Anxiety, Depressive, and Posttraumatic Stress Disorders Among Mothers of Pediatric Survivors of Hematopoietic Stem Cell Transplantation
Objective. To examine the prevalence and predictors of anxiety, depression, and posttraumatic stress disorder among mothers of children who underwent hematopoietic stem cell transplantation (HSCT).
Methods. A total of 111 mothers of children who survived HSCT completed self-report measures of psychological functioning at the time of HSCT and self-report measures and a structured psychiatric interview 18 months after HSCT. Demographic data and medical data were also collected.
Results. Approximately 20% of mothers had clinically significant distress reactions. When subthreshold posttraumatic stress disorder was included, nearly one third of mothers met the criterion for persistent distress. Mothers with 1 of the 4 psychiatric diagnoses reported a significantly lower quality of life, suggesting that the diagnoses had an impact on their general quality of life. Mothers who had 1 of the 4 diagnoses at 18 months were younger and had higher anxiety and depressive symptoms at the time of HSCT.
Conclusions. A subset of mothers of children who survive HSCT may require psychological intervention. Mothers who are most at risk are younger and evidence anxiety and depressive symptoms at the time of transplantation.
Hematopoietic stem cell transplantation (HSCT) is a physically demanding treatment procedure that involves a preparatory regimen of intensive chemotherapy or radiation or both, followed by an infusion of marrow or peripheral stem cells from a donor (allogeneic transplantation) or from the patient (autologous transplantation). HSCT is considered to be a first-line therapy for many life-threatening hematologic and oncologic diseases. Advances in supportive care such as treatment of graft-versus-host disease (GvHD) have improved survival rates, and improvements in donor marrow preparation options have resulted in an increased donor pool and wider variety of diseases for which HSCT is applicable. Both advances have resulted in a dramatic increase in the utilization of HSCT in the treatment of pediatric malignancies and congenital and acquired disorders in the past decade.1
Although potentially curative, HSCT is still associated with significant morbidity and mortality. The preparation for HSCT involves high doses of chemotherapy that are frequently combined with total body irradiation, which are extremely toxic. Patients typically undergo several weeks of marrow aplasia, which can result in infection and bleeding. Overall, HSCT can result in lengthy hospital stays and recovery periods. Short-term medical complications include severe mucositis, hepatic toxicities, viral infections, septicemia, intracranial bleeding, and acute GvHD.2–5 Long-term medical complications include pulmonary disease,6 oral and craniofacial and oral complications,7 chronic GvHD,8 and second malignancies and disease recurrence.9,10 Despite drastically improved survival statistics, HSCT remains a risky procedure. Depending on the type of transplant and disease, transplant-related mortality ranges between 20% and 80% 1 year after transplantation.1
The significant medical risks along with the difficult experience of witnessing one’s child undergoing an aversive medical procedure such as HSCT can result in both acute and persistent psychological distress reactions among parents. The available literature on the psychological impact of HSCT suggests that the time immediately before and after the transplant can be particularly upsetting for parents. Nelson et al11 found that 66% of mothers had scores on a depression inventory that were in the clinical range at the time of HSCT admission, with rates of depression declining to 52% by hospital discharge. Barrera et al12 found that 8% of mothers reported depressive symptoms in the clinical range and 50% of mothers reported anxiety levels over the clinical cutoff immediately before transplantation. Streisand et al13 assessed parents a number of times between 1 week before and 21 days after HSCT and found that stress levels were highest between prehospital admission and 7 days after HSCT infusion. Few studies have used psychiatric interview methods to quantify clinically significant distress reactions. In our previous work, we evaluated a large sample of mothers (N = 115) at the time of HSCT using questionnaire instruments as a screening tool for additional assessment using a psychiatry interview.14 Among the subset of mothers who were administered the psychiatric interview (N = 64), 20% received a diagnosis of generalized anxiety disorder (GAD), major depressive disorder (MDD), or panic disorder (PD).
Relatively few studies have evaluated long-term distress reactions among parents of children who survive HSCT. Most of the published work has been qualitative or consisted of clinical descriptions.15,16 Sormanti et al17 conducted 1 of the earliest long-term follow-up studies, interviewing parents of children who had undergone transplantation between 1 and 10 years before to identify sources of stress. Parents reported concerns about recurrence, insurance, their own physical and mental health, and finances. In addition to anxiety and depressive symptoms, recent long-term studies have reported that parents of children who have survived HSCT may experience posttraumatic stress reactions.18 Indeed, learning that one’s child has a life-threatening disease is a qualifying traumatic event for posttraumatic stress disorder (PTSD).19 PTSD symptoms include intrusive worries or reexperiencing aspects of the traumatic event, avoidance of reminders of the event or a numbing of emotions, and hypervigilance or increased physiologic arousal. Little is known about the prevalence of long-term posttraumatic stress reactions among parents of children who survive HSCT. In our previous work, we examined the prevalence of maternal PTSD 6 months after HSCT.18 Ninety mothers completed a questionnaire assessment of PTSD symptoms, but only a subset of mothers who met a cutoff score on the paper-and-pencil measure (N = 40) were administered the psychiatric interview for PTSD. Approximately 8% of mothers received a diagnosis of current PTSD. Some information about parental PTSD reactions may be obtained from studies of parents of children who undergo other types of transplantation, as the stressors are similar. Young et al20 examined the incidence and severity of PTSD symptoms among parents of pediatric solid organ transplant recipients, assessing parents between 10 and 38 months after the most recent transplant. No studies have reported on the psychological impact of pediatric HSCT on parents beyond 6 months after HSCT or have formally assessed a spectrum of psychiatric responses and assessed parents at a standardized point in time after HSCT. Such information is important to health care providers who work with families because it can assist in delineating long-term psychosocial needs of parents.
