of its efficacy for morning fatigue, exercise is the only evidenced-based recommendation for CRF.⁸⁶

Linked to a poorer functional status is the higher comorbidity burden in the Both Moderate and Both High classes. As noted in one review,⁸⁷ patients with higher levels of anxiety have a disproportionately higher number of comorbid conditions. While not assessed for morning fatigue, similar associations were found between CRF and a higher comorbidity burden.⁸⁸ Given that most patients with cancer have one or more comorbid conditions,⁸⁹ oncology clinicians need to work collaboratively with primary care providers to ensure effective management of these conditions and associated symptoms.

Mechanistic considerations

Undoubtedly, the mechanisms that underlie the co-occurrence of state anxiety and morning fatigue and associated increases in stress are extremely complex. However, evidence suggests that increases in inflammatory responses are associated with increases in anxiety,¹⁸ fatigue,^19,20 and stress.^21,22 In addition, in a review of neuroimaging studies,⁹⁰ changes in the function and/or structure of key brain regions (e.g., amygdala, hypothalamus, basal ganglia) were associated with greater anxiety, PTSD, and stress in oncology patients. Of note, it is well established that exposure to ACEs increases the risk for a major anxiety disorder in adulthood.^91,92 This relationship is attributed to increases in inflammation and allostatic load.⁹³ In terms of morning fatigue, recent work from our group found that a number of inflammatory pathways (e.g., nuclear factor kappa-light-chain enhancer of activated B cells, natural killer cell-mediated cytotoxicity) were associated with higher levels of morning fatigue.¹⁹ Additional research is warranted to determine the mechanisms associated with higher levels of state anxiety, morning fatigue, and specific types of stress.

Limitations

Several limitations warrant consideration. Given that this study is the first to identify subgroups of patients with distinct anxiety and morning fatigue profiles, despite the large sample size, the differences among the classes in various risk factors warrant confirmation in future studies. Given that the major reason for refusal was that patients were overwhelmed with their cancer treatment, these findings may underestimate patients’ symptom burden and stress. In addition, a comparable analysis is warranted using scores for evening fatigue. Because the patients were recruited during their first or second cycle of chemotherapy and followed for only two cycles, future studies need to enroll patients prior to the initiation of treatment and follow them for a longer period of time. Equally important, because the stress measures were only done at enrollment, no causal relationships can be demonstrated and warrant evaluation in future studies.

Implications for Clinical Practice and Research

In addition to the clinical implications noted throughout the Discussion, findings from this study can assist clinicians in identifying patients at increased risk for clinically meaningful levels of anxiety and morning fatigue. Based on the findings from this study, clinicians need to perform a comprehensive evaluation of patients’ levels of anxiety, morning fatigue, and stress, particularly the occurrence and impact of ACEs. Based on the findings related to coping and resilience, these patients may have difficulty garnering the resources that they need to adapt to the stressors associated with cancer and its treatments. Trauma-informed care is likely to benefit these patients (see www.healthcaretoolbox.org).⁹⁴ Referrals to psycho-oncology specialists may be needed to provide individually tailored interventions to support these patients.

Additional research is warranted on the relationships between changes in anxiety and morning fatigue and various types of stress. Additional studies on the efficacy of a variety of coping strategies, with or without stress management and exercise interventions, to be able to decrease anxiety, morning fatigue, and/or stress are needed to inform the prescription of individualized, evidenced-based interventions. In addition, the mechanisms that underlie the co-occurrence of these two symptoms warrant evaluation.

Table 1.

Latent Profile Solutions and Fit Indices for One through Four Classes for State Anxiety and Morning Fatigue Scores over Six Assessments:

Model	LL	AIC	BIC	Entropy	VLMR
1 Class	−38208.03	76532.05	76833.46	n/a	n/a
2 Class	−37280.12	74702.23	75071.20	0.85	1855.82a
3 Classb	−36991.66	74151.33	74587.85	0.87	576.90a
4 Class	−36780.82	73755.64	74259.71	0.82	NS

Baseline Entropy and VLMR are not applicable for the one-class solution.

^ap < .00005.

^bThe 3-class solution was selected because the BIC for that solution was lower than the BIC for the 2-class solution. In addition, the VLMR was significant for the 3-class solution, indicating that three classes fit the data better than two classes. Although the BIC was smaller for the 4-class than for the 3-class solution, the VLMR was not significant for the 4-class solution, indicating that too many classes were extracted.

Abbreviations: AIC, Akaike’s Information Criterion; BIC, Bayesian Information Criterion; LL, log-likelihood; n/a, not applicable; NS, not significant; VLMR, Vuong-Lo-Mendell-Rubin likelihood ratio test for the K vs. K-1 model.