Heart rate variability and psychosocial symptoms in adolescents and young adults with cancer.

BACKGROUND: Heart Rate Variability (HRV) is a valid, scalable biomarker of stress. We aimed to examine associations between HRV and psychosocial outcomes in adolescents and young adults (AYAs) with cancer.
METHODS: This was a secondary analysis of baseline data from a randomized trial testing a resilience intervention in AYAs with cancer. Two widely used HRV metrics, the standard deviation of normal to normal beats (SDNN) and root mean square of successive differences (RMSSD), were derived from electrocardiograms. Patient-reported outcome (PRO) survey measures included quality of life, anxiety, depression, distress, and resilience. Linear regression models were used to test associations between HRV and PRO scores. The Wilcoxon rank sum test was used to test differences in median HRV values among participant subgroups.
RESULTS: Among the n = 76 patients with available electrocardiograms, the mean age was 16 years (SD 3y), 63% were white, and leukemia/lymphoma was the most common diagnosis. Compared to healthy adolescents, AYAs with cancer had lower median HRV (SDNN [Females: 31.9 (12.8-50.7) vs 66.4 (46.0-86.8), p<0.01; Males: 29.9 (11.5-47.9) vs 63.2 (48.4-84.6), p<0.01]; RMSSD [Females: 28.2 (11.1-45.5) vs 69.0 (49.1-99.6), p<0.01; Males: 27.9 (8.6-48.6) vs 58.7 (44.8-88.2), p<0.01]). There was no statistically significant association between PRO measures and SDNN or RMSSD in either an unadjusted or adjusted linear regression models.
CONCLUSION: In this secondary analysis, we did not find an association between HRV and psychosocial PROs among AYAs with cancer. HRV measures were lower than for healthy adolescents. Larger prospective studies in AYA biopsychosocial research are needed.

Introduction

Psychological distress is prevalent during adolescent development, and adolescents and young adults (AYAs) with serious illness like cancer experience even higher rates of depression, post-traumatic stress, and suicide [1-3]. Poor mental health directly translates to adverse cancer-related outcomes, including increased physical symptom burden [4] and poorer survival [5, 6]. In contrast, positive psychological well-being has been associated with protective effects, including lower risk of relapse and cancer-related mortality [7]. Identifying the mechanisms underlying these biobehavioral connections is a growing focus of investigation and may offer a novel tool for symptom-based risk stratification and intervention.

Psychosocial and environmental stressors can activate a cascade of signaling pathways that have overlapping behavioral and biologic implications. For example, a perceived threat (such as a new cancer diagnosis) triggers the hypothalamic-pituitary-adrenal (HPA) axis to ultimately secrete glucocorticoids into circulation, which have well-documented metabolic and immunomodulatory effects that could influence cancer-related outcomes [8]. The autonomic nervous system (ANS) is another key physiologic mediator of the relationship between patient experience and outcomes in biobehavioral oncology. Activation of the parasympathetic and sympathetic branches of the ANS has been associated with symptoms of anxiety, depression, and fatigue in patients with cancer [9, 10]. Conversely, interruption of sympathetic adrenergic signaling using β-blocker medications is associated with improved metastatic and inflammatory biomarkers, as well as lower cancer-specific mortality in patients with breast cancer [11-13].

Heart rate variability (HRV), which is the physiologic fluctuation between successive heartbeats, is a surrogate marker of ANS status and has been widely applied in both social science and biomedical settings [14, 15]. HRV is measured by capturing electrical signals from the heart using electrocardiograms (ECGs), Holter monitors, pulse oximeters, or newer wearable devices and smartwatches. There are published guidelines for measuring and interpreting HRV [16], and although gold standard for HRV measurement is a 24-hour recording, components of HRV can reliably be determined from short (1 or 5 minute) or ultra-short (10 second) monitoring [17, 18].

Lower HRV, signaling reduced autonomic flexibility, is associated with important outcomes including infection, mortality [19], depression and anxiety disorders [15, 20], and psychosocial stress [21]. Adult patients with cancer are known to have lower HRV compared to healthy controls, with this difference more pronounced in those with advanced cancer [22]. Reduced HRV has also been reported in young children with leukemia [23, 24] and dysautonomia indexed by HRV has been documented in adult survivors of childhood cancer [25]. Importantly, the degree of HRV reduction may predict cancer-related fatigue and chronic pain [26, 27] as well as cancer progression and overall survival [28]. The precise role of the ANS in these conditions is incompletely understood, but the direct interplay between psychosocial risk factors, the ANS, and inflammation provide a biologically plausible explanation for the connection between mental and physical symptoms in cancer [29].

To date, nearly all oncology research using HRV has focused on adults, aside from a small number of studies in younger patients with leukemia [23]. There have been no explicit studies investigating the utility of HRV as a potential stress biomarker in older AYA oncology patients. Given the extreme physical and psychosocial stress associated with cancer diagnosis and treatment in adolescents and young adults [1-3], HRV measurement may be particularly helpful as a simple, non-invasive assessment of patient wellbeing and function. Here, we present a cross-sectional secondary analysis of baseline HRV and patient-centered psychosocial outcomes from a randomized trial among AYAs with cancer.