In the present study, we focused on distress responses of mothers. We targeted mothers because previous studies of parents of pediatric cancer patients have shown that mothers report more distress than other family members.21 One reason that mothers may be at particular risk is that they are typically the primary caregivers of their children during and after transplantation13,22 and thus may carry more parenting stress and witness a greater number of upsetting medical events than fathers.
Understanding why some mothers develop long-term severe distress reactions such as anxiety and depressive disorders and PTSD and other mothers do not has received relatively little empirical attention. To identify risk factors for long-term distress reactions, it may prove beneficial to evaluate the role of characteristics and resources of mothers who are present at the time of the transplantation. Focusing on risk factors that are present at the time of transplantation may prove beneficial for 2 reasons. From a clinical perspective, this is the time when health care staff conduct comprehensive psychosocial evaluations of families and also the time when staff can provide intensive psychosocial services to mothers while the child is hospitalized. From a theoretical perspective, conceptualizations of how people process traumatic experiences focus on perceptions and reactions during the “exposure” to the stressful event, psychological vulnerability factors, and the resources that the person brings to the experience as key predictors of later reactions.23,24
In terms of reactions and perceptions during the stressful experience, severe emotional reactions to the experience at the time of the stressful event are likely to be a risk factor because they are likely to indicate greater appraisals of fear and greater perceived life threat to the child. Indeed, studies of PTSD have shown that emotional reactions during the event (assessed retrospectively) are predictive of later PTSD.25,26 More severe emotional reactions at the time of transplantation may also indicate greater vulnerability to psychiatric problems. Two potential psychological vulnerability factors may predict mothers’ long-term distress reactions. First, psychiatric history before the event is likely to be a key predictor of long-term emotional distress.27 Second, both theoretical conceptualizations of reactions to traumatic life experiences28 and a large body of literature consider exposure to previous traumatic life experiences as a critical vulnerability factor in the development of posttraumatic stress reactions,29 depression,30 and PD.31 Traumatic experiences before the HSCT may sensitize the mother by adding to her sense of life as being difficult and traumatic. In terms of resilience factors, one of the most well-documented resources to facilitate recovery and prevent the development of distress reactions to traumatic life experiences is social support.32 Parent stress at the time of HSCT11,33 and after HSCT17 has been shown to be associated with social support.
Finally, we consider demographic characteristics and medical risk and post-HSCT course factors in maternal distress reactions. Demographic characteristics such as the gender of the child, the age of the child and the parent, maternal ethnicity, education, and income may each contribute to the development of long-term distress reactions. Mothers of children who undergo more risky transplantations and have a more difficult posttransplantation course may be more likely to develop long-term reactions. A more risky transplantation with a more difficult post-HSCT course would represent greater exposure to potentially upsetting experiences and a greater potential threat to the child’s life. According to traumatic stress theories, these factors increase vulnerability to long-term psychological problems.34
The present study was designed to address limitations of previous studies of long-term distress responses among mothers of children who underwent HSCT. We used a longitudinal, prospective design and used a structured psychiatric interview to evaluate the presence of clinical levels of anxiety, depression, and PTSD. Vulnerability and resilience factors were assessed at the time of HSCT and used to predict individual differences in the presence of clinical levels of severe reactions 18 months after HSCT. A variety of posttransplantation medical course factors were assessed. The study had 4 main aims: 1) to document MDD, GAD, PD, and PTSD at 18 months after HSCT; 2) to estimate the comorbidity of MDD, GAD, PD, and PTSD; 3) to evaluate whether demographic variables, emotional reactions at the time of HSCT, psychological vulnerability factors (previous psychiatric history, previous traumatic life events), and psychological resilience factors (social support), assessed at the time of HSCT, as well as post-HSCT medical course, were associated with the 4 psychiatric disorders 18 months later; and 4) to examine the association between these disorders and maternal quality of life. It was hypothesized that mothers who reported more distress at the time of HSCT, who had a previous psychiatric history and had experienced more lifetime traumatic events, and who had less social support and had children with more post-HSCT complications would be more likely to have 1 of the 4 disorders at 18 months after HSCT. We also predicted that mothers with 1 of the 4 disorders would report a lower quality of life. The results reported in this article are part of a larger study of psychological impact of HSCT on mothers.
