A Prospective Study of Anxiety, Depression, and BehavioralChanges in the First Year After a Diagnosis of Childhood

Acute Lymphoblastic Leukemia

A Report From the Children’s Oncology Group

Regina M. Myers, MD1; Lyn Balsamo, PhD1; Xiaomin Lu, PhD2; Meenakshi Devidas, PhD2; Stephen P. Hunger, MD3;

William L. Carroll, MD4; Naomi J. Winick, MD5; Kelly W. Maloney, MD3; and Nina S. Kadan-Lottick, MD, MSPH1

BACKGROUND: The authors prospectively assessed anxiety, depression, and behavior in children with standard-risk acute lympho-

blastic leukemia (SR-ALL) during the first year of therapy and identified associated risk factors. METHODS: A cohort study was per-

formed of 159 children (aged 2 years-9.99 years) with SR-ALL who were enrolled on Children’s Oncology Group protocol AALL0331

at 31 sites. Parents completed the Behavior Assessment System for Children, the General Functioning Scale of the Family Assessment

Device, and the Coping Health Inventory for Parents at approximately 1, 6, and 12 months after diagnosis. RESULTS: Overall, mean

scores for anxiety, depression, aggression, and hyperactivity were similar to population norms. However, more children scored in the

at-risk/clinical range for depression than the expected 15% at 1 month (21.7%; P 5.022), 6 months (28.6%; P<.001), and 12 months

(21.1%; P 5.032). For anxiety, more children scored in the at-risk/clinical range at 1 month (25.2% vs 15%; P 5.001), but then reverted

to expected levels. On adjusted analysis, unhealthy family functioning was found to be predictive of anxiety (odds ratio [OR], 2.24;

P 5.033) and depression (OR, 2.40; P 5.008). Hispanic ethnicity was associated with anxiety (OR, 3.35; P 5.009). Worse physical

functioning (P 5.049), unmarried parents (P 5.017), and less reliance on social support (P 5.004) were found to be associated with

depression. Emotional distress at 1 month predicted anxiety (OR, 7.11; P 5 .002) and depression (OR, 3.31; P 5.023) at 12 months.

CONCLUSIONS: Anxiety is a significant problem in a subpopulation of patients with SR-ALL immediately after diagnosis, whereas

depression remains a significant problem for at least 1 year. Children of Hispanic ethnicity or those with unhealthy family functioning

may be particularly vulnerable. These data suggest that clinicians should screen for anxiety and depression throughout the first year

of therapy. Cancer 2014;120:1417–25. VC 2014 American Cancer Society.

KEYWORDS: childhood acute lymphoblastic leukemia, anxiety, depression, family functioning.

INTRODUCTIONGreater than 90% of children with acute lymphoblastic leukemia (ALL) with standard risk (SR) features will be cured.1

The improved survival rate is, in part, due to more intensive therapies with durations of 2.5 to 3.5 years.2 These therapiesinvolve multiple different chemotherapeutic agents that can affect the child’s emotional functioning. In particular, corti-costeroids, a key component of therapy for ALL, affect mood, behavior, and cognition.3

To our knowledge, the current study is the first prospective, longitudinal study of emotional and behavioral func-tioning in a large sample of children receiving active treatment for SR-ALL who did not receive cranial radiation.4 Infor-mation regarding the behavioral and emotional health of children with cancer is largely based on studies of children aftertherapy, rather than those receiving active treatment.5 Those studies have used cross-sectional designs and yielded mixedresults regarding emotional distress in patients with ALL.4,6,7 The few longitudinal studies that are available were small,

Corresponding author: Nina S. Kadan-Lottick, MD, MSPH, Section of Pediatric Hematology/Oncology, Yale University School of Medicine, PO Box 208064, 333

Cedar St, LMP-2073, New Haven, CT 06520-8064; Fax: (203) 737-2228; Nina.Kadan-Lottick@Yale.Edu

1Section of Pediatric Hematology/Oncology, Yale University School of Medicine and Yale Comprehensive Cancer Center, New Haven, Connecticut; 2Department

of Biostatistics, College of Medicine, College of Public Health and Health Professions, University of Florida, Gainesville, Florida; 3Department of Pediatrics, Univer-

sity of Colorado School of Medicine and Children’s Hospital Colorado, Aurora, Colorado; 4New York University Cancer Institute, New York, New York; 5Division of

Pediatric Hematology/Oncology, University of Texas Southwestern School of Medicine, Dallas, Texas.

In this prospective study of the psychological adjustment of children receiving active therapy for standard-risk acute lymphoblastic leukemia, the authors found

that a significant subpopulation of children displayed symptoms of anxiety and depression that merit intervention. Particularly vulnerable were children of His-

panic descent or those with unhealthy family functioning.

