A Comparative Study of Sleep Parameters in Adult Survivors of Childhood Acute Lymphoblastic Leukemia and Healthy Peers: Insights from Accelerometer Data

,


Introduction
Acute lymphoblastic leukemia (ALL) is the most common type of malignant tumor in children [1,2].In developed countries, the fve-year survival rate now exceeds 90% [3].However, in this specifc population, survival may be complicated by impaired quality of life [4][5][6] and by the large number of late treatment-related adverse sequelae [7,8].Very often, sleep problems are among them in survivors of childhood cancer [9][10][11].
Sleep and sleep problems in long-term survivors of childhood ALL may be key factors that can negatively afect mental health [9,12], lead to problems with health behaviors [13], and may also be signifcantly related to fatigue [14,15].
Tese sleep problems may occur months to years after the treatment has been completed [16].Poor sleep quality has been confrmed not only in children [17] but also in adolescent survivors of childhood ALL [18].Survivors of childhood ALL may also very often fail to meet the sleep recommendations [19].
Sleep problems have been addressed in research studies both during treatment (in children) and after the end of treatment, e.g., in the frst period after diagnosis [20], during the frst year of treatment [21], three years after the end of treatment [17], after two to seven years [19], and with an average number of years after diagnosis of 7.4 years [18].Measuring selected sleep parameters using actigraphy [22] is coming to the forefront in population epidemiological studies compared with self-reported methods.Actigraphy is a feasible alternative to polysomnography, and its huge advantage is its application to a natural environment [23].Instrumentally measured sleep parameters should play a major role in future studies [24], and sleep assessment should become a part of annual patient care [25].
In our study, we are interested in whether sleep problems can persist in other stages of life and what sleep parameters in the areas of time in bed (TIB), sleep (SL), and sleep efciency (SE), which are one of the important indicators of sleep quality [26], are achieved by adults who became ill in childhood.We have not been able to fnd studies that characterize (using the method of actigraphy) sleep with such a long-term (>ten years) distance from the completion of treatment or from diagnosis in this target group.Given that research studies in the healthy population also point to sex-related diferences in sleep parameters-women achieve longer sleep times and higher SE compared to men [27,28]-we are interested in whether these diferences also occur in adult survivors of childhood acute lymphoblastic leukemia (ALS).
Te main objective of the study is to characterize the selected sleep parameters (TIB, SL, and SE) within a normal week in ALS with regard to gender and to compare these characteristics with the characteristics of their healthy peers.Another partial aim of the study is to determine the fulfllment of recommendations in the areas of SL and SE.

Methods
Tis cross-sectional study involved a total of 40 participants.20 participants formed the group ALS, where the criterion for inclusion in the research was an age of 18-30 years, having undergone active treatment for ALL in childhood at departments of pediatric hematology and hemato-oncology in the Czech Republic, and with the minimum time since their diagnosis being more than ten years.Te exclusion criterion was disability or medical condition not related to the treatment.Te control group (CG) consisted of 20 healthy participants who had not experienced any cancer and had no disability or health impact.All the members of both groups (ALS and CG) had their permanent residence in the Czech Republic.

Recruitment of Participants.
At present, addressing potential research participants with such a long time since the completion of treatment is very demanding organizationally.One of the main ways to reach this target group is to reach out through patient organizations.However, these organizations only register patients who are interested.For the purposes of our research, this target group was addressed through the patient organizations-Šance Association and Společně s úsměvem (together with smile).Out of a total of 31 ALS addressed who met the criteria for inclusion in the research, 21 ALS showed interest in participating.As part of the research survey, each ALS had the opportunity to involve a CG sampled from a healthy population in the research.In this way, 21 participants forming the CG were recruited.Because of the failure to meet the minimum wear time of the sleep monitoring device, the data of one ALS and one CG were excluded from the study.Te total number of participants meeting the research criteria is 20 ALS and 20 CG.Te recruitment of the participants is shown in Figure 1.