Materials and methods

This was a secondary analysis of a completed randomized controlled trial (RCT) testing the Promoting Resilience in Stress Management (PRISM) intervention in AYAs with cancer (NCT02340884). We conducted a cross-sectional analysis of available baseline data prior to receipt of the PRISM intervention [30]. Heart rate variability (SDNN and RMSSD) was our independent variable, and patient-reported psychosocial measures served as our dependent variable. We hypothesized that lower HRV would be associated with adverse psychosocial states (higher anxiety, depression, distress; and lower quality of life and resilience). The Seattle Children’s Institutional Review Board approved this study.

Participants and setting

The Phase II PRISM Randomized Trial was conducted at a single institution (Seattle Children’s Hospital) from January 2015 to October 2016. Eligible patients were English speaking, aged 12–25 years, and diagnosed with a new or relapsed/refractory cancer requiring treatment. Demographic and disease-related variables were requested in surveys and collected from the medical record of consented participants. Of the n = 92 AYAs with baseline patient reported psychosocial measures, n = 76 also had available ECGs to derive HRV and were included in the present analysis. Of the n = 16 participants with missing ECGs, 7 did not require ECGs for their specific treatment plan, and there was no documented reason for omission in 9 cases.

Measures

Heart rate variability [31]

HRV was derived using 10 second 100Hz ECGs obtained at the time of cancer diagnosis or relapse as part of routine clinical care. These screening ECGs were obtained as part of patients’ medical care and were not part of the clinical trial. Individual paper ECGs were retrospectively extracted from the medical record and scanned into a digital format using a high-resolution scanner. These digital ECG files were then uploaded into an image processing software (WebPlotDigitizer [32]) to extract R-R intervals from the ECG tracing images [33]. R-waves were first identified using the software algorithm and then manually reviewed for accuracy and artifacts. The discrete R-R intervals (in milliseconds) could then be identified from the image and converted to their numerical form. Using the open-source R software, RHRV [34], we then derived the two most widely used time domain parameters: standard deviation of normal to normal beats (SDNN) and root mean square of successive differences (RMSSD) per published guidelines [16, 35]. These two HRV measures were chosen because 1) they are commonly used in behavioral and clinical research; 2) they can be derived from standard ECGs [17] and 3) published age-appropriate normative values exist [36].

Psychological variables

All participants were invited to complete a survey consisting of AYA age–validated instruments upon enrollment and received a $25 gift card upon survey completion.

Pediatric Quality of Life (PedsQL) Generic and Cancer Module teen reports [37, 38]. The PedsQL 4.0 Generic and 3.0 Cancer Modules include a combined total of 50 items evaluating QOL of AYAs with cancer. Subscales assess physical, emotional, social, and school well-being, plus cancer-related domains such as pain, nausea, procedural anxiety, and perceived physical appearance. In healthy populations, a score of <70 is considered at risk for poor QOL. In patients with cancer, mean scores for the Generic and Cancer PedsQL Modules are reported at 70.9 (SD 17.2) [37] and 65.3 (SD 16.3) [30], respectively.

Hospital Anxiety and Depression Scale (HADS) [39]. The HADS assesses depressive and anxious symptoms in patients with serious illness. It has been validated in AYAs with chronic illness [40] as well as AYA cancer survivors [41]. The mean score for adolescents with cancer is 11 (SD 6.2) [30]. A ‘case’ of anxiety and depression is defined as ≥8 points.

Connor-Davidson Resilience Scale (CD-RISC) [42]. The CD-RISC is a reliable and widely used instrument to measure inherent resiliency. The 10-item instrument has been used in diverse populations including adolescents, parents, and patients with cancer. The mean score among healthy US adults is 31.8 (SD 5.4) [42] and 28 (SD 5.8) in adolescents with cancer [30].

Kessler-6 general psychological distress scale (K6) [43]. This 6-item scale measures level of psychological distress experienced in the past month. The instrument has been extensively cross validated, including among adolescents. The average score for healthy adolescents is 5.8 (SD 4.7) [44], and 7.0 (SD 4.7) [30] for adolescents with cancer. Previous studies have shown that scores > 6 are consistent with high distress and those ≥ 13 meet criteria for serious or debilitating psychological distress [43].