Participants were 111 mothers of children who were undergoing HSCT. The present investigation uses data from participants in a larger longitudinal study of maternal distress and coping after pediatric bone marrow transplantation. Potential participants were recruited at 1 of 6 participating hospital units where HSCTs are performed. Eligibility requirements for inclusion were 1) mother was able to read and write English, 2) mother was the primary caregiver for the child undergoing HSCT, 3) mother was >18, and 4) child undergoing HSCT was ≤21. Of 416 mothers approached for study participation, 130 (31%) mothers declined participation and 286 consented. The most common reasons cited for nonparticipation included being overwhelmed (25%) and not wanting to leave the child’s bedside (13%). At 18 months, 12 (4%) mothers were not included because they were not yet due for their 18-month assessment, 67 (23%) dropped out of the study, 10 (3%) mothers did not complete the Structured Clinical Interview for DSM-IV, NonPatient Version (SCID-NP) interview, 1 (0.3%) mother completed the interview but not a questionnaire at the 18-month follow-up, and for 85 (30%) mothers, the child died before the assessment date. The final sample comprised 111 mothers. Descriptive information on the study sample is contained in Table 1, and descriptive medical information about the children is contained in Table 2.
Before the date of HSCT infusion, mothers completed the anxiety and depressive symptom inventories, the Impact of Events Scale (IES), the traumatic events inventory, the social support inventory, and the assessment of previous and current psychiatric care (baseline survey). Baseline surveys were obtained before the date of the HSCT infusion in all but 3 cases (3 surveys were administered between 1 and 4 days after the infusion). The average time of baseline survey administration was 2 days before HSCT infusion.
Eighteen months after HSCT infusion, mothers completed the Axis I SCID-NP35 and a quality-of-life measure (18-month post-HSCT survey). The average time between the baseline and 18-month assessments was 19 months (range: 17.5–22 months). Medical information was collected from the child’s medical chart at the time of HSCT and 18 months later.
Measures Administered at the Time of HSCT
Demographic variables included the mother’s age, marital status, ethnicity, education level, family income, child gender, and child age.
Anxiety and depressive symptoms were measured using the Beck Anxiety Inventory (BAI)36 and the Beck Depression Inventory (BDI).37 The BAI is a 21-item self-report instrument that is used to assess symptoms of anxiety and is specifically developed to focus on symptoms that are distinct from depression symptoms. Internal consistency for the present sample was excellent (α = .91). The BDI is a widely used 21-item scale to assess depressive symptoms. Internal consistency in the present sample was excellent (α = .86). For purposes of the present study, the BAI and BDI scores were summed to calculate a total distress score.
Posttraumatic stress symptoms were measured using the IES.38 The IES is a 15-item, widely used scale to assess posttraumatic stress symptoms. The scale has 2 subscales, Intrusions and Avoidance. The 7-item Intrusive thoughts scale was used to measure the frequency with which mothers had intrusive thoughts about her child’s cancer and its treatment. The 8-item Avoidance scale measured attempts to avoid thinking about the child’s cancer and its treatment. Mothers rated frequency of these behaviors over the course of the past month using a 4-point Likert scale (0 = not at all; 5 = often; α = .84). Despite poor content validity as a measure of PTSD, the IES has been shown to discriminate a variety of traumatized groups from nontraumatized cohorts,39 to differentiate between those who receive a PTSD diagnosis and those who do not,40 to correlate highly with other measures of posttraumatic stress,41 and to be sensitive to changes in response to trauma.41,42
Previous Traumatic Events
The Traumatic Stress Schedule43 is an 8-item, face-valid screening instrument that assesses the lifetime occurrence of traumatic events, including such events as being the victim of a crime or sustaining injury or damage as a result of a natural disaster or serious accident. Scores can range from 0 to 8.
Three questions assessed whether the mother had ever sought counseling for emotional problems, seen a therapist or counselor for drug or alcohol problems, and ever been hospitalized for psychological problems (yes/no). Scores can range from 0 to 3. Mothers were also asked whether they had seen a religious counselor, professional counselor or therapist, or support or therapy group in the past month (yes/no). Scores can range from 0 to 3.
Family and Friend Support
Twenty-six items adapted from the emotional support and cognitive information subscales of Inventory of Socially Supportive Behaviors44,45 and the positive reappraisal and acceptance scales of the Ways of Coping46 and COPE47 assess enacted social support. Items reflected attempts by family and friends to assist the mother in efforts to look at the situation in a more meaningful or a more positive light, as well as items that reflect any assistance in finding some degree of acceptance in the situation. Internal consistency in the present sample was .88.
Measures Administered at the 18-Month Follow-up
The SCID-NP35,48,49 is a semistructured clinical interview that results in a clinical diagnosis in accordance with Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition criteria. Four modules of the SCID-NP were administered: the MDD, GAD, PD, and PTSD modules.36 The SCID-NP-PTSD assesses whether the stressor criterion has been met (criterion A) and the presence of 17 different PTSD symptoms. Criterion A assesses whether the symptoms occurred after exposure to an event involving actual or threatened death or physical injury or a threat to the physical integrity of self or others. The response to the event must involve fear, helplessness, or horror. Criterion B symptoms assess persistent reexperiencing of the event. Criterion C assesses avoidance of stimuli associated with the stressor and a numbing of emotional responsiveness. Criterion D assesses physiologic arousal occurring after the experience. A diagnosis of PTSD requires at least 1 criterion B, 3 criterion C, and 2 criterion D symptoms. Symptoms must have been present for at least 1 month (criterion E) and be associated with impaired social and occupational functioning (criterion F). The responses are rated using a 4-choice rating: inadequate information, absent, subthreshold, and threshold. Only responses rated as meeting threshold are considered to have met the criterion for each symptom. Questions were specifically linked to the child’s HSCT. It should be noted that all items were asked, even when the mother did not endorse both symptoms for criterion A.