DOI: 10.1002/cncr.28578, Received: August 20, 2013; Revised: October 16, 2013; Accepted: November 18, 2013, Published online January 28, 2014 in Wiley

Online Library (wileyonlinelibrary.com)

Cancer May 1, 2014 1417

Original Article

included diverse cancer populations, and/or excludedpatients with ALL who were aged< 5 years.8-10 Variablesassociated with children’s psychosocial adjustmentinclude older age,11,12 female sex,13,14 lower householdincome,14 parental distress,14 and more intensivetreatment.12

The contribution of family functioning and copingbehaviors to children’s emotional functioning is of partic-ular interest because there are emerging data that suggestthese factors may be modifiable. For example, a random-ized clinical trial among adolescent survivors of childhoodcancer and their families found that a 1-day family-basedintervention improved symptoms of posttraumaticstress.15 In addition, there is compelling evidence of theefficacy of family interventions in other childhood illnesspopulations such as in type 1 diabetes.16

We prospectively evaluated the emotional and be-havioral functioning of a large representative sample ofchildren with ALL who were enrolled on a frontlineChildren’s Oncology Group (COG) therapeutic studyduring the first year of therapy. We sought to: 1) describethe longitudinal trajectory of the psychological adjust-ment of children during their first year of therapy bymeasuring symptoms of anxiety, depression, and behav-ioral disturbances; and 2) identify factors associated withworse psychological functioning in children, includingpotentially modifiable variables related to family func-tioning and coping that could be targeted in futureinterventions.

MATERIALS AND METHODS

Study Population

We conducted a prospective, longitudinal study of emo-tional and behavioral outcomes in children with SR-ALLwho were enrolled on the COG AALL0331 protocolbetween 2005 and 2009 at 31 sites. We focused on a sub-set of children who were classified as having average-riskALL, defined as SR by National Cancer Institute criteria(initial white blood cell count< 50,000/lL and aged 1.0-9.99 years) without central nervous system or testicularleukemia who had a good early response to therapy basedon bone marrow morphology and minimal residual dis-ease burden at end induction, and other variables.17,18

Additional eligibility requirements for this study of emo-tional and behavioral outcomes included age� 2 yearsand at least 1 parent with reading comprehension of Eng-lish or Spanish, the languages for which validated surveysexist. The participating sites were chosen from all COGsites to include a combination of community-based and

tertiary care centers with available staffing for this ancil-lary study.

Patients received a 3-drug, 4-week induction withvincristine, pegaspargase, and dexamethasone (at a dose of6 mg/m2/day 3 28 days), as well as intrathecal chemo-therapy. No patient received cranial radiation. There were2 therapeutic randomizations: 1) standard consolidationversus intensified consolidation, which added 2 doses ofcyclophosphamide and pegaspargase; and 2) standard in-terim maintenance with oral methotrexate versus aug-mented interim maintenance with escalating intravenousmethotrexate as postconsolidation therapy. In 2008, thesecond randomization was halted based on the results ofthe Children’s Cancer Group-1991 SR-ALL trial.2

A total of 194 patients who were enrolled onAALL0331 at the participating sites met the eligibility cri-teria for this ancillary study. Of these, 24 patients declinedand 170 consented to participate. Of those who con-sented, 4 patients withdrew from the AALL0331 protocolbefore the first required survey evaluations and 7 were notgiven the evaluations because of errors at the study sites.The 159 participants (82% of eligible participants) weresimilar to the 35 eligible nonparticipants in terms of age atdiagnosis and sex, with some differences in ethnicitynoted (Table 1).

Procedures

The Institutional Review Board of each participating cen-ter as well as the Yale University Human InvestigationCommittee approved the current study. Informed con-sent, and assent when indicated, was obtained for all par-ticipants. The identified primary caregiver (the child’smother in 84% of instances) completed surveys at 3selected timepoints during the first year of therapy: day 1of consolidation (approximately 1 month after diagnosis),the end of the delayed intensification (approximately 6months after diagnosis), and 6 months after the initiationof maintenance (approximately 12 months after diagno-sis). Of the 159 participants, 145 individuals, 131 indi-viduals, and 136 individuals completed the evaluations atthe first, second, and third timepoints, respectively.

Measures

Emotional and behavioral functioning was assessed by theBehavioral Assessment System for Children, SecondEdition: Parent Report Scale (BASC-2 PRS), a valid andreliable instrument that has been used successfully in pedi-atric oncology populations.19,20 Children aged� 8 yearsalso completed the BASC-2 Self-Report of Personality.However, only 17 of the children in the current study

Original Article

1418 Cancer May 1, 2014

(10.7%) were aged at least 8 years at the time of diagnosis,and therefore there were inadequate data with which tocompare their self-report outcomes. The BASC-2 PRSyields standardized T-scores on a variety of clinical scales.Scores of 60 to 69 represent the at-risk range andscores� 70 represent the clinically significant range. TheBASC-2 PRS has been standardized on normative dataobtained from a random sample of 12,350 children whoare representative of the US population based on sex, eth-nicity, socioeconomic status, geographic region, and cul-ture.21 Expected frequencies of elevated scores in thenormative (ie, healthy comparison) population of chil-dren are available in the BASC manual.19 For the currentstudy, the hyperactivity, aggression, anxiety, and depres-sion scales were analyzed. Anxiety and depression are of-ten comorbid conditions, but in this instrument there arediscrete scales for each.

Family functioning was evaluated using the GeneralFunctioning Scale of the Family Assessment Device (FAD-GF).22 In this 12-item scale, parents indicate the degree to

which they believe each statement describes their family(eg, “We are able to make decisions about how to solveproblems”). Possible scores range from 1 to 4. Scores� 2reflect unhealthy family functioning.22

Family coping was assessed using the Coping HealthInventory for Parents (CHIP), which has been validatedfor children with a variety of chronic illnesses.23 In this45-item checklist, parents rate how helpful a specific cop-ing behavior is on a 4-point scale ranging from “not help-ful” to “extremely helpful.” A higher score on each of the3 subscales (ie, 1) maintaining family integration and op-timism; 2) maintaining social support and self-esteem;and 3) understanding the medical situation) indicates agreater reliance on that particular coping pattern, butthere are no normative scores.