Data analysis

In this secondary analysis, we used baseline ECGs and survey data to examine the relationship between HRV parameters and patient-reported quality of life and resilience, as well as symptoms of anxiety, depression, and distress. We summarized these baseline measures using means/medians, standard deviations, frequencies, and proportions. All variables were reported as continuous, with some variables converted to ordinal or dichotomous when appropriate using clinically relevant cut points (e.g., a HADS depression subscale score of ≥ 8). We conducted linear regression modeling to test associations between HRV and PRO scores. In adjusted models, we controlled for age, gender, cancer type (Leukemia/Lymphoma or CNS/Non-CNS Solid Tumors), and cancer status (relapsed or newly diagnosed). Patient age was transformed into an ordinal variable based on age groups thought to be the most similar developmentally and physiologically. Exploratory stratified analyses were also performed to assess the HRV-PRO relationship among differing age categories, gender, cancer type, and cancer relapse status. Additionally, we compared median HRV values of participants to published normative data for sex-matched healthy adolescents [31], as well as among patients with relapsed versus newly diagnosed cancer in our sample using the Wilcoxon rank sum test. All data analysis was conducted using Stata 14 software (StataCorp, College Station, TX).

Results

There were 76 patients with both ECGs and surveys at baseline. Just under half of participants were female, and the mean age at study entry was 16 years (SD 3 years) (Table 1). The most common cancer diagnosis was leukemia/lymphoma, and most participants identified as white. On average, surveys were collected 7 days after ECGs (rage 85 days before to 64 days after).

Table 1

Patient characteristics.

	Total N = 76
Sex
Female	43%
Age Range	12-25y
Mean (SD)	16y (3y)
12-15y	43%
16-19y	43%
20-25y	13%
Race
White	63%
Black/African American	3%
Asian	7%
Other*	28%
Cancer Diagnosis
Leukemia/Lymphoma	75%
CNS Tumor	3%
Non-CNS Solid Tumor	22%
Psychological Instrument Score	Mean (SD)
HADS Total	11.1 (6.3)
HADS Depression	5.1 (3.4)
Above clinical cutoff on HADS Depression	29%
HADS Anxiety	6.0 (3.6)
Above clinical cutoff on HADS Anxiety	28%
CD-RISC	28.9 (6.0)
Kessler-6	6.9 (4.8)
Above clinical cutoff	49%
Quality of Life (PedsQL)	60.3 (19.1)
Cancer Specific QOL	65.6 (16.8)
Elapsed time between survey and ECG **
Mean (range) in days	7.7 (-85–64)

*Other = mixed race, missing, or other.

**Three outliers were removed with gaps of >120 days between survey and ECG.

Psychological instrument results

At baseline, AYA participants reported a mean HADS score of 11.1 (SD 6.3), with nearly one-third of participants meeting criteria for clinically relevant anxiety or depression. Mean psychological distress score was 6.9 (SD 4.8), with 49% meeting criteria for elevated distress. The mean resilience score was 28.9 (SD 6), and general and cancer specific QOL scores were 60.3 (19.1) and 65.6 (SD 16.8), respectively.

HRV results

The median values (IQR) for SDNN and RMSSD were 30.9 (12.7–50.3) and 31.2 (20.4–38.6), respectively (Table 2). Compared to published values for healthy adolescents, participants had statistically significantly lower median SDNN [Females: 31.9 (12.8–50.7) vs 66.4 (46.0–86.8), p<0.01; Males: 29.9 (11.5–47.9) vs 63.2 (48.4–84.6), p<0.01] and RMSSD [Females: 28.2 (11.1–45.5) vs 69.0 (49.1–99.6), p<0.01; Males: 27.9 (8.6–48.6) vs 58.7 (44.8–88.2), p<0.01]. Newly diagnosed patients tended to have higher median HRV values [SDNN = 33.0 (13.2–50.4), RMSSD = 29.5 (11.6–47.6)] than those patients who enrolled in the trial at the time of relapse [SDNN = 23.6 (10.2–50.2), RMSSD = 20.8 (10.0–43.1)], but this was not statistically significant. Fig 2 illustrates HRV measures stratified by age, gender, cancer type, and cancer relapse status. Additionally, values for both SDNN and RMSSD demonstrated a right skew, with more patients falling on the lower end of the HRV spectrum (Fig 1).

Table 2

Heart rate variability measures.

	SDNN median (IQR)	RMSSD median (IQR)
All patients (n = 76)	30.9 (12.7–50.3)	31.2 (20.4–38.6)
Age
12–15 (n = 33)	25.7 (12.8–52.9)	23.0 (10.6–40.7)
16–19 (n = 33)	39.5 (8.8–49.0)	27.9 (8.1–47.7)
20–25 (n = 10)	34.3(17.4–47.9)	30.9 (15.8–56.9)
Sex
M (n = 43)	29.9 (11.5–47.9), p <0.01*	27.9 (8.6–48.6), p <0.01*
F (n = 33)	31.9 (12.8–50.7), p <0.01*	28.2(11.1–45.5), p <0.01*
Cancer Type
Leukemia/Lymphoma (n = 57)	34.2 (12.8–50.7)	28.2 (12.2–45.5)
Non-CNS Solid Tumor (n = 17)	28.2 (17.4–46.7)	30.8 (8.1–48.6)
CNS Solid Tumor (n = 2)	10.1 (7.0–13.2)	6.7 (5.1–8.1)
Cancer Status
Newly diagnosed (n = 58)	33.0 (13.2–50.4)	29.5 (11.6–47.6)
Relapsed (n = 18)	23.6 (10.2–50.2), p = 0.47**	20.8 (10.0–43.1), p = 0.35**

All results reported in milliseconds.