Interviews were conducted by research assistants at each study site. The interviewers were trained in the administration of the SCID-NP using the SCID-NP manual, audiotaped sample interviews, and practice interviews. Training was conducted by a licensed psychologist with a specialization in SCID-NP administration and scoring (J.D.). Additional training and supervision was provided during the course of data collection (J.D.). During these sessions, symptom endorsement and diagnosis for participants were discussed, to guard against interviewer drift and ensure accuracy of scoring.
For establishing interrater reliability for MDD, GAD, and PD, a randomly selected subset of 23 audiotapes (21% of taped interviews) were reviewed and scored by a second rater, a licensed psychologist with a specialization in SCID-NP administration and scoring. The rater was blind to other assessment scores and identifying information about mother and child. For establishing interrater reliability for PTSD, a randomly selected subset of 25 (23%) audiotapes were reviewed using the same procedures. For establishing interrater reliability between the interviewers and the SCID-NP expert, their ratings were compared using κ coefficients. Standard criteria for acceptable κ levels do not currently exist. Fleiss50 characterized κ of .40 to .60 as fair, .60 to .75 as good, and >.75 as excellent.
Quality of life in the month preceding the 18-month assessment was measured with the physical and mental health summary scales of the SF-36 Health Survey,51 scored as described by Ware.52 The mental health scale is composed of items that measure vitality, social functioning, the impact of emotional factors on normative role functioning, and mental health. The physical health scale is composed of items that measure general health perceptions, difficulty completing activities of daily living, interference of physical problems on daily activities, and pain. The SF-36 has demonstrated excellent internal reliability in past research.52
Variables that reflected characteristics of the child’s transplant were collected from the medical chart at the time of the transplant hospitalization and over the course of the 18-month post-HSCT period. The variables were as follows:
Time from initial disease diagnosis to transplant, calculated in days
Disease risk, which was based on how advanced the child’s disease was as well as the likelihood of transplant success; the variable had 4 categories: nonmalignant disease, early-stage disease (eg, acute lymphocytic leukemia in continuous remission), intermediate-stage disease (eg, acute lymphocytic leukemia in second or greater remission), and advanced-stage disease (eg, refractory leukemias)
Transplant risk, which was based on the type of transplant and disease; there were 3 risk categories: good (eg, disease in continuous remission), intermediate (eg, matched unrelated donor), and poor risk (eg, refractory or relapsed disease, unmatched donor)
Total number of transfers to the intensive care unit during the HSCT hospitalization
Bearman organ toxicity,53 which was calculated at HSCT discharge (range: 0-10)
Presence of acute GvHD, calculated 3 months after discharge (yes/no)
Total number of nonlocalized infections, calculated 3 months after HSCT discharge
Number of days hospitalized for the HSCT
Whether the child developed chronic GvHD at any point over the 18-month follow-up
Whether the disease relapsed during the 18-month follow-up
Comparisons of Participants Versus Study Noncompleters
The 111 mothers who provided complete data were compared with the 78 mothers who either dropped out of the study after the baseline survey or did not provide complete data at 18 months with regard to available demographic, medical, and baseline data. There were significant differences with regard to ethnicity and psychiatric history. Mothers who did not complete the follow-up assessment were more likely to be nonwhite (χ2  = 4.2, P < .05; percent nonwhite, participant group: 22%; percent nonwhite, study drops: 36%) and less likely to have a psychiatric history (χ2 = 7.9, P < .05; percent previous psychiatric contact, participant group: 53%; percent previous psychiatric contact, study drops: 33%). There were no significant differences with regard to baseline anxiety, depressive symptoms, IES scores, maternal or child demographic or child medical risk, or available post-HSCT course variables.
Prevalence of Psychiatric Diagnoses 18-Months After HSCT
Inter-rater reliability for the SCID-NP in the present study was evaluated using the κ statistic. Using Fleiss’s criteria, results indicated that κ for SCID-NP diagnoses ranged from good to excellent. The κ for MDD was .65, the κ for GAD was 1.0, and the κ for PTSD was .75. The κ for PD could not be computed because none of the cases used for interrater agreement was diagnosed with PD.