Data Collection.
Tose interested in participating in the research survey contacted the principal investigator of the research, who provided them with all the necessary information.Teir informed consent to the research was required for them to participate in the research survey.Te study was approved by the Ethics Committee of the Faculty of Physical Culture of Palacký University in Olomouc under no.20/2021.Te monitoring equipment was handed over to both research groups in person one or two days before the scheduled start of the research.Tis period was chosen to reduce the risk of battery discharge and thus not meeting the specifed number of days of measurement.Te actual data collection took place for seven days and was carried out from June 14 to 20, 2021.After the end of the data collection, the device was sent back by post at the expense of the project investigator for evaluation.Information regarding treatment (in the ALS) and current characteristics such as age, height, and weight (in both the ALS and CG) were obtained by a questionnaire survey.

Sleep Monitoring.
Te Axivity AX3 accelerometer (Axivity Ltd., Newcastle upon Tyne, UK), a small waterproof device (23 × 32.5 × 7.6 mm; 11 g), was used to measure sleep quantity and quality.Te Axivity AX3 was used in the UK Biobank study [29] and has been previously used in cancer survivors [30][31][32].Te participants were instructed to wear the device on their nondominant wrist for seven consecutive days, except while swimming.Te device was initialized to measure data at 100 Hz with an 8G bandwidth using the OmGui software (version 1.0.0.43,Newcastle University, UK).Te raw accelerometer data were processed using the R package GGIR (v2.1-0, https://cran.r-project.org/web/packages/GGIR/)[33].Te default sleep algorithm [34] was used to identify periods of sleep (sustained periods of inactivity with no change greater than 5 °over at least fve min), the onset of sleep, and waking time.

Sleep Parameters.
Te indicators for assessing sleep quantity and quality were determined as time in bed (TIB), sleep (SL), and sleep efciency (SE).TIB was defned as the diference between the onset of sleep and the time of waking, SL was defned as the sum of all the recorded periods of sleep during the night, and SE was defned as the percentage of time spent sleeping within the TIB [34].Te optimal SL for this study was determined to be 420 min/day [35].SE can 2 European Journal of Cancer Care provide essential information about the health status of an individual [36]; the threshold of ≥85% was used as an indicator of good sleep quality [37].A chi-squared test as recommended by Campbell [38] was used for the comparison of proportions.

Results
In terms of descriptive characteristics of the study participants ( In the ALS group, the median age at the time of the disease was 7.5 (IQR: 8.0), the median duration of treatment was 18.0 (IQR: 13.0), and the median time from diagnosis was 15.5 years (IQR: 8.9).
Te overall results (Table 2) showed no signifcant differences between the ALS and CG groups.For the TIB parameter, the instrumentally measured resulting values were 405.5 min/day for the ALS (CG: 428.2 min/day), SL 372.7 min/day for the ALS (CG: 382.9 min/day), and SE 88.0% for the ALS (CG: 88.5%).
An overview of the selected sleep parameters evaluated on the basis of gender is shown in Table 3. Again, there were no signifcant diferences between the ALS and CG.No significant diferences were found even when the gender of the participants was taken into account (TIB ALS male vs. female, p � 0.71; TIB CG male vs. female, p � 0.70; SL ALS male vs. female, p � 0.88; SL CG male vs. female, p � 0.94; SE ALS male vs. female, p � 0.37; SE CG male vs. female, p � 0.06).
In terms of established sleep recommendations (Table 4), >420 min/day was achieved by 15% of the ALS and 19% of the CG.SE > 85% was achieved by 80% of the ALS and 80% of the CG.
No gender diferences were found in terms of meeting the sleep recommendations (SL ALS male × female, p � 0.42; SL CG male × female, p � 0.60; SE ALS male × female, p � 0.83; SE CG male × female, p � 0.12).