*Compared to sex-matched published normative values in healthy adolescents.

**Compared to newly diagnosed patients.

Fig 2

Scatterplots of the two HRV measures SDNN and RMSSD plotted against patient reported psychosocial symptoms with simple linear regression fitted lines in red.

Fig 1

Histograms showing the distribution of SDNN and RMSSD for the entire cohort of n = 76 participants.

Vertical red lines represent normal median values for healthy adolescents.

Associations between HRV and PROs. There were no linear associations between either measure of HRV and baseline anxiety, depression, distress, quality of life, or resilience. There was no statistically significant association between PRO measures and SDNN or RMSSD in either an unadjusted or adjusted linear regression model (Table 3, Fig 2). In exploratory stratified analyses of patient-reported anxiety and depression scores there were no statistically significant relationships among other subgroups (S1 Table). However, the association between depression and RMSSD approached statistical significance in females (beta coefficient = 0.54, p = 0.09), as did depression and SDNN in patients aged 20–25 years (beta coefficient = 1.06, p = 0.09).

Table 3

Association of heart rate variability with patient reported outcomes.

	SDNN			RMSSD
Model 1	β coefficient	95% CI	p-value	β coefficient	95% CI	p-value
Anxiety	0.2	[-0.2–0.6]	0.3	0.1	[-0.2–0.5]	0.46
Depression	0.16	[-0.2–0.5]	0.36	0.2	[-0.2–0.6]	0.32
Distress	0.3	[-0.2–0.8]	0.25	0.2	[-0.3–0.7]	0.42
Resilience	-0.05	[-0.7–0.6]	0.89	-0.03	[-0.7–0.6]	0.93
General QOL	0.6	[-1.4–2.6]	0.57	0.8	[-1.3–2.8]	0.45
Cancer-specific QOL	-0.6	[-2.4–1.1]	0.48	-0.02	[-1.8–1.8]	0.99
Model 2
Anxiety	0.17	[-0.2–0.5]	0.35	0.2	[-0.2–0.5]	0.42
Depression	0.1	[-0.2–0.5]	0.44	0.2	[-0.2–0.5]	0.3
Distress	0.2	[-0.2–0.7]	0.3	0.2	[-0.3–0.7]	0.4
Resilience	-0.02	[-0.6–0.5]	0.94	-0.1	[-0.7–0.5]	0.71
General QOL	0.7	[-1.0–2.5]	0.42	0.6	[-1.2–2.4]	0.5
Cancer-specific QOL	-0.4	[-1.8–0.9]	0.51	-0.2	[-1.5–1.2]	0.86

Linear regression models with patient reported outcomes as primary outcome of interest, and SDNN or RMSSD as the predictor of interest. Model 1 = unadjusted model, Model 2 = adjusted for Age, Gender, and Cancer Relapse Status.

Discussion

In this exploratory post-hoc analysis, we did not find a statistically significant association between heart rate variability and psychosocial patient reported outcomes in AYAs with cancer. However, median HRV parameters were substantially lower than in the healthy adolescent population, which may have broader implications for overall well-being. Results of this study provide baseline normative HRV values in the largest published cohort of AYA oncology patients to date. Given the growing interest in using HRV as a measure of ANS function in supportive care research, documenting the normative values and distribution is an important first step in building a larger program of research in biopsychosocial AYA oncology.

Cancer treatment, especially chemotherapy agents like microtubule inhibitors and anthracyclines, can directly damage the nervous and cardiac systems and cause autonomic dysfunction. It is therefore not surprising that reduced HRV has been documented in the leukemia population (where Vincristine and Doxorubicin are often present in treatment regimens) [24] and adult survivors of childhood cancers [25]. Indeed, our data suggest there could be a more pronounced HRV reduction in the relapsed patients, likely representing treatment effect in addition to the compounded psychosocial stress of multiple cancer diagnoses.

However, newly diagnosed patients also had median HRV parameters that appeared lower than healthy adolescents, raising the question of other non-treatment-related sources for autonomic dysfunction that may come with a new cancer diagnosis. These could include disruption of normal neuro-cardiovascular physiology by cancer itself (e.g. a large mass in the mediastinum or catecholamine-secreting tumor) or the intensified psychosocial stressor of the diagnosis in the AYA population (and accompanying social, developmental, financial, and existential threats). As HRV is easily measured through non-invasive methods, it could provide an innovative tool to help risk stratify patients who may require additional forms of supportive care.

Participants in this study were enrolled on a larger randomized trial testing the resilience intervention, PRISM. Primary results from this trial indicated the intervention was associated with improved resilience and cancer-specific quality of life, as well as decreased distress [30]. Because we did not have parallel longitudinal screening ECG recordings in our patient cohort, we were unable to evaluate possible effects of the PRISM intervention on HRV. However, understanding the physiologic correlates of psychosocial intervention is an important area for future investigation, particularly in AYA oncology.