Prevalence figures are reported in Table 3. In terms of MDD, 10.8% of mothers received a diagnosis of current depressive disorder. Approximately 4% of mothers received a diagnosis of GAD, and ∼5% received a diagnosis of PD. Approximately 12% received a diagnosis of PTSD. Partial PTSD, which is defined as meeting the stressor criterion A and criteria for 2 of the 3 PTSD symptom clusters, was present in an additional 7.2% mothers (n = 8). In terms of PTSD symptom clusters endorsed, 32% of mothers met criterion A (stressor criterion), 49% of mothers met criterion B (reexperiencing), 21% met criterion C (avoidance, numbing), and 32% of mothers met criterion D (physiologic arousal). Thus, nearly half of mothers had significant post-HSCT reexperiencing symptoms, and one third reported physiologic arousal (irritability, jumpiness). The most commonly endorsed symptoms that met threshold criteria on criterion B were reminders of the transplant evoking significant stress reactions (30%) and recurrent and intrusive recollections about the transplant (29%). The most commonly endorsed symptoms on criterion C were the sense of a foreshortened future (30%) and feelings of detachment or estrangement from others (21%). The most commonly endorsed symptoms on criterion D were difficulties with sleep (35%), irritability (26%), difficulty concentrating (25%), and hypervigilance (22%).
Comorbidity of Depression, Anxiety, and PTSD
A total of 20 (18%) mothers received 1 or more of the 4 diagnoses on the basis of the SCID-NP. Of these 20 mothers, 7 (35%) had comorbid diagnoses. Six of the 7 mothers had PTSD as a comorbid diagnosis. One mother received a diagnosis of PD, GAD, and MDD; one mother received a diagnosis of PD and PTSD; 2 mothers received a diagnosis of MDD and PTSD; 2 mothers received a diagnosis of MDD, GAD, and PTSD; and 1 mother received a diagnosis of MDD, GAD, PD, and PTSD.
Psychological, Demographic, and Objective Medical Factors That Predict Psychiatric Diagnoses 18 Months After HSCT
We used a dichotomous variable indicating whether the participant received any of the 4 psychiatric diagnoses as the dependent variable in this analysis. This strategy was selected for 2 reasons. First, there were a small number of diagnoses of each disorder and combining the diagnoses would result in a larger group for statistical comparison. Second, as noted above, there was evidence of comorbidity. The analyses that examined predictors of SCID-NP diagnosis were conducted in 2 steps. First, 3 separate logistic analyses were conducted for each set of predictors to determine whether the predictor was associated with the outcome variable. This step was taken to identify a smaller set of predictors from the larger set of 18 potential predictors. Step 1 included demographic variables: hospital site, maternal age, maternal education, marital status (married or not married), ethnicity (white and nonwhite), child gender, and family size (child and maternal age are highly correlated so only maternal age was used). Step 2 included the 10 transplant predictors. Step 3 included previous psychiatric history at baseline, therapy use in the month before the baseline, lifetime traumatic events, family and friend support at baseline, and IES score and psychological distress at baseline. Results suggested the following significant predictors for each category: step 1, maternal age; step 2, days hospitalized for HSCT; and step 3, psychological distress at baseline. Younger mothers, mothers of children who had shorter hospitalizations, and mothers who reported more anxiety and depressive symptoms at baseline were more likely to receive 1 of the 4 diagnoses.
On the basis of these analyses, these 4 variables were used as predictors in a stepwise logistic regression. Maternal age was entered at step 1, time hospitalized was entered at step 3, and baseline depressive and anxiety symptoms were included at step 3. Table 4 reports the results of the logistic regression. Maternal age and anxiety and depressive symptoms at the time of HSCT emerged as significant predictors. Table 5 presents descriptive information about mothers with and without a diagnosis. Younger mothers and mothers who reported more anxiety and depressive symptoms at the time of transplantation were more likely to have 1 of the 4 disorders.
Psychiatric Diagnoses and Quality of Life
We compared the physical and mental health functioning of mothers with 1 of the 4 psychiatric disorders with the functioning of mothers who did not receive a psychiatric diagnosis, using t tests. Results indicated significant differences in terms of mental health functioning scores (meandiagnosis: 36.5; standard deviation [SD]: 15.1; meanno diagnosis: 52.3; SD: 7.3; t : 4.3, P < .001) but nonsignificant differences in physical health functioning scores (meandiagnosis: 55.9; SD: 11.2; meanno diagnosis: 55.7; SD: 5.1; t  = −0.16). Having 1 of the 4 psychiatric diagnoses was associated with lower mental health scores on the SF-36.
Watching one’s child undergo transplantation can be a traumatic experience for mothers and can potentially lead to the development of long-term psychological distress responses. The purpose of this study was to evaluate the prevalence of severe distress 18 months after HSCT and to identify psychological vulnerability factors and resilience factors present at the time of HSCT, as well as objective post-HSCT risk factors that predict these long-term psychological distress reactions. We had 4 key findings. First, ∼20% of mothers of children who survived HSCT received a diagnosis of MDD, GAD, PD, or PTSD. Nearly half of the mothers evidenced significant posttraumatic stress reactions such as intrusive worries about their child who underwent HSCT and reminders of the transplant evoking significant stress reactions. Second, there was evidence of comorbidity among the 4 disorders, particularly with PTSD. Third, we were able to identify relatively few predictors of later psychiatric problems. Only maternal age and emotional distress at the time of HSCT were associated with later problems. Fourth, mothers with 1 of the 4 psychiatric disorders evidenced a lower quality of life.