Physical functioning was measured by the “pain andhurt” and “nausea” subscales of the Pediatric Quality ofLife Inventory (PedsQL) 3.0 Cancer Module.24 In thisquestionnaire, parents rate how much of a problem eachsymptom has been within the past month. Scores aretransformed on a scale from 0 (worst health) to 100 (besthealth).

A parent demographic survey included questionsregarding ethnicity, household income, marital status,maternal education, and family size.

Statistical Analysis

Patient characteristics, including age at diagnosis, sex, race,and ethnicity, were summarized and compared betweenparticipants and eligible nonparticipants using an exact chi-square test to evaluate the potential for response bias.

The primary outcomes of interest were the BASC-2PRS subscales for anxiety and depression. The percentagesof patients in the “at least at-risk” and clinical ranges werecompared with the corresponding percentages in the nor-mative population, a comparison group of healthy chil-dren, using a 1-sided binomial exact test.

Both univariate and multivariate longitudinal analy-ses were conducted. For univariate analysis, a logisticregression model was tested with the dichotomizedBASC-2 PRS scores (ie, elevated vs not elevated scores)for anxiety and depression as dependent variables, takinginto consideration the dependence of repeated measure-ments at 3 timepoints for each subject. The following in-dependent variables were analyzed: age at diagnosis, sex,race and ethnicity, household income, maternal educa-tion, marital status, family size, pain and hurt subscaleand nausea subscale, as well as repeated measures of gen-eral family functioning and coping behaviors. The multi-variate regression modeling included the patient and

TABLE 1. Comparison of Participants With EligibleNonparticipants

CharacteristicParticipants

(n 5 159)

EligibleNonparticipants

(n 5 35) P

Age group at diagnosis, no. .134

Preschool (ages 2-4 y) 86 (54.1%) 24 (68.6%)

School age (ages 5-9 y) 73 (45.9%) 11 (31.4%)

Sex, no. .576

Female 76 (47.8%) 19 (54.3%)

Male 83 (52.2%) 16 (45.7%)

Child ethnicity, no. .011

White, non-Hispanic 108 (67.9%) 16 (45.7%)

Black, non-Hispanic 11 (6.9%) 1 (2.9%)

Hispanic 26 (16.4%) 9 (25.7%)

Other 14 (8.8%) 9 (25.7%)

Marital status of parents, no.

Married 105 (66.0%)

Not married 45 (28.3%)

Missing data 9 (5.7%)

Maternal highest level of education, no.

Less than college 92 (57.9%)

At least some college 55 (34.6%)

Missing data 12 (7.5%)

Family income, no.

<$50,000 72 (45.3%)

$50,000-$79,999 25 (15.7%)

�$80,000 30 (18.9%)

Missing data 32 (20.1%)

Therapeutic randomization, no.

Standard CS/standard IM-DI 42 (26.4%)

Intensified CS/standard IM-DI 51 (32.1%)

Standard CS/augmented

IM-DI

37 (23.3%)

Intensified CS/augmented

IM-DI

29 (18.2%)

Abbreviation: CS, consolidation; DI, delayed intensification; IM, interim

maintenance.

Psychological Adjustment of Children With ALL/Myers et al

Cancer May 1, 2014 1419

family factors that were found to be associated with ele-vated anxiety and depression scores by univariate analysisat P< .1. All analyses were performed using SAS statisticalsoftware (version 9.2; SAS Institute Inc, Cary, NC).

RESULTS

Participants

The participants had a mean age of 4.9 6 2.2 years at thetime of diagnosis; 16.4% were of Hispanic ethnicity and66.0% had married parents (Table 1).

Frequency of Emotional and BehavioralProblems

Mean scores for parental report of anxiety, depression,aggression, and hyperactivity symptoms were stable andin the average range at all 3 timepoints. However, the fre-quency of anxiety and depression scores in the at-risk orclinically significant range was greater than expected com-pared with the normative population of children (Fig. 1).At 1 month after diagnosis, a greater percentage of chil-dren scored in the at-risk or clinically significant range foranxiety (25.2% vs 15%; P 5 .001) compared with thenormative population, but then reverted to expected levelsat 6 months (17.5% vs 15%; P 5 .253) and 12 months(14.2% vs 15%; P 5 .542) after diagnosis. The frequencyof anxiety scores that were elevated to the clinically signifi-cant range was greater than expected at 1 month (10.4%vs 4%; P 5 .001) and 6 months (8.7% vs 4%; P 5 .013)after diagnosis, but then declined to expected levels by 12months after diagnosis (4.5% vs 4%; P 5 .448).

For depression, a higher percentage of children hadscores in the at-risk or clinically significant range thanexpected throughout the first year of therapy (1 month:21.7% vs 15% [P 5 .022]; 6 months: 28.6% vs 15%[P< .001]; and 12 months: 21.1% vs 15% [P 5 .038]).However, the frequency of depression scores in the clini-cally significant range was not significantly different fromexpected levels at any timepoint (1 month: 5% vs 4%[P 5 .316]; 6 months: 6.4% vs 4% [P 5 .133]; and 12months: 6.8% vs 4% [P 5 .087]).