There are several possibilities for why we did not find a relationship between HRV and psychological states when this has been reported in other adolescent populations [15]. Importantly, as a secondary analysis of a larger trial, our study was not powered to detect our associations of interest. Additionally, ECGs were not obtained as part of a strict protocol to control known external influences on basal HRV such as time of day, recent physical activity, and controlled respiration [37, 45]. Nonetheless, the use of baseline ECGs (as opposed to ECGs obtained for clinical concern) in a large trial of AYA oncology patients provides a unique opportunity for data collection to inform subsequent study design.

There were some specific HRV-PRO relationships that approached statistical significance and warrant further examination in larger prospective studies. For example, SDNN and depression in female patients, as well as those patients in the older age group (20-25y). Future studies should focus on collecting extended (24h) HRV recordings in parallel with psychosocial patient reported outcome measures. The optimal study protocol would also avoid data collection during acute medical situations that could confound HRV results in patients with cancer (e.g., serious infection, chemotherapy toxicity, significant cardiac medication adjustments).

This study has additional key limitations. We had a relatively small, demographically homogeneous sample, and thus our results may not be generalizable to other populations. Given the nature of multiple comparisons in our analysis, there was also an increased risk for bias in our results. Additionally, there are other variables known to contribute to baseline HRV, such as physical fitness level, that were not collected as part of this study and thus could not be accounted for in the analysis. However, despite these limitations, this is the first study to report baseline HRV data in the adolescent cancer population. This provides a useful framework to develop larger prospective studies in biobehavioral AYA oncology. Future research should include longer protocolized HRV measurement, as well as collect HRV and psychosocial PRO data in parallel to test its utility as a biomarker of well-being in this population.

Conclusions

In this secondary analysis, we did not find evidence of an association between HRV measures and patient-reported psychosocial outcomes. However, the implications for reduced HRV in AYA oncology patients compared to healthy adolescents warrants further investigation.

Exploring the physiologic underpinnings of patient-centered outcomes could offer novel intervention strategies in patients with serious illness.

Association of heart rate variability with depression and anxiety stratified by sex, cancer type and age.

Association of patient reported anxiety and depression with HRV measures stratified by gender, cancer type, and age using linear regression models. Models adjusted for age and cancer type when not the stratum of interest. Coefficients interpreted as change in PRO score for every 10msec change in HRV. CNS and non-CNS solid tumor categories collapsed for analysis.

(DOCX)

Click here for additional data file.

(XLSX)

Click here for additional data file.

7 Jul 2021

PONE-D-21-16445

Heart Rate Variability and Psychosocial Symptoms in Adolescents and Young Adults with Cancer

PLOS ONE

Dear Dr. Taylor,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Aug 19 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Michael Kaess, M. D.

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Requires revision

This is a very interesting study and certainly worthy of publication. However, the presentation needs improving.

General issues

As indicated above, the presentation needs improving as the tabular presentations do not convey much meaning. For example, the ß coefficients of Table 4 (non-of which are statistically significant) do not give an immediate expression of how important (or not) they are as they refer to the change in PRO for unit change of (say) SDNN. Thus, for Anxiety ß = 0.02 for a 1 ms increase in SDNN whereas a 10ms increase of SDNN would bring a 0.02 � 10 = 0.2 change in PRO. Think of plotting the weights (the y variable) of 10 children whose ages (the x variable) range from 5 to 15 years using days, months or weeks, rather than years. For the first of these analyses the ß coefficient is very small, the second larger, the third larger still but not as large as the fourth. So, looking at Figure 2, a 10ms x-scale might be a better choice for the analysis of this study. This of course would not change the conclusions but ease the interpretation.

More importantly, Table 4 and Figure 2 (interesting though they are) do not give much indication of the relationship between the HRV measure and the PRO. What is needed are some scatter plots of HRV against PRO for all the 76 patients in the study. That is 6 plots for SDNN and 6 for RMSSD. Several statistical packages allow such plots to be collated into two Figures of 6 panels. The fitted simple linear regression model fitted should be superimposed on each panel. Such plots would allow the investigators, and PLoS readers, to get a general impression of what is going on. Are the points close to the fitted lines? Is the relationship linear? Are there any unusual features?

Breaking these down into patient subgroups could then be done using multiple regression techniques with patient characteristics as covariates.

Specific issues

Page Line

2 14-16 These do not give the magnitudes of the differences there are between these patients and their sex-matched population. In any event, I am not sure this analysis needs to be conducted for the purpose of this study.

2 21 As indicated by the authors, the information in this study should form the basis of any proposed ‘Larger prospective study’. However, all statistical tests are reported as non-significant but the lack of significance could be due to (no or a very small effect) or that there are insufficient numbers in this study. So, are any differences observed by the authors in this study suggestive of clinically important effects that could be demonstrated in a larger study. If so, it would be useful to indicate these in their Discussion section.