Because no longitudinal studies have examined the prevalence of severe psychological distress reactions among mothers of pediatric HSCT survivors, we cannot directly compare the results of the present study with other studies. In terms of PTSD, we can compare the present results with our previous study of maternal PTSD assessed 6 months after HSCT,18 with studies of mothers of children who have survived other types of treatment for childhood cancer54 or organ transplantation20 as well with general population prevalence figures. Our figures for PTSD from the present sample are slightly higher than figures published in our previous study of this sample of mothers, which reported that 7.8% of mothers met PTSD criteria 6 months after transplantation,18 and slightly higher than figures reported in a long-term follow-up study of mothers of pediatric cancer survivors.54,55 However, our PTSD figures are much lower than the 25% figure reported by Pelcovitz et al55 and lower than the 27.1% figure recently reported by Young et al20 in their study of parents of pediatric organ transplant survivors. Our 12% PTSD figure is slightly higher than the 9.2% prevalence figure reported after exposure to trauma in general population studies.56 It is possible that our figures are lower than other studies of mothers of transplant survivors because we assessed mothers at a later time after transplantation. Because we studied mothers at a later time point after HSCT, the children in the present study survived the most risky phase after HSCT and their chances of long-term survival were relatively high. It is also possible that PTSD was lower in our study because we used a structured clinical interview rather than a paper-and-pencil measure, which typically results in lower rates of diagnosis. Although the rates of mothers who met full PTSD criteria were not extremely high, subthreshold PTSD was relatively high. Criterion B was particularly common among mothers, with 33 of the 94 mothers who did not have full PTSD meeting this criterion. Because subthreshold PTSD has been associated with social and occupational impairment comparable to full PTSD,57 this is an important phenomenon for clinicians to attend to in their assessments.
In terms of anxiety and depressive disorders, comparisons of studies of mothers at the time of transplantation are hampered by differences in assessment methods and procedures. Barrera et al12 did not use an interview measure, and our previous report from this population interviewed only a subset of mothers.18 Thus, we cannot compare prevalence of anxiety and depressive disorders at the time of transplantation with long-term prevalence. If we use BDI scores reported in other studies as an indicator of prevalence, then the prevalence of MDD at 18 months (10.5%) was similar to the 8% reported at the time of HSCT by Barrera et al12 but lower than the 15% figure that we reported at the time of HSCT in our previous study.18 Overall, our findings suggest that the prevalence of maternal depressive disorders decline over the 18 months after transplantation. If we use BAI scores reported in our previous study to compare prevalence, then the prevalence of severe distress reactions is lower than the 17% figure that we reported at the time of HSCT in our previous study.18 However, the prevalence rates reported in the present study may be an underestimation of actual prevalence. Mothers who dropped out of our study were significantly more depressed and anxious at the time of the transplantation. It is likely that our prevalence estimates are lower than actual prevalence in this sample if fewer mothers dropped out of our study.
The reported comorbidity of anxiety, depressive, and posttraumatic stress disorders was relatively high (∼35%), as evidenced by overlap in SCID-NP diagnoses. These findings are consistent with the literature. For example, Roy-Byrne et al58 found that 35.8% of primary care patients who received a diagnosis of PD also received a diagnosis of GAD within the past year, and 33.3% of the PD patients also received a diagnosis of MDD. An interesting finding from the present study regarded the overlap of anxiety and depressive disorders with PTSD. Comorbidity between PTSD and other psychiatric problems, particularly MDD, has been well documented in the literature.59–61 A number of reasons for this overlap have been proposed: 1) anxiety and depressive disorders could also be reactions to the traumatic event, 2) the burden of having PTSD could cause other psychological problems, and 3) the diffuse nature of the PTSD symptom constellation results in overlap in the symptoms that compose these diagnoses with other anxiety and mood disorders.62 Only the last hypothesis has been disproved.63
We found that 2 factors were predictive of later psychiatric disorders among mothers: maternal age and anxiety and depressive symptoms at the time of HSCT. Other studies that examined correlates of psychiatric disorders among parents of children who underwent HSCT12,64 or parents of children who survived cancer treatment have not reported that younger mothers evidence more distress.65 Studies that examined PTSD have reported inconsistent findings with regard to age. Studies of veterans have found that a younger age at trauma was a risk factor for PTSD,66 whereas studies of the general population have not found age at the time of the trauma to predict PTSD.67 One potential explanation for our findings is that younger mothers have less experience coping with difficult life events and thus may be more likely to have severe, persistent distress reactions.
The predictive association between distress at the time of transplantation and later psychiatric problems is consistent with the general psychiatric literature suggesting that previous psychiatric history is associated with later disorders.68,69 There are 2 potential explanations for the present findings. First, severe distress responses could reflect preexisting psychiatric disorders present before the transplantation. Second, emotional distress at the time of the traumatic experience heightens the mother’s attention to the negative and aversive aspects of the child’s HSCT and thereby raises the salience of the traumatic experience and the likelihood of persistent distress reactions. We did not formally assess lifetime psychiatric history in our SCID interview, and we did not assess attention given to negative and aversive aspects of the HSCT experience.