At the 3 timepoints, the percentage of children withboth anxiety and depression scores in the at-risk/clinicallysignificant range was 12.6%, 15.1%, and 8.3%, respec-tively. The frequency of elevated hyperactivity and aggres-sion scores was similar to that expected in a normativepopulation (data not shown).

Longitudinal Analysis

Compared with children with anxiety scores in the averagerange, those with anxiety scores in the at-risk/clinically

significant range 1 month after diagnosis were 7.70 timesas likely to have elevated scores 6 months after diagnosis(95% confidence interval [95% CI], 2.39-24.85;P< .001) and 7.11 times as likely to have elevated scores12 months after diagnosis (95% CI, 2.08-24.30;P 5 .002). Children with scores in the at-risk/clinicallysignificant range 6 months after diagnosis were 20.64times as likely to have scores in the at-risk/clinically signif-icant range 12 months after diagnosis (95% CI, 6.02-70.74; P< .001).

Compared with children with depression scores inthe average range, those with depression scores in the at-risk/clinically significant range 1 month after diagnosiswere 3.51 times as likely to have scores in the at-risk/clini-cally significant range 6 months after diagnosis (95% CI,1.33-9.26; P 5 .015) and 3.31 times as likely to havescores in the at-risk/clinically significant range 12 monthsafter diagnosis (95% CI, 1.20-9.10; P 5 .023). Childrenwith scores in the at-risk/clinically significant range 6months after diagnosis were 5.11 times as likely to havescores in the at-risk/clinically significant range 12 monthsafter diagnosis (95% CI, 1.94-13.48; P< .001) (data notshown).

Predictors of Anxiety and Depression

Table 2 displays the results of the univariate analysisadjusted for time elapsed since diagnosis. Significant pre-dictors of anxiety and depressive symptoms by parentalreport included unhealthy family functioning and lessreliance of each of the 3 coping patterns measured by theCHIP. Hispanic ethnicity was associated with worse anxi-ety symptoms, but not depressive symptoms. Conversely,worse physical functioning as measured by the pain andhurt subscale of the PedsQL was associated with depres-sion, but not anxiety. There were no differences detectedamong the 4 treatment groups.

Table 3 displays the results of a multivariate model,which included the patient and family factors that werefound to be at least marginally significant (P� .1) on uni-variate analysis. In this adjusted analysis, Hispanic ethnic-ity (odds ratio [OR], 3.35; 95% CI, 1.36-8.24) andunhealthy family functioning (OR, 2.24; 95% CI, 1.07-4.70) remained significant predictors of worse anxietysymptoms.

The significant predictors of worse depressive symp-toms by adjusted analysis were unhealthy family function-ing (OR, 2.40; 95% CI, 1.26-4.56), unmarried parents(OR, 2.36; 95% CI, 1.17-4.75), worse physical function-ing (P 5 .049), and less reliance on maintaining socialsupport coping behaviors (P 5 .004).

Original Article

1420 Cancer May 1, 2014

DISCUSSIONTo the best of our knowledge, the current study is the firstprospective, longitudinal study of emotional and behav-ioral functioning in a large sample of children receivingactive treatment for SR-ALL who did not receive cranialradiation. Although the majority of children had anxietyand depression scores similar to those of controls, a signif-icant subpopulation of children displayed symptoms thatmerit intervention. We found that depressive symptomswere a significant problem from the end of the first monthof therapy to 12 months after diagnosis. In contrast, thefrequency of anxiety was elevated at the end of the firstmonth of therapy, but then declined to levels expected ina normative population at 6 months and 12 months afterdiagnosis. Anxiety and depression scores at 1 month afterdiagnosis significantly predicted persistence of symptoms

throughout the first year of therapy. In adjusted analysis,unhealthy family functioning and self-reported Hispanicethnicity were the variables that were found to have thestrongest association with emotional functioning. Chil-dren with unhealthy family functioning were 2.24 timesas likely to have anxiety symptoms and 2.40 times as likelyto have depressive symptoms. Hispanic children were3.35 times as likely as white non-Hispanic children tohave anxiety symptoms, but were not at an increased riskof depressive symptoms. Age, sex, family socioeconomicstatus, and therapeutic randomization did not predictemotional functioning. We did not find behavioralchanges to be a significant problem.

The current study is unique in that our sample was ahomogeneous group of patients with ALL with a highexpected cure rate who were also enrolled on a randomized

Figure 1. Prevalence of elevated anxiety and depression scores in the first year after diagnosis of acute lymphoblastic leukemiaare shown.

Psychological Adjustment of Children With ALL/Myers et al

Cancer May 1, 2014 1421

clinical trial. Thus, we were able to account for any differ-ences due to therapeutic randomization. Because we en-rolled patients at 31 sites from a range of community andtertiary care centers and rural and urban regions, the resultsof the current study can be generalized more than single-institution studies. Furthermore, we had a high participa-tion rate (82% of eligible), which greatly reduced thepotential for selection bias.