5 & 6 Table 1 & Line 12 My understanding is that ‘normal or reference ranges’ are usually indicated by mean � 2SD which encompasses most of the data, and not the ICR which only covers the middle 50% of the possible data values.

7 10 Suggest replace ‘7’ by ‘7.0’

Reviewer #2: Title: "Heart Rate Variability and Psychosocial Symptoms in Adolescents and Young Adults with Cancer"

Summary

In their manuscript, the authors describe secondary analyses considering HRV in association with psychosocial outcomes in adolescents and young adults with cancer, as well as a comparison of HRV measures between oncology patients and sex- and age-matched population norms. Analyses were conducted using clinical baseline data from a randomized intervention study and cardiac autonomic measures (calculated in the time-domain) acquired using ultra-short (10 sec) ECG recordings. While the authors reported no evidence for an association between cardiac autonomic measures and any of the psychosocial outcomes, they report significant deviations of SDNN and RMSSD values in newly diagnosed and relapsed cancer patients compared to sex- and age-matched population norms.

This concise article treats an important topic and has the potential to drive further research evaluating autonomic biomarkers of risk stratification in young individuals with cancer.

Yet, I think that currently the article is lacking some in-depth elaboration, while certain figures and tables might be revised / removed.

In the comments below are suggestions to the authors with the aim to help improve the manuscript.

Major comments

1). I would suggest to entirely remove Figure 1 from the manuscript. Instead, the authors should elaborate on the supposed pathways in the according paragraph of the introduction section, while citing relevant literature.

2). I was wondering whether the authors checked for differences in demographics / psychosocial outcomes between patients with and without available ECG recordings?

2). Reporting HRV measurement, the authors may consider following and citing current guidelines: Quintana, D., Alvares, G. & Heathers, J. Guidelines for Reporting Articles on Psychiatry and Heart rate variability (GRAPH): recommendations to advance research communication. Transl Psychiatry 6, e803 (2016). https://doi.org/10.1038/tp.2016.73

3). Table 1 seems redundant, and I would suggest to remove it from the manuscript. In addition, Table 4 could be integrated in the supplementary table. Figure 2 also seems largely redundant, instead, the additional information provided could be integrated in Table 3. Comparisons with population norms might be depicted instead.

4). Can the authors please clarify whether ECG recordings were collected before, at the same time as, or after psychosocial measures, and how much time elapsed between measurements?

6). In their discussion, the authors state "Importantly, as a secondary analysis of a larger trial, our study was not powered to detect our associations of interest". I was wondering whether the authors could explicate how the respective associations of interest should be investigated in future studies?

7). In the discussion the authors also state "However, newly diagnosed patients also had median HRV parameters that appeared lower than healthy adolescents, raising the question of other non-treatment-related sources for autonomic dysregulation." I was wondering whether the authors might enrich their discussion by elaborating on potential "non-treatment-related sources for autonomic dysregulation" in this population, highlighting important research gaps.

Minor comments

8). I was wondering whether the authors checked for differences in demographics / psychosocial outcomes between patients with and without available ECG recordings?

9). The R-package used (RHRV) should be properly cited to acknowledge the authors’ work:

Leandro Rodriguez-Linares, Xose Vila, Maria Jose Lado, Arturo Mendez, Abraham Otero and Constantino Antonio Garcia (2019). RHRV: Heart Rate Variability Analysis of ECG Data. R package version 4.2.5. https://CRAN.R-project.org/package=RHRV

10). For group comparisons, besides p-values the authors may also report according effects sizes.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: David Machin

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

4 Aug 2021

We have provided our de-identified data and uploaded to the supplemental files.

Reviewer 1 Comments:

Comment: As indicated above, the presentation needs improving as the tabular presentations do not convey much meaning. For example, the ß coefficients of Table 4 (non-of which are statistically significant) do not give an immediate expression of how important (or not) they are as they refer to the change in PRO for unit change of (say) SDNN. Thus, for Anxiety ß = 0.02 for a 1 ms increase in SDNN whereas a 10ms increase of SDNN would bring a 0.02 � 10 = 0.2 change in PRO. Think of plotting the weights (the y variable) of 10 children whose ages (the x variable) range from 5 to 15 years using days, months or weeks, rather than years. For the first of these analyses the ß coefficient is very small, the second larger, the third larger still but not as large as the fourth. So, looking at Figure 2, a 10ms x-scale might be a better choice for the analysis of this study. This of course would not change the conclusions but ease the interpretation.

Change: We have adjusted the presentation of the data to make the ß coefficients easier to interpret in Table 3 (formerly Table 4).

Comment: More importantly, Table 4 and Figure 2 (interesting though they are) do not give much indication of the relationship between the HRV measure and the PRO. What is needed are some scatter plots of HRV against PRO for all the 76 patients in the study. That is 6 plots for SDNN and 6 for RMSSD. Several statistical packages allow such plots to be collated into two Figures of 6 panels. The fitted simple linear regression model fitted should be superimposed on each panel. Such plots would allow the investigators, and PLoS readers, to get a general impression of what is going on. Are the points close to the fitted lines? Is the relationship linear? Are there any unusual features?

