Yoga as a Complementary and Alternative Therapy in Children with Hematologic and Oncologic Disease

Abstract

A diagnosis of a hematologic or oncologic disease in a child can be stressful for the patient and the family. Yoga as an intervention has been reported to decrease stress in adults diagnosed with chronic disorders but few studies have been reported with children and their families. A convenience sample of patients diagnosed with cancer or a blood disorder (ages 7-17 years) and their caregivers was selected to participate in a single bedside yoga class. Participants were surveyed pre and post yoga with the Spielberger State Trait Anxiety Scale. Children were also surveyed pre and post yoga with the Wong-Baker Faces Pain Scale. Children had a significant decrease in pain post yoga but no change in anxiety. Adolescents and parents had a significant decrease in anxiety post yoga intervention.

Chronic illness in a child is a life-changing experience for the child and the family. The effects of chronic illness, such as cancer, can hinder quality of life by contributing to depression, stress, and anxiety in families (Commodari, 2010; Danhauer et al., 2017, Levine et al., 2017; Pasacreta, et al., 2006). Although traditional allopathic medicine may offer pharmacological interventions to assist in depression and stress, complementary and alternative medicine may provide alternative interventions to improve quality of life (Carlson et al., 2017; Jastrowski et al., 2013; Kanitz et al., 2013; Mustian et al., 2013; Sung et al., 2004; Thrane, 2013; Wurz et al., 2014). In 2007, the National Center for Complementary and Integrative Health (2007) ranked yoga in the top 10 treatment modalities for mind–body therapies to reduce stress. A recent meta-analysis reviewed the efficacy of yoga on physiological markers of stress. Populations reported included school-aged children and adolescents. The evidence suggested that yoga asana (type of yoga most often reported in which poses are used) practice may affect physiological parameters associated with stress such as decreased cortisol, decreased blood pressure, and a decrease in inflammatory markers such as cytokines. McNamara et al. (2016) reported on a study of patients diagnosed with cystic fibrosis (ages 7-20 years) who participated in a series of five to six yoga sessions over a 10-week period (n = 20). Their results suggested that anxiety improved not only for pre/post session evaluation but overall from Session 1 to the end of the program, although there was no reported significance in depressive or sustained anxiety symptoms. There was also a significant decrease in joint pain pre/post sessions. Moody et al. (2017) recently reported the use of yoga as a nonpharmacological intervention to reduce pain in children hospitalized with vaso-occlusive crisis. In their study, patients were offered daily yoga sessions, which included mindfulness, poses, breathing instructions, and guided relaxation, compared with a similar control group, which was offered relaxation music. Seventy-three participants completed one to eight sessions. In this sample, the greatest effect on pain was noted after the first session compared with baseline but was not sustained throughout the sessions. There was no significant reduction in anxiety or use of opioids for either group. The researchers reported a high accrual rate (88%), a high participation rate (95% of the sessions) and positive feedback suggesting that this is a feasible intervention for hospitalized children.

In adult populations with chronic illness, yoga has been reported to reduce stress, but there are limited data to support the use of yoga in children and their caregivers as an adjunctive measure (Carlson et al., 2017; Hainsworth et al., 2014; Pascoe et al., 2017, Telles et al., 2015).

Previous research studies in children with cancer and their guardians reported that individualizing yoga was feasible (Diorio et al., 2015) and decreased the need for pain medication, reduced anxiety and agitation, and improved mood and sleep (Diorio et al., 2015; Hooke et al., 2015). Other studies have measured yoga as an intervention for pain (Diorio et al., 2015; Geyer et al., 2011), but these studies measured multiple yoga sessions over several weeks using the PedsQL 4.0 cancer module. Wurz et al. (2014) suggested that twice weekly sessions with children with cancer in the outpatient setting was feasible; however, this was not supported in a recent report by Stein et al. (2019) using a weekly individualized program in this population over 10 weeks. Thygeson et al. (2010) reported the use of yoga as an alternative therapy to address anxiety for children and adolescents with hematologic and oncologic diseases, and parents before and after a single yoga class. In that sample population, the adolescents and parents experienced decreases in anxiety but the children did not. There is currently no research on yoga as a complementary therapy measuring pain with a specific pain scale before and after one single yoga session.

The purposes of this study were to determine if participation in one bedside yoga class 1) decreased anxiety and pain in hospitalized children diagnosed with a cancer or a blood disorder and 2) decreased anxiety for their parents.