Limited published data are available regarding thelongitudinal psychosocial functioning of children cur-rently receiving chemotherapy. The results of the currentstudy can be most closely compared with the prospective

cohort study of 38 patients by Sawyer et al, who alsofound that children experience considerable emotionaldistress in the immediate postdiagnosis period.9,25 How-ever, in contrast to the current study, Sawyer et al con-cluded that by 1 year after diagnosis, children treated forcancer had psychological functioning similar to that ofchildren in the community. This other study had a smallersample size and included a heterogeneous group ofpatients with childhood cancer. A more recent study byFurlong et al prospectively assessed quality of life in chil-dren with SR and high-risk ALL, some of whom receivedcranial radiation. Although their outcomes included

TABLE 2. Univariate Association Between Patient and Family Factors and Anxiety and Depression

Characteristic

Anxiety Depression

OR (95% CI) P OR (95% CI) P

Age group at diagnosis

Preschool (ages 2-4 y) Reference group Reference group

School age (ages 5-12 y) 0.62 (0.37-1.05) .076 0.78 (0.49-1.26) .314

Sex

Male Reference group Reference group

Female 1.62 (0.97-2.72) .067 1.27 (0.79-2.02) .325

Race/ethnicity

White, non-Hispanic Reference group Reference group

Hispanic 3.32 (1.80-6.15) .000 1.36 (0.73-2.53) .335

Black, non-Hispanic 0.83 (0.23-2.96) .769 1.22 (0.46-3.25) .696

Other 1.61 (0.68-3.83) .277 1.59 (0.75-3.37) .226

Annual family income

�$50,000 Reference group Reference group

<$50,000 1.11 (0.63-1.98) .720 1.17 (0.68-2.01) .564

Maternal education

At least some college Reference group Reference group

Less than college 1.15 (0.64-2.04) .642 1.24 (0.74-2.09) .410

Marital status of parents

Married Reference group Reference group

Not married 1.30 (0.74-2.29) .354 1.65 (0.99-2.75) .054

General family functioninga

Healthy family functioning Reference group Reference group

Unhealthy family functioning 3.01 (1.76-5.15) <.001 2.37 (1.45-3.85) .001

Maintaining family integration

coping behaviorsb

0.97 (0.94-0.99) .009 0.96 (0.94-0.98) .001

Maintaining social support

coping behaviorsc

0.96 (0.94-0.99) .005 0.95 (0.93-0.97) <.001

Understanding the medical

situation coping behaviorsd

0.95 (0.90-1.00) .038 0.93 (0.89-0.98) .003

Pain and hurt by parental reporte 0.99 (0.98-1.00) .062 0.99 (0.98-1.00) .016

Nausea by parental reportf 0.99 (0.97-1.01) .186 0.99 (0.98-1.01) .230

Therapeutic randomization

SC/SIM-SDI Reference group Reference group

IC/SIM-SDI 1.20 (0.60-2.40) .598 0.95 (0.50-1.78) .862

SC/AIM-ADI 1.23 (0.59-2.57) .585 1.00 (0.51-1.95) .991

IC/AIM-ADI 1.02 (0.45-2.28) .970 1.14 (0.56-2.33) .713

Abbreviations: 95% CI, 95% confidence interval; AIM-ADI, augmented interim maintenance and augmented delayed intensification; IC, intensified consolida-

tion; SC, standard consolidation; OR, odds ratio; SIM-SDI, standard interim maintenance and standard delayed intensification.a Measured by the General Functioning Scale of the Family Assessment Device (FAD-GF).b Measured by the Coping Health Inventory for Parents (CHIP) subscale 1.c Measured by the CHIP subscale 2.d Measured by the CHIP subscale 3.e Measured by the Pediatric Quality of Life Inventory (PedsQL) 3.0 Cancer Module pain and hurt subscale.f Measured by the PedsQL 3.0 Cancer Module nausea subscale.

Original Article

1422 Cancer May 1, 2014

global quality-of-life health utilities instead of clinicalmeasures of psychological functioning, Furlong et alfound that meaningful distress improved throughouttherapy in a pattern similar to the results presentedherein.10 This study also differed from the current one inthat the instrument used (the Health Utilities Index) wasonly validated for children aged� 5 years, whichexcluded> 50% of the SR-ALL sample. The most recentlongitudinal report measured the prevalence and distressof cancer-related symptoms in children throughout ther-apy.8 This study, which included parents of 89 childrenwith different cancer diagnoses, also found that psycho-logical distress improved throughout therapy, but was stillan issue for a subset of children at the end of therapy.

We found that family functioning was an importantpredictor of emotional functioning among children withcancer. To the best of our knowledge, the majority of pre-vious research has focused on family functioning as anoutcome variable. Maurice-Stam et al studied off-therapychildren and found both positive and negative correla-tions between family functioning and quality of life,

depending on the age of the child.26 In patients with pedi-atric asthma, family dysfunction has been associated withchildren’s mental health.27 The current study used theFAD-GF, which measures perceived family cohesion andthe ability of family members to communicate with eachother. The results of the current study suggest that familieswho demonstrate worse cohesion and communicationshould be considered to be at higher risk and be offeredmore psychosocial support. There are available psychoso-cial risk screening measures such as the validated Psycho-social Assessment Tool (PAT2.0), which can integrateassessments of family functioning.28 Furthermore, familyfunctioning may be a modifiable variable, and thus theresults of the current study support developing family-based interventions that target family functioning.