Change: We have now created the scatterplot panels with the fitted simple linear regression model overlayed to facilitate easier visualization of the data (Figure 2).

Comment: Breaking these down into patient subgroups could then be done using multiple regression techniques with patient characteristics as covariates.

Change: We agree this analysis is informative, and so have included these subgroup multiple regression results in the supplemental information (Supplemental Table 1). Given the relatively small sample size for each patient subgroup, we have limited graphical representations to the full cohort as above.

Comment: Page 2, lines 14-16 These do not give the magnitudes of the differences there are between these patients and their sex-matched population. In any event, I am not sure this analysis needs to be conducted for the purpose of this study.

Change: We have now included the magnitudes of the differences in the abstract (as they are reported in the results), as we feel highlighting the differences in HRV between healthy adolescents and those with cancer is an important adjunct to the other analyses. [Page 2]

Comment: Page 2, Line 21 As indicated by the authors, the information in this study should form the basis of any proposed ‘Larger prospective study’. However, all statistical tests are reported as non-significant but the lack of significance could be due to (no or a very small effect) or that there are insufficient numbers in this study. So, are any differences observed by the authors in this study suggestive of clinically important effects that could be demonstrated in a larger study. If so, it would be useful to indicate these in their Discussion section.

Change: We have included some examples of specific relationships that warrant further study based on results approaching statistical significance: “There were some specific HRV-PRO relationships that approached statistical significance and warrant further examination in larger prospective studies. For example, SDNN and depression in female patients, as well as those patients in the older age group (20-25y).” [Page 13]

Comment: Page 5 & 6 Table 1 & Line 12 My understanding is that ‘normal or reference ranges’ are usually indicated by mean � 2SD which encompasses most of the data, and not the ICR which only covers the middle 50% of the possible data values.

Change: Yes, we agree this is typically true. However, in the cited study (Sharma, et al) HRV for healthy adolescents was not normally distributed, and thus the data was reported as medians (and 25th – 75th percentiles). [Pages 2 and 10]

Comment: Page 7, Line 10 Suggest replace ‘7’ by ‘7.0’

Change: This has been changed.

Reviewer 2 Comments:

Comment: I would suggest to entirely remove Figure 1 from the manuscript. Instead, the authors should elaborate on the supposed pathways in the according paragraph of the introduction section, while citing relevant literature.

Change: We have removed Figure 1 and included additional text and references for relevant mechanistic pathways in biobehavioral oncology: “Psychosocial and environmental stressors can activate a cascade of signaling pathways that have overlapping behavioral and biologic implications. For example, a perceived threat (such as a new cancer diagnosis) triggers the hypothalamic-pituitary-adrenal (HPA) axis to ultimately secrete glucocorticoids into circulation, which have well-documented metabolic and immunomodulatory effects that influence cancer-related outcomes. The autonomic nervous system (ANS) is another key physiologic mediator of the relationship between patient experience and outcomes in biobehavioral oncology. Activation of the parasympathetic and sympathetic branches of the ANS has been associated with symptoms of anxiety, depression, and fatigue in patients with cancer. Conversely, interruption of sympathetic adrenergic signaling using β-blocker medications is associated with improved metastatic and inflammatory biomarkers, as well as lower cancer-specific mortality in patients with breast cancer.” [Page 3]

Comment: I was wondering whether the authors checked for differences in demographics / psychosocial outcomes between patients with and without available ECG recordings?

Change: We did not specifically look at differences by demographics or psychosocial outcomes, as in nearly half of the missing cases, pre-treatment ECGs were not medically necessary. This would leave a very small group (n=9 participants) to analyze, which would not likely yield meaningful results. We have now included documented reasons for missing ECGs: “Of the n=16 participants with missing ECGs, 7 did not require ECGs for their specific treatment plan, and there was no documented reason for omission in 9 cases.” [Page 5]

Comment: Reporting HRV measurement, the authors may consider following and citing current guidelines: Quintana, D., Alvares, G. & Heathers, J. Guidelines for Reporting Articles on Psychiatry and Heart rate variability (GRAPH): recommendations to advance research communication. Transl Psychiatry 6, e803 (2016). https://doi.org/10.1038/tp.2016.73

Change: Thank you for this excellent reference. In addition to including the citation, we have updated our Methods section to more closely align with GRAPH standards: “HRV was derived using 10 second 100Hz ECGs obtained at the time of cancer diagnosis or relapse as part of routine clinical care. These screening ECGs were obtained as part of patients’ medical care and were not part of the clinical trial. Individual paper ECGs were retrospectively extracted from the medical record and scanned into a digital format using a high-resolution scanner. These digital ECG files were then uploaded into an image processing software (WebPlotDigitizer) to extract R-R intervals from the ECG tracing images. R-waves were first identified using the software algorithm and then manually reviewed for accuracy and artifacts. The discrete R-R intervals (in milliseconds) could then be identified from the image and converted to their numerical form. Using the open-source R software, RHRV, we then derived the two most widely used time domain parameters: standard deviation of normal to normal beats (SDNN) and root mean square of successive differences (RMSSD) per published guidelines.” [Pages 5-6]

Comment: Table 1 seems redundant, and I would suggest to remove it from the manuscript. In addition, Table 4 could be integrated in the supplementary table. Figure 2 also seems largely redundant, instead, the additional information provided could be integrated in Table 3. Comparisons with population norms might be depicted instead.

