Health-Related Quality of Life in Children With Congenital Adrenal Hyperplasia
Health-Related Quality of Life in Children With Congenital Adrenal Hyperplasia
Health-Related Quality of Life in Children With Congenital Adrenal Hyperplasia
Karen Lin-Su b
a
Department of Pediatric Endocrinology, Icahn School of Medicine at Mount Sinai, and b Department of Pediatric
Endocrinology, New York Presbyterian Hospital, Weill Cornell Medical Center, New York, N.Y., USA; c Department of
E-Mail [email protected]
1 Gustave Levy Place, Box 1198, New York, NY 1198 (USA)
www.karger.com/hrp
E-Mail mabelyaumd @ gmail.com
for CAH, such as dexamethasone, may negatively im- Subjects and Methods
pact HRQOL [2], while others report impaired health
Children with CAH and hypothyroidism between the ages of 8
status and working ability among adults with the classic and 18 years who were evaluated at our university-affiliated pedi-
form of the disease [2]. Among women with CAH, con- atric endocrinology clinic were recruited. Recruitment for CAH
cerns regarding fertility, increased anxiety about dis- was also performed through the CARES (Congenital Adrenal Hy-
closing the diagnosis to others, and decreased satisfac- perplasia Research Education and Support) Foundation, an inter-
tion with sexual functioning have been reported [3, 4]. national nonprofit patient support organization. Exclusion criteria
were an inability to read English or independently complete the
However, other data also suggest that patients function survey, additional chronic illness, or another acute disease that
normally and have a good overall psychological adjust- may impact the subjects’ clinical conditions. Written parental con-
ment [5]. Most recently, a study of 588 CAH adults in sent and child assent were obtained prior to participation in the
Sweden showed that the total cohort of CAH patients study.
did not differ greatly from the general population in HRQOL was assessed using the PedsQL 4.0 generic core scales,
which is a validated pediatric HRQOL inventory [9]. Information
terms of education, employment, income, sick leave, on demographics and medical management was obtained by an
disability pension, social welfare, marriage, or children additional intake form that was administered at the same time as
[6]. However, a subanalysis revealed that outcomes were the PedsQL. The PedsQL 4.0 includes an age group-specific self-
worse among women and those affected with the classic report for the patient as well as a parent proxy report. It consists of
form of the disease [7]. 23 questions that are subdivided into 4 domains: physical, emo-
tional, social, and school functioning; each domain has specific
Pediatric outcome data on HRQOL in CAH are sparse. itemized questions. The physical domain has 8 items: difficulty
Adults have identified experiences during their child- with walking, running, sports or exercise, lifting, bathing, house
hood that have negatively impacted them, including the chores, feeling pain, and having low energy. The emotional do-
processes involved in understanding or becoming aware main has 5 items: fear, sadness, anger, insomnia, and worrying
of CAH, hospitalizations for adrenal crises, and surgical about the future. The social domain contains 5 items, which in-
clude having trouble getting along, feeling unwanted, feeling
interventions to correct genitourinary conditions in fe- teased, not being able to do things others their age can do, and
males [4]. In contrast, a recent study of children with keeping up with their peers. The school domain has 5 items: focus,
CAH in a Dutch population showed few negative effects memory, keeping up with school/work load, absences due to
of CAH on physical, social, and societal functioning based health status, and absences due to doctor appointments and hos-
on a self-designed questionnaire [6]. Clearly, more re- pital visits.
The scoring for each question is on a 5-point scale: from 0 =
search is needed to identify the aspects of a child’s QOL ‘never a problem’ to 4 = ‘almost always a problem’. Each score of
that may be impaired. 0–4 was converted to a 0–100 metric scale and then averaged with-
The objective of this study was to evaluate changes in in its categories as well as individually analyzed. Higher scores rep-
HRQOL in children affected by CAH. Beyond concerns resent better HRQOL. Categorical scores were only averaged if
related to chronic illness, children with CAH may face ≥50% of the items in the scale were answered.
unique challenges related to possible sexual precocity, Statistical Analysis
hyperandrogenism, or genital ambiguity. To under- Descriptive statistics were used to calculate the mean and stan-
stand if CAH poses an additional burden on QOL com- dard deviation (SD) scores of the child self-report and the parent
pared to children with other chronic endocrine disor- proxy report for each PedsQL domain. A subgroup analysis was
ders, we also tested children affected with hypothyroid- performed for various types of CAH.
The mean scores for the general pediatric population for each
ism [8]. Similar to CAH, the treatment of hypothyroidism functional domain in the PedsQL 4.0 questionnaire were pub-
requires the administration of daily medication and lished by Varni and colleagues [10–12] and were compared to the
continuous medical assessment through childhood, mean scores of children with CAH. The minimal clinically im-
with frequent laboratory testing and medical visits. Hy- portance difference (MCID) was determined for each domain
pothyroidism can be congenital or acquired and, thus, using the standard error of measurement (SEM), as previously
described [10]. MCID is defined as the ‘smallest difference in a
has a bimodal age at presentation similar to CAH, which score of a domain of interest that patients perceive to be benefi-
can present either in the newborn period as classic dis- cial and that would mandate, in the absence of troublesome side
ease or later on in life as nonclassic disease. Contrary to effects and excessive costs, a change in the patient’s management’
CAH, however, hypothyroidism does not carry the [12].
physical stigmata that may be associated with preco- Impaired HRQOL or ‘at risk’ status was defined as a PedsQL
score of at least 1 SD below the general pediatric population