An important conclusion of the current study is thatanxiety and depression at 1 month after diagnosis signifi-cantly predict the persistence of symptoms throughoutthe first year of therapy. Clinicians have long advocatedthe importance of addressing acute symptoms as part ofmultidisciplinary supportive care during cancer therapy.29

TABLE 3. Multivariate Analysis of the Association Between Patient and Family Factors and Anxiety andDepression

Characteristic

Anxiety Depression

OR (95% CI) P OR (95% CI) P

Age group at diagnosis

Preschool (ages 2-4 y) Reference group Reference group

School age (ages 5-12) 0.49 (0.24-1.01) .053 0.77 (0.42-1.40) .387

Sex

Male Reference group Reference group

Female 1.57 (0.79-3.16) .206 1.24 (0.68-2.25) .478

Race/ethnicity

White, non-Hispanic Reference group Reference group

Hispanic 3.35 (1.36-8.24) .009 0.52 (0.20-1.39) .192

Black, non-Hispanic 0.85 (0.16-4.54) .846 0.86 (0.25-2.99) .815

Other 1.39 (0.42-4.52) .592 1.10 (0.42-2.87) .849

Marital status of parents

Married Reference group Reference group

Not married 1.15 (0.49-2.56) .797 2.36 (1.17-4.75) .017

General family functioninga

Healthy family functioning Reference group Reference group

Unhealthy family functioning 2.24 (1.07-4.70) .033 2.40 (1.26-4.56) .008

Maintaining family integration

coping behaviorsb

0.99 (0.94-1.05) .771 1.04 (0.99-1.10) .085

Maintaining social support

coping behaviorsc

0.98 (0.94-1.03) .366 0.94 (0.91-0.98) .004

Understanding the medical

situation coping behaviorsd

1.00 (0.90-1.10) .964 0.95 (0.88-1.04) .283

Pain and hurt by parental reporte 0.99 (0.98-1.00) .152 0.99 (0.98-1.0) .049

Abbreviations: 95% CI, 95% confidence interval; OR, odds ratio.a Measured by the General Functioning Scale of the Family Assessment Device (FAD-GF).b Measured by the Coping Health Inventory for Parents (CHIP) subscale 1.c Measured by the CHIP subscale 2.d Measured by the CHIP subscale 3.e Measured by the Pediatric Quality of Life Inventory (PedsQL) 3.0 Cancer Module pain and hurt subscale.

Psychological Adjustment of Children With ALL/Myers et al

Cancer May 1, 2014 1423

The results of the current study further highlight the im-portance of the early identification of and interventionamong distressed children to avoid long-term emotionaldistress. Previous studies in the general population havesuggested that anxiety in children may result in the laterdevelopment of depressive disorders and substanceabuse.30 Studies in non-cancer pediatric populations havealso found that there are inherited factors that contributeto an individual’s risk of responding to adversity in earlylife with depression and anxiety.31 Such stressors canresult in changes in the corticotropin-releasing factor sys-tem in genetically predisposed individuals.

Although symptoms of anxiety and depression maylessen throughout therapy, it is important to recognize thedistress they cause and provide appropriate psychosocialinterventions. The National Institutes of Health State-of-the-Science Conference on Symptom Management in Can-cer: Pain, Depression, and Fatigue concluded that: “Allpatients with cancer should have optimal symptom controlfrom diagnosis throughout the course of illness, regardlessof personal and cultural characteristics.”32 A wealth of psy-chosocial interventions exists for children with cancer,including cognitive behavioral therapy, social and recrea-tional activities, and psychoeducational interventions.33

Many interventions use family-based methods, which havebeen associated with beneficial outcomes for children.34

Understanding the efficacy of various interventions is anongoing area of research.33

The association between Hispanic ethnicity andanxiety is novel in children with ALL. To make sure thiswas not due to a methodological error in using an inap-propriate comparison group, we verified that the frequen-cies of Hispanic children in the BASC-2 PRS normativecomparison group (16.5%-20%) were similar to that inthe current study population group (16.4%).19 There is asingle report that demonstrated poorer emotional func-tioning in Hispanic children with cancer, but it onlyaddressed the off-treatment period.35 There are other dif-ferences between Hispanic and non-Hispanic childrenwith ALL, including inferior survival rates in Hispanicchildren.1,36 We do not currently have the data availableto explain how Hispanic ethnicity leads to worse psycho-logical functioning; but given these differences, this areaof study deserves further attention.

The current study has some methodological characteris-tics that should be considered in interpreting the results. First,some patients enrolled in the study did not complete the eval-uations at all the required timepoints due to withdrawalsfrom the therapeutic study, administrative errors at studysites, and/or incomplete forms. Second, only parent report

was available for the majority (89.3%) of children in the cur-rent study because there was no self-report option of theBASC-2 available for children aged< 8 years. In fact, to thebest of our knowledge, there are no validated, objective self-report assessment tools for emotional functioning for childrenaged< 5 years and few for those aged 5 to 8 years. Youngchildren may lack the language abilities to report on and/orthe cognitive capacity to reflect on one’s own behaviors orfeelings. Children as old as 11 years tend to report fewer psy-chiatric symptoms and are unreliable in reporting about timefactors, such as duration or frequency of symptoms,37 andtherefore collateral information is emphasized.38 Assessingchildren of preschool age is typically accomplished throughparent report and observation.39 In addition, we were able toassociate family functioning with emotional functioning, butwe were unable to determine the direction of the associationfrom our data. It may be that children with better emotionalfunctioning lessen the burden on their families.