Change: We have now removed Table 1 and integrated the additional information from Figure 2 into Table 3 (now Table 2). We have also moved the histograms depicting HRV values compared to population norms from the Supplemental data to the main manuscript (now labeled Figure 1). As Table 4 (now Table 3) presents the primary data analysis for this study, we feel it is important to include this data in the main manuscript but have adjusted the reporting of the data to make the interpretation of ß coefficients more meaningful. To further ease reader interpretation, we have created scatter plots instead for easier visualization (now Figure 2).

Comment: Can the authors please clarify whether ECG recordings were collected before, at the same time as, or after psychosocial measures, and how much time elapsed between measurements?

Change: As this study was a secondary analysis, we did not have control over timing of ECG collection, and this varied in relation to survey administration. We have now incorporated this data into Table 1 and included the following statement in the Results section: “On average, surveys were collected 7 days after ECGs (rage 85 days before to 64 days after).” [Page 8]

Comment: In their discussion, the authors state "Importantly, as a secondary analysis of a larger trial, our study was not powered to detect our associations of interest". I was wondering whether the authors could explicate how the respective associations of interest should be investigated in future studies?

Change: We have now given more explicit suggestions for future study design: “Future investigations should prospectively collect extended (24h) HRV recordings in parallel with psychosocial patient reported outcome measures. The optimal study protocol would also avoid data collection during acute medical situations that could confound HRV results in patients with cancer (e.g. serious infection, chemotherapy toxicity, significant cardiac medication adjustments).” [Page 13]

Comment: In the discussion the authors also state "However, newly diagnosed patients also had median HRV parameters that appeared lower than healthy adolescents, raising the question of other non-treatment-related sources for autonomic dysregulation." I was wondering whether the authors might enrich their discussion by elaborating on potential "non-treatment-related sources for autonomic dysregulation" in this population, highlighting important research gaps.

Change: We have included specific examples of theorized non-treatment-related etiologies for autonomic dysregulation in AYAs with cancer: “However, newly diagnosed patients also had median HRV parameters that appeared lower than healthy adolescents, raising the question of other non-treatment-related sources for autonomic dysfunction that may come with a new cancer diagnosis. These could include disruption of normal neuro-cardiovascular physiology by the cancer itself (e.g. a large mass in the mediastinum or catecholamine-secreting tumor) or the intensified psychosocial stressor of the diagnosis in the AYA population (and accompanying social, developmental, financial, and existential threats). As HRV is easily measured through non-invasive methods, it could provide an innovative tool to help risk stratify patients who may require additional forms of supportive care.” [Page 12-13]

Comment: The R-package used (RHRV) should be properly cited to acknowledge the authors’ work:

Leandro Rodriguez-Linares, Xose Vila, Maria Jose Lado, Arturo Mendez, Abraham Otero and Constantino Antonio Garcia (2019). RHRV: Heart Rate Variability Analysis of ECG Data. R package version 4.2.5. https://CRAN.R-project.org/package=RHRV

Change: We have now included this citation. [Page 6]

Comment: For group comparisons, besides p-values the authors may also report according effects sizes.

Change: We do agree that effect sizes generally assist with interpretability; however, in the case of the nonparametric testing that is required of comparing HRV population values to our study cohort, the effect sizes do not necessarily provide meaningful information. And so although we can provide numerical effect sizes from the modified ranksum testing performed, we feel that in this case comparing raw differences in medians (and IQRs) gives readers a clearer sense of how the two groups might be different.

Submitted filename: Response to Reviewers.doc

Click here for additional data file.

19 Oct 2021

Heart Rate Variability and Psychosocial Symptoms in Adolescents and Young Adults with Cancer

PONE-D-21-16445R1

Dear Dr. Taylor,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Michael Kaess, M. D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The addition of Figure 2 adds substantially to the previous version.

All my comments can be released to the authors

Reviewer #2: I would like to thank the authors for their careful revision and for addressing all my comments. Elaborations on why certain suggestions were not implemented seem plausible to me, and I think the manuscript has improved much during the revision process. In addition, it is great that the underlying dataset has now been shared by the authors! All in all, I think this study is now suitable for publication.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

25 Oct 2021

PONE-D-21-16445R1

Heart rate variability and psychosocial symptoms in adolescents and young adults with cancer

Dear Dr. Taylor:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Prof. Dr. Michael Kaess

Academic Editor

PLOS ONE