Discussion
Te main fnding of this study was that the results for the ALS and CG did not difer for the sleep parameters that were selected.It can be assumed that ALS may achieve the same values in individual sleep parameters as the healthy population.
When comparing the selected sleep parameters measured using an actigraph in the ALS, we could not fnd a study where it would be possible to compare these parameters objectively.A study of the pediatric population with acute lymphoblastic leukemia [20] achieved signifcantly higher values for TIB  [19].With regard to our study, it seems that with the passage of treatment, SE may increase.Since this is one of the important indicators of sleep quality, it would be appropriate to carry out more instrumentally measured studies on this topic.In terms of gender, the sleep parameters did not difer between the ALS and CG.Tese results do not correspond with the fndings of other studies dealing with healthy populations [28,39], which suggest that men sleep less on average than women across age categories.
Te recommended sleep duration >7 h/day (420 min/ day) [35] showed almost no diference between the groups (ALS and CG).However, only a small number of probands met this sleep recommendation in both groups.Since sleep lasting between seven and eight hours per day is associated with health benefts [35], it is necessary to place much greater emphasis on this recommendation.
One of the limitations of our study is the sample size, which may not be large enough to conduct robust associative analyses between sleep quality and various demographic, social, and psychological parameters.Exploring these relationships could provide a more comprehensive understanding of sleep functioning in childhood cancer survivors, and we acknowledge that this is an important area for future research.
Other limitation of our study stems from the inability to collect data on the reasons why some childhood cancer survivors chose not to participate.While we were able to engage with those who expressed interest in the study, we did not have a mechanism in place to gather information from those who did not respond to our invitation.Tis is a signifcant limitation as it leaves unanswered questions about whether the nonparticipants experienced sleep problems or other mental health difculties, which could have potentially biased our fndings.
In addition, we did not collect data on the use of medications afecting sleep and anxiety levels by the participants.Tese substances could potentially infuence sleep duration and quality, and having information on their use could have added valuable insights to our fndings.We can speculate that the use of these substances might be present in both our study and control groups, and thus their impact on our fndings might be mitigated to some extent.We recognize that this is a speculative assumption and that concrete data would have strengthened our study.
Nevertheless, we believe that the ALS target group with a long time lag (in our study, an average of >15 years) since the established diagnosis is a strength of the study since this target group is not the subject of current research studies.Also, instrumental measurement in the natural environment provides a greater advantage over polysomnography.
Adopting a longitudinal approach to assess sleep problems in patients undergoing therapy and patients after treatment is crucial, with objective, regular sleep assessments in natural settings being paramount.Tese assessments may not only facilitate a comprehensive understanding of sleep patterns and potential disruptions but also enable the implementation of preventive measures aimed at mitigating or resolving sleeprelated issues.Te integration of instrumentally measured sleep parameters should be considered an essential component of long-term patient follow-up post-treatment.Already, a minimal 2 week intervention using actigraphy-based sleep feedback can yield positive efects on sleep quality and overall well-being [40].Tis highlights the potential of regular sleep assessments and feedback in positively infuencing patients' outcomes.In addition, combining objective sleep monitoring with ecological momentary assessment [41] may ofer a promising approach.Incorporating these methodologies into future research could signifcantly enhance our understanding of sleep problems in patients undergoing therapy and after treatment and contribute to the development of tailored interventions to improve sleep quality and overall well-being.

Conclusion
Te evidence of our study suggests that adult survivors of childhood acute lymphoblastic leukemia may achieve the same values as the healthy population for the selected sleep parameters.No signifcant gender diferences were found.Only a small number of the ALS and CG probands met the recommended total length of sleep time.For the SE parameter, the recommendation was met to a much greater extent in both groups.

2. 5 .
Statistical Data Processing.Descriptive statistics (median and interquartile range) were used to provide the basic characteristics of the study sample.To compare the groups (gender and treatment), the nonparametric Mann-Whitney U test was used.Te signifcance level was set to α � 0.05.Te coefcients of the efect size were evaluated as follows: 0.2 ≤ d < 0.5: small efect size, 0.5 ≤ d < 0.8: medium efect size, and d ≥ 0.8: large efect size.Te IBM SPSS Statistics 25 statistical software (IBM SPSS, Inc. Chicago, IL, USA) was used.

Table 1 :
Descriptive characteristics of the study sample (n � 40).

Table 2 :
Overall overview of the selected sleep parameters.

Table 3 :
Overview of the selected sleep parameters by gender.

Table 4 :
Overview of meeting recommendations for sleep and sleep efciency.