Method and Participants

Institutional review board approval was granted by the University of California, San Diego for this project to be conducted at Rady Children’s Hospital, San Diego. A convenience sample of pediatric inpatients aged 7 to 17 years and/or their parents or guardians were recruited from July 2016 to October 2016 from the inpatient hematology/oncology unit at a tertiary care pediatric hospital in Southern California. Patients were eligible if they (a) spoke English, (b) were hospitalized on the inpatient hematology–oncology unit with an oncologic or hematologic diagnosis, (c) were able to give assent according to institutional guidelines, and (d) had parental consent to participate. Parents or guardians were eligible to participate provided they (a) spoke and read English, (b) had a child of any age who was hospitalized on the inpatient unit with a diagnosis of cancer or blood disorder, and (c) consented to participate in the study. Parents could participate in the study with or without their child since some hospitalized children are younger than the eligibility age for the study and others chose not to participate. Similarly, patients could participate in the study even if their parents chose not to participate.

Design

This study is modelled after an earlier report by Thygeson et al. (2010), with the addition of a pre and post administration of the Wong–Baker Faces Score to incorporate a measurement for pain. The yoga intervention was presented as a single session administered in the patient’s room lasting 20 to 60 minute depending on the participant’s level of interest and fitness. All the yoga classes were administered in English by one of the researchers, a certified yoga instructor. Depending on the patient’s physical abilities, interests, and previous experience with or exposure to yoga, different methods of yoga were administered based on the instructor’s assessment. These included deep breathing, meditation/guided imagery, mat yoga, chair yoga, and yoga in bed. For example, if a child liked animals, animal names were used to describe the pose. Each class included an element of breathing and ended with a final silence in relaxation pose, savasana.

Instruments

Previous research reported with yoga as an intervention for children with cancer recorded quality of life in terms of anxiety. For consistency in this study, the researchers considered extended anxiety as a tool to measure the effects of yoga as an intervention (Thygeson et al., 2010).

The Spielberger State Trait Anxiety Inventory (STAI) scale was used to measure psychological distress and the general sense of well-being for the children and adults 13 years of age and older. The inventory consists of 20 items that measure trait anxiety, which is a measure of relatively stable individual differences for anxiety and 20 items that measure state anxiety, which is more situational. The STAI S-Anxiety scale refers to state anxiety. In responding to the STAI S-Anxiety scale, subjects choose the number on the standard inventory form to the right of each item-statement that best describes the intensity of their feelings on a 4-point Likert scale ranging from 1 (not at all) to 4 (very much so). In responding to the STAI T-Anxiety scale, or trait anxiety, subjects are instructed to indicate how they generally feel by rating the frequency of their feelings of anxiety on a 4-point Likert scale ranging from 1 (almost never) to 4 (almost always). The range of scores for each scale is 20 to 80 points. The internal consistency (Cronbach’s α) estimate for the STAI ranges from 0.90 to 0.94. The mean score for adults ages 19 to 39 years is 36.54 (SD = 10.4) for males and 36.17 (SD = 10.61) for females. The mean score for participants over 13 years old is 39.45 (SD = 9.74) for males and 40.54 (SD = 12.86) for females (Spielberger et al., 1983).

The STAI Child Inventory (STAIC) was administered to the children ages 7 to 12 years. The STAIC S-Anxiety scale allows the child to report how they feel at a particular moment in time. The STAIC S-Anxiety is scored on a 3-point Likert-type scale with descriptive responses ranging from 1 (e.g. very calm) to 3 (e.g. not calm) to describe how the child feels at that moment in time. The STAIC T-Anxiety scale allows the child to report how they generally feel. The STAIC T-Anxiety scale has 20-items with each question beginning with “I feel” followed by a 3-point Likert-type scale ranging from 1 (hardly ever) to 3 (often). Scores range from 20 to 60. The internal consistency estimate (Cronbach’s α) for the STAIC is 0.82 to 0.87. The mean score on the STAIC for children is 31 (SD = 5.71) for males and 30.7 (SD = 6.01) for females. Higher scores are correlated with higher levels of anxiety (Spielberger et al., 1983).

There are currently no published studies available that evaluate yoga and its effects on pain in children with cancer and blood disorder. In this institution, pain was measured using the Wong–Baker Faces Scale™.The Wong–Baker Faces Scale™ was created to help children communicate about their pain in order to improve assessment and pain intervention (Wong-Baker FACES Foundation, 2016). This instrument is a 10-point scale used to measure pain with 0 being “No Hurt” and 10 being “Hurts Worst.” The Wong–Baker Faces Scale™ has been validated in hospitalized children ages 3 to 18 years undergoing various painful events including intravenous insertion, injections, spinal taps, and body pains (Wong & Baker 1998). Given that the current policy in this institution uses this instrument to measure self-reported pain in patients, it was used in this study to measure pain in patient’s pre and post yoga intervention for all ages.