An interesting finding of this study was that objective transplant-related indicators of the child’s health did not predict long-term distress reactions. In this study, we used a relatively large set of potential indicators of the post-HSCT medical course and found that only 1 indicator of transplant severity or medical complications, the number of days hospitalized after HSCT, predicted long-term psychiatric diagnoses. However, the direction of this relationship was contrary to what would was predicted; that is, mothers of children with longer hospitalizations were less likely to received a diagnosis of a psychiatric disorder. In addition, when other factors were entered into the analyses, the contribution of this factor was no longer significant. Overall, these findings are not surprising given that objective indicators of child health have rarely been associated with parent distress.20
The association between impairment in mental health functioning and in social and occupational role functioning associated with having 1 of the 4 diagnoses is not surprising, as the disruption in quality of life has been reported in studies of people with psychiatric disorders.70 However, these findings point to the pervasive negative effects of these disorders on mothers’ quality of life and suggest that clinical intervention to address psychological problems will likely result in an improvement in mothers’ overall functioning.
Before discussing clinical implications, limitations of the present study and directions for future research should be discussed. First, anxiety and depression could reflect preexisting, long-standing psychological disorders that were not caused or exacerbated by the HSCT or the child’s medical diagnosis. Evaluating lifetime disorders or evaluating mothers before HSCT would partially clarify this issue, but it would be practically difficult to conduct pre-HSCT evaluations because many families are not accessible until the immediate time period before the hospitalization. Even a pre-HSCT evaluation would not capture psychiatric issues present before the child’s diagnosis. Second, the present sample was homogeneous in terms of ethnicity and primary language. Almost 90% of the mothers were white, and all mothers were English speaking. Thus, interpretation of the results to other samples of mothers of pediatric HSCT recipients, particularly Spanish-speaking mothers, should be made cautiously. A third limitation, mentioned previously, is sample selection bias. Mothers who dropped out of our study were more likely to be nonwhite and less likely to have received psychiatric care in the past. In addition, the most common reason for refusal was that the mother stated that she was overwhelmed. These differences may indicate that our prevalence estimates are lower than actual prevalence in this sample and suggest that mothers who are most in need of services may not participate in psychosocial research. Future studies should use measures that have been translated into Spanish and make special attempts to recruit and retain nonwhite mothers. Fourth, our sample included children with a wide age range. Fourth, although child age was not associated with maternal emotional distress, it is possible that developmental status of the child could have an impact on the mother’s experiences. Future studies should consider other ways of assessing developmental stage in addition to child age. Fifth, we did not include assessments of fathers or partners. Future studies should consider including fathers as well as mothers. Sixth, we did not assess child physical and emotional functioning, as reported by the child. As mother and child functioning are likely closely intertwined, future studies should include child-report ratings. Finally, we did not have psychiatric interview data on MDD, PD, and GAD prevalence for all mothers at baseline. Future studies should include similar methods (ie, interview methods) at all assessments.
Despite these limitations, these findings suggest that HSCT is a significant stressor for a subset of mothers. Approximately 20% of mothers have clinically significant distress reactions, and when subthreshold PTSD is included, slightly less than one third of mothers would meet this criterion for persistent distress. Mothers with 1 of the 4 psychiatric diagnoses reported a significantly lower quality of life, suggesting that the diagnoses had an impact on their daily functioning. Our study has several clinical implications. First, our findings suggest risk factors that clinicians can use to identify mothers who are most at risk for severe distress reactions: younger mothers and mothers who report high levels of anxiety and depressive symptoms at the time of HSCT. Second, our results suggest that it may prove beneficial to evaluate mothers at the time of HSCT and that to target interventions to mothers who are most at risk will reduce the likelihood that they will develop severe, persistent psychological distress reactions in the long-term. Such intervention programs might benefit from targeting mothers who, at the time of transplantation, evidence high levels of anxiety or depressive symptoms. Third, our results suggest that the content of psychological interventions should target immediate reductions in anxiety and depressive symptoms.
This work was supported by grant MH 57738 awarded by the National Cancer Institute and the National Institutes of Mental Health.
We thank Dr Jennifer Edwards for providing SCID-NP reliability ratings and training interviewers. We also thank Jean Greiff, Ellen Kolton, Alyssa Lowther, Anne McDevitt, Erin Olivo, Jennifer Soriano, Jennifer Morone, Chris Martinez, Tina Nikolopoulos, Daniela San Martin, Lisa Wu, and Nina Babat, PsyD, for assistance in collection of study data and Julian Silva and Suzanne Vickberg for data management. We appreciate the assistance of the oncologists on the transplant services, including Farid Boulad, Michael Amylon, and Morris Kletzel. Finally, we extend our appreciation to the mothers who participated in this study.
- Received August 4, 2003.
- Accepted October 15, 2003.
- Reprint requests to (S.M.) Fox Chase Cancer Center, 333 Cottman Ave, Philadelphia, PA 19111. E-mail:
- ↵Guinan E, Krance R, Lehmann L. Stem cell transplantation in pediatric oncology. In: Pizzo PA, Poplack DG (eds). Principles and Practice of Pediatric Oncology. Philadelphia, PA: Lippincott Williams & Wilkins; 2002:429–451
- ↵da Fonseca M. Long term oral and craniofacial complications following pediatric bone marrow transplantation. J Pediatr Dent.2000;22 :57– 62
- ↵Baker K, DeFor T, Burns L, Ramsay N, Neglia J, Robinson L. New malignancies after blood or marrow stem-cell transplantation in children and adults: incidence and risk factors. J Clin Oncol.2003;21 :1352– 1358
- ↵Nelson AW, Miles MS, Belyea MJ. Coping and support effects on mothers’ stress responses to their child’s hematopoietic stem cell transplantation. J Pediatr Oncol Nurs.1997;14 :202– 212
- ↵Streisand R, Rodrigue JR, Houck C, et al. Brief report: parents of children undergoing bone marrow transplantation: documenting stress and piloting a psychological intervention program. J Pediatr Psychol.2000;25 :331– 337
- ↵Manne S, Nereo N, DuHamel K, et al. Anxiety and depression in mothers of children undergoing bone marrow transplant: symptom prevalence and use of the Beck Depression Inventory and the Beck Anxiety Inventory as screening tools. J Consult Clin Psychol.2001;69 :1037– 1047
- ↵Manne S, DuHamel K, Nereo N, et al. Predictors of PTSD in mothers of children undergoing bone marrow transplantation: the role of cognitive and social processes. J Pediatr Psychol.2002;27 :607– 617
- ↵Diagnostic and Statistical Manual of Mental Disorders. 4th ed. Text revision. Washington, DC: American Psychiatric Association; 2000
- ↵Young GS, Mintzer L, Seacord D, et al. Symptoms of posttraumatic stress disorder in parents of transplant recipients: incidence, severity and related factors. Pediatrics.2000;111 :725– 731
- ↵Ostroff J, Woolverton K, Moadel A, et al. Witnessing trauma: patterns of traumatic stress responses among mothers of pediatric bone marrow transplant patients. Ann Behav Med.1997;19 :S192 (abstr)
- ↵Lang P. The cognitive psychophysiology of emotion: fear and anxiety. In: Tuma A, Maser J (eds). Anxiety and the Anxiety Disorders. Hillsdale, NJ: Erlbaum; 1985:131–170
- ↵Tait R, Silver R. Coming to terms with major negative life events. In: Uleman J, Bargh J (eds). Unintended Thought. New York, NY: Guilford Press; 1992:315–382
- ↵McFarlane A, Yehuda R. Resilience, vulnerability and the course of posttraumatic reactions. In: van der Kolk B, McFarlane A, Weisaeth L (eds). Traumatic Stress. New York, NY: Guilford Press; 1996:155–181
- ↵March J. What constitutes a stressor? The criterion A issue. In: Davidson J, Foa E (eds). Posttraumatic Stress Disorder: DSMIV and Beyond. Washington, DC: American Psychiatric Press; 1993:37–54
- ↵Spitzer RL, Williams JBW, Gibbon M, et al. User’s Guide for the Structured Clinical Interview for DSM-III-R SCID. Washington, DC: American Psychiatric Press; 1990
- ↵Beck A, Steer R. Beck Depression Inventory. San Antonio, TX: The Psychological Corporation; 1987
- ↵Horowitz M, Wilner N, Alvarez W. Impact of event scale: a measure of subjective stress. Psychosom Med.1979;41 :209– 218
- ↵Briere J. Psychological Assessment of Adult Posttraumatic States. Washington, DC: American Psychological Association; 1997
- ↵Corcoran L, Fisher J. Measures for Clinical Practice for Adults: A Sourcebook. New York, NY: Free Press; 1994
- ↵Sundin E, Horowitz M. Horowitz’s Impact of Event Scale: evaluation of 20 years of use. Psychosom Med.2003;65 :870– 876
- ↵Gibbon M, Williams RL, Janet BW. User’s Guide for the Structured Clinical Interview for DSM-IV Axis I Disorders-Research Version. New York, NY: Biometrics Research; 1996
- ↵Fleiss JL. Statistical Methods for Rates and Proportions. New York, NY: Wiley; 1981
- ↵Ware JE Jr, Snow KK, Kosinski M, et al. SF-36 Health Survey: Manual and Interpretation Guide. Boston, MA: The Health Institute, New England Medical Center; 1993
- ↵Ware JE Jr, Kosinski M. SF-36 Physical & Mental Health Summary Scales: A Manual for Users of Version 2. 2nd ed. Lincoln, RI: QualityMetric Inc; 2001
- ↵Bearman SI, Appelbaum FR, Buckner CD, et al. Regimen-related toxicity in patients undergoing bone marrow transplantation. J Clin Oncol.1988;6 :1562– 1568
- ↵Manne S, DuHamel KN, Gallelli K, et al. Post-traumatic stress disorder among mothers of pediatric cancer survivors: diagnosis, comorbidity, and utility of the PTSD checklist as a screening instrument. J Pediatr Psychol.1998;23 :6:357– 366
- ↵Bromet E, Sonnega A, Kessler R. Risk factors for DSM III-R posttraumatic stress disorder: findings from the National Comorbidity Survey. Am J Epidemiol.1998;147 :353– 361
- Copyright © 2004 by the American Academy of Pediatrics