Based on the results of this large, multisite, cohortstudy of children treated for SR-ALL, we conclude thatsymptoms of depression and anxiety are a significant prob-lem in the immediate postdiagnosis period. Although anxi-ety symptoms lessen after the first month of therapy,depressive symptoms appear to persist throughout at leastthe first year. We also found that we can identify children at1 month after diagnosis who are substantially more likely tohave worse psychological functioning throughout the firstyear of therapy. These results are highly relevant to pediatriconcologists who should be screening for clinical levels of anx-iety and depression starting in the early postdiagnosis periodfor at least 1 year after diagnosis. Furthermore, the results ofthe current study highlight high-risk groups who shouldreceive additional psychosocial support, including childrenof Hispanic ethnicity or those with parent-reported unheal-thy family functioning. Further studies are needed to developevidence-based interventions, including those targeting fam-ily functioning and patient subgroups at greatest risk, to pre-vent and treat anxiety and depression in children with ALL.

FUNDING SUPPORTThis research was supported by grants from the National Institutesof Health to the Children’s Oncology Group including CA13539,CA98543, and a Community Cancer Oncology Program grantfrom the Division of Cancer Prevention of the National CancerInstitute to the Children’s Oncology Group.

CONFLICT OF INTEREST DISCLOSURESDr. Hunger is the Ergen Family Chair in Pediatric Cancer. Dr.Kadan-Lottick is supported in part by American Cancer SocietyScholar Grant 119700-RSGHP-10-107-01-CPHPS and a TeamBrent St. Baldrick’s Foundation Scholar award.

Original Article

1424 Cancer May 1, 2014

REFERENCES1. Hunger SP, Lu X, Devidas M, et al. Improved survival for children

and adolescents with acute lymphoblastic leukemia between 1990and 2005: a report from the children’s oncology group. J ClinOncol. 2012;30:1663-1669.

2. Matloub Y, Bostrom BC, Hunger SP, et al. Escalating intravenousmethotrexate improves event-free survival in children with standard-risk acute lymphoblastic leukemia: a report from the Children’s On-cology Group. Blood. 2011;118:243-251.

3. Hochhauser CJ, Lewis M, Kamen BA, et al. Steroid-induced altera-tions of mood and behavior in children during treatment for acutelymphoblastic leukemia. Support Care Cancer. 2005;13:967-974.

4. Sung L, Yanofsky R, Klaassen RJ, et al. Quality of life during activetreatment for pediatric acute lymphoblastic leukemia. Int J Cancer.2011;128:1213-1220.

5. Pickard AS, Topfer LA, Feeny DH. A structured review of studieson health-related quality of life and economic evaluation in pediatricacute lymphoblastic leukemia. J Natl Cancer Inst Monogr. 2004;(33):102-125.

6. Shankar S, Robison L, Jenney ME, et al. Health-related quality of lifein young survivors of childhood cancer using the Minneapolis-Manchester Quality of Life-Youth Form. Pediatrics. 2005;115:435-442.

7. Waters EB, Wake MA, Hesketh KD, et al. Health-related quality oflife of children with acute lymphoblastic leukaemia: comparisonsand correlations between parent and clinician reports. Int J Cancer.2003;103:514-518.

8. Heden L, Poder U, von Essen L, Ljungman G. Parents’ perceptionsof their child’s symptom burden during and after cancer treatment. JPain Symptom Manage. 2013;46:366-375.

9. Sawyer M, Antoniou G, Toogood I, et al. Childhood cancer: a 4-year prospective study of the psychological adjustment of childrenand parents. J Pediatr Hematol Oncol. 2000;22:214-220.

10. Furlong W, Rae C, Feeny D, et al. Health-related quality of lifeamong children with acute lymphoblastic leukemia. Pediatr BloodCancer. 2012;59:717-724.

11. Meeske K, Katz ER, Palmer SN, et al. Parent proxy-reported health-related quality of life and fatigue in pediatric patients diagnosed withbrain tumors and acute lymphoblastic leukemia. Cancer. 2004;101:2116-2125.

12. Landolt MA, Vollrath M, Niggli FK, Gnehm HE, Sennhauser FH.Health-related quality of life in children with newly diagnosed cancer:a one year follow-up study. Health Qual Life Outcomes. 2006;4:63.

13. Mounir GM, Abolfotouh MA. Assessment of health related qualityof life among school children with cancer in Alexandria. J Egypt Pub-lic Health Assoc. 2007;82:219-238.

14. Klassen AF, Anthony SJ, Khan A, et al. Identifying determinants ofquality of life of children with cancer and childhood cancer survi-vors: a systematic review. Support Care Cancer. 2011;19:1275-1287.

15. Kazak AE, Alderfer MA, Streisand R, et al. Treatment of posttrau-matic stress symptoms in adolescent survivors of childhood cancerand their families: a randomized clinical trial. J Fam Psychol. 2004;18:493-504.

16. Chesla CA. Do family interventions improve health&quest; J FamNurs. 2010;16:355-377.

17. Smith M, Arthur D, Camitta B, et al. Uniform approach to riskclassification and treatment assignment for children with acute lym-phoblastic leukemia. J Clin Oncol. 1996;14:18-24.