Electronic medical records from children and their family members who consented to the study were reviewed for the following: age, diagnosis, gender, ethnicity, reason for admission, time since diagnosis, and number of days admitted as an inpatient during the current admission prior to the yoga treatment. Additionally, it was noted who chose to participate, reason for refusal to participate, observations from yoga instructor on participation and challenges, location of class, number of classes attended, length of time for class, previous participation in yoga or mindfulness classes.

Analysis

Responses were analyzed using descriptive statistics to describe the sample with analysis of open-ended responses from participant comments and evaluations for clarification. The primary aim of this study was to determine if there was a change in STAI scores or pain before and after yoga. A paired t test was used to determine if there was a difference between STAI scores before and after yoga. With n = 40 patients, this test had an 80% power at the α = .05 level to detect a difference between measures equal to 50% of the observed standard deviation. The Wilcoxon signed-rank test was also considered as a robust alternative if the assumptions of the t test were not met. A similar statistical framework was used to assess the difference in pain scores before and after yoga.

In addition, we used a logistic or linear regression model for the change in STAI or pain scores with demographic and physiological characteristics as covariates. For any patients with more than one recorded visit, linear mixed effects models were used and included a random intercept term to account for intrapatient correlation. Summary statistics (mean, standard deviation, quartiles, counts, and percentages) and plots were produced for all demographic and outcome variables. Analyses were conducted using the contemporary version of R (R Foundation for Statistical Computing, Vienna, Austria; http://www.r-project.org/).

Results

During the 6-month enrollment period, 15 patients participated in the study (Table 1). The mean age of the children was 11.8 years (SD = 3.28 years) and the ages ranged from 7 to 17 years. The patients were evenly distributed across gender. The majority of children were Hispanic (n = 8) or non-Hispanic Caucasian (n = 6). The patients’ diagnoses included acute lymphoblastic leukemia (n = 10), neuroblastoma (n = 2), acute myelogenous leukemia (n = 1), Burkitt lymphoma (n = 1), Wilms tumor (n = 1).

Table 1.

Patient and Parent Demographics.

	Parent	Patient
Age	n = 17	n = 15
	36.47 (20.00, 49.00)	11.80 (7.00, 17.00)
Gender	n = 18	n = 15
Female	15 (83.3%)	8 (53.3%)
Male	3 (16.7%)	7 (46.7%)
Race/ethnicity	n = 18	n = 15
Asian	1 (5.6%)	0 (0.0%)
Black	0 (0.0%)	1 (6.7%)
Caucasian (non-Hispanic)	7 (38.9%)	6 (40.0%)
Hispanic	9 (50.0%)	8 (53.3%)
Other	1 (5.6%)	0 (0.0%)

Note. Continuous data are presented as mean (range), and categorical data are presented as count (percentage).

Parents were approached during the same time frame and 18 parents/guardians agreed to participate. Parents in the study were primarily female (n = 15) and Hispanic (n = 9) or non-Hispanic Caucasian (n = 7).

For patients, the changes on the STAIC State and Trait scores before and after yoga were not statistically significant. The scores on the STAIC State decreased by 0.38 (95% confidence interval [CI]: −1.17 to 1.92, p = .584) after a single session of yoga, and STAIC Trait scores decreased by 1.88 (95% CI: −0.71 to 4.46, p = .130; Figure 1, Table 2). Neither the change in STAIC State or Trait scores became significant when controlling for subject gender and age. No significant effect was found for gender in the STAIC State or Trait scores, indicating that males and female reacted similarly to yoga. The change in pain scores after a single session of yoga was statistically significant (p = 0.045; Table 2).

Figure 1.

Boxplots of STAIC State and Trait scores in patients before and after yoga.

Table 2.

Means (Standard Deviations) of Outcomes Collected Pre and Post Yoga.

	Pre-yoga score	Post-yoga score	Difference (95% CI)	p Value
Parents
STAI State	38.17 (10.39)	28.54 (7.72)	9.62 (6.26, 12.99)	< 0.001
STAI Trait	37.46 (7.85)	33.42 (7.69)	4.04 (1.07, 7.02)	0.010
Patients
STAIC State	26.12 (3.87)	25.75 (4.46)	0.38 (−1.17, 1.92)	0.584
STAIC Trait	32.88 (8.90)	31.00 (9.55)	1.88 (−0.71, 4.46)	0.130
WBF Pain	1.36 (1.91)	0.71 (1.27)	0.64 (0.02, 1.27)	0.045

Note. CI = confidence interval; STAI = Spielberger State Trait Anxiety Inventory; STAIC = STAI Child Inventory; WBF Pain = Wong–Baker Faces Scale™ Pain Score. Significance is determined by a paired t test.

For parents or guardians, scores on the STAI State decreased by 9.62 (95% CI: 6.26 to 12.99, p < .001) after a single yoga session, and STAI Trait scores decreased by 4.04 (95% CI: 1.07 to 7.02, p = .01; Figure 2, Table 2). Changes in STAI State and Trait scores remained significant when controlling for age and gender in a linear mixed model. Fathers appeared to receive less benefit from yoga than mothers in STAI trait scores, but the sample size for male participants was too small in the parent group to conclude any inferences (n = 3).

Figure 2.

Boxplots of STAI State and Trait scores in parents before and after yoga.

Discussion

This study demonstrated that it was feasible for children and adolescents with a hematologic or oncologic diagnosis and their parents/guardians to participate in yoga. Similar to the results reported by Thygeson et al. (2010), children did not demonstrate a change on the STAI State and Trait scores post-yoga session. Results of this study also suggest that age and gender were not significant variables to predict change for the STAI State and Trait scores. Children and adolescents showed a significant reduction in pain after yoga measured by the Wong–Baker Faces scale™. It has been reported that 74.4% of patients experience pain within the first month of cancer therapy (Levine et al., 2017). Pain has traditionally been treated with pharmaceutical interventions (World Health Organization, 2012). In this study, patients regardless of age reported significant improvement in pain. This study also demonstrated that bedside yoga was feasible for parents/guardians attending to their children. One single yoga episode was significant enough for parents/guardians to reduce both their state and trait anxiety scores. Similarly, a study reported by Manne et al. (2016) demonstrated that in their sample parents/guardians of patients had a significant reduction in anxiety after using relaxation techniques. Furthermore, after teaching parents coping mechanisms using a DVD with focused breathing and progressive muscle relaxation, parents reported a significant level of stress reduction within one month of the intervention.

Limitations and Future Recommendations

This study was limited by convenience sampling on a small sample size of English speaking families on a single inpatient unit, the small number of male parents who participated, and the intervention being limited to a single yoga episode. Although the yoga instructor asked at the time of the session as to whether the participant had prior experience with yoga, the study did not record this measure as a variable. Previous exposure to yoga may have affected the findings. Generalizability may be affected by the inclusion criteria.

Wurz et al. (2014) reported that yoga may have a continued benefit to children diagnosed with cancer on both their quality of life and overall fitness and activity duration. Although their study was limited by a small sample size, results indicated that this twice a week intervention was feasible for this population. The introduction of a yoga program early in treatment with regular scheduled sessions may be of benefit to the child’s overall well-being as they complete their treatment protocols. Additionally, the finding that yoga may result in reduction of pain in children with cancer suggests yoga may be of benefit to all patients. The diagnosis of cancer in a child is stressful for the parents and the finding from this study suggest that other family members may also benefit from the use of yoga for both the state and trait of anxiety which suggests participation may benefit the family unit during this stressful time.

The actual number of sessions, timing of sessions, and specific type of sessions that are of benefit is yet to be determined (Danhauer et al., 2017). Given these results, future studies including scheduled yoga sessions offered to children and their families once diagnosed with a cancer or blood disorder, a tool to practice yoga at home, and early introduction in the treatment plan of yoga may yield long-term benefits in this population and should be investigated. Additionally, the paucity of reports using alternatives to traditional medicine in this population suggests that the use of other complementary and alternative medicine therapies in this population should be evaluated for potential efficacy to promote health and well-being for these children and their families.

Author Biographies

Julia S. Fukuhara, MSN, BS, RN, CPHON, PHN, RYT-200, is an oncology nurse, NASM certified person trainer, RYT-200 yoga instructor, Real Ryder and Schwinn cycling instructor. Her extensive knowledge of yoga and endurance training, and her understanding of anatomy from a medical perspective are mutually beneficial in her practices.

Judith O’Haver, PhD, RN, CPNP-PC, is an assistant professor and pediatric nurse practitioner. She holds faculty appointments at the Mayo Clinic College of Medicine, University of Arizona College of Medicine, Creighton University School of Medicine (Phoenix Regional Campus), and Arizona State University College of Nursing and Health Innovation.

James A. Proudfoot, MSc, is a senior statistician in the Department of Ophthalmology at the University of California, San Diego. When he is not analyzing data, he enjoys playing bass in a classic rock cover band and reading sci-fi and fantasy.

Jeanie M. Spies, MSN, RN-C, PNP, is an oncology nurse practitioner at Rady Children’s Hospital, San Diego. She is a graduate of the University of Arizona Fellowship in Integrated Medicine.

Dennis J. Kuo, MD, MS, is an associate clinical professor in the Division of Pediatric Hematology-Oncology at Rady Children’s Hospital, San Diego and the University of California, San Diego. He is also the medical director for the division’s cancer predispositions program and cancer survivorship program.