18. Myers R, Balsamo L, Carroll W, Hunger S, Winick M, et al. Emo-tional and behavioral functioning in the first year after diagnosis ofstandard risk (SR) acute lymphoblastic leukemia (ALL): a reportfrom Children’s Oncology Group (COG) AALL0331. Presented at:American Society of Pediatric Hematology/Oncology 25th AnnualMeeting; May 9-12, 2012; New Orleans, LA.

19. Reynolds CR, Kamphaus RW. Behavior Assessment System for Chil-dren. 2nd ed. Circle Pine, MN: American Guidance Service; 2004.

20. Wolfe-Christensen C, Mullins LL, Stinnett TA, Carpentier MY,Fedele DA. Use of the Behavioral Assessment System for Children

2nd Edition: Parent Report Scale in pediatric cancer populations. JClin Psychol Med Settings. 2009;16:322-330.

21. Carpentieri SC, Meyer EA, Delaney BL, et al. Psychosocial and be-havioral functioning among pediatric brain tumor survivors. J Neuro-oncol. 2003;63:279-287.

22. Miller IW, Bishop DS, Epstein NB, et al. The Mcmaster FamilyAssessment Device-Reliability and Validity. J Marital Fam Ther.1985;11:345-356.

23. McCubbin H, McCubbin M, Nevin R, et al. Coping Health Inven-tory for Parents. In: McCubbin HI, Thompson AI, eds. FamilyAssessment Inventories for Research and Practice. Madison, WI:University of Wisconsin-Madison; 1987.

24. Varni JW, Burwinkle TM, Katz ER, Meeske K, Dickinson P. ThePedsQL in pediatric cancer: reliability and validity of the PediatricQuality of Life Inventory Generic Core Scales, Multidimensional Fa-tigue Scale, and Cancer Module. Cancer. 2002;94:2090-2106.

25. Sawyer M, Antoniou G, Toogood I, Rice M. Childhood cancer: atwo-year prospective study of the psychological adjustment of chil-dren and parents. J Am Acad Child Adolesc Psychiatry. 1997;36:1736-1743.

26. Maurice-Stam H, Grootenhuis MA, Brons PP, Caron HN, Last BF.Psychosocial indicators of health-related quality of life in childrenwith cancer 2 months after end of successful treatment. J PediatrHematol Oncol. 2007;29:540-550.

27. Sawyer MG, Spurrier N, Whaites L, et al. The relationship betweenasthma severity, family functioning and the health-related quality oflife of children with asthma. Qual Life Res. 2000;9:1105-1115.

28. Pai AL, Patino-Fernandez AM, McSherry M, et al. The PsychosocialAssessment Tool (PAT2.0): psychometric properties of a screener forpsychosocial distress in families of children newly diagnosed withcancer. J Pediatr Psychol. 2008;33:50-62.

29. van de Wetering MD, Schouten-van Meeteren NY. Supportive carefor children with cancer. Semin Oncol. 2011;38:374-379.

30. Kendall PC, Safford S, Flannery-Schroeder E, Webb A. Child anxi-ety treatment: outcomes in adolescence and impact on substance useand depression at 7.4-year follow-up. J Consult Clin Psychol. 2004;72:276-287.

31. Nemeroff CB, Vale WW. The neurobiology of depression: inroadsto treatment and new drug discovery. J Clin Psychiatry. 2005;66(suppl 7):5-13.

32. Patrick DL, Ferketich SL, Frame PS, National Institutes of HealthState-of-the-Science Panel. National Institutes of Health State-of-the-Science Conference Statement: Symptom management in cancer:pain, depression, and fatigue, July 15-17, 2002. J Natl Cancer InstMonogr. 2004;(32):9-16.

33. Pai AL, Drotar D, Zebracki K, Moore M, Youngstrom E. A meta-analysis of the effects of psychological interventions in pediatric on-cology on outcomes of psychological distress and adjustment. JPediatr Psychol. 2006;31:978-988.

34. Meyler E, Guerin S, Kiernan G, et al. Review of family-based psy-chosocial interventions for childhood cancer. J Pediatr Psychol. 2010;35:1116-1132.

35. Meeske KA, Patel SK, Palmer SN, et al. Factors associated withhealth-related quality of life in pediatric cancer survivors. PediatrBlood Cancer. 2007;49:298-305.

36. Goggins WB, Lo FF. Racial and ethnic disparities in survival of USchildren with acute lymphoblastic leukemia: evidence from theSEER database 1988-2008. Cancer Causes Control. 2012;23:737-743.

37. Schwab-Stone M, Fallon T, Briggs M, Crowther B. Reliability ofdiagnostic reporting for children aged 6-11 years: a test-retest studyof the Diagnostic Interview Schedule for Children-Revised. Am JPsychiatry. 1994;151:1048-1054.

38. Chrisman A, Egger H, Compton SN, et al. Assessment of childhooddepression. Child Adolesc Mental Health. 2006;11:111-116.

39. Luby J, Tandon M. Assessing the preschool-age child. In: DulcanM, ed. Dulcan’s Textbook of Child and Adolescent Psychiatry.Arlington, VA: American Psychiatric Publishing; 2010.

Psychological Adjustment of Children With ALL/Myers et al

Cancer May 1, 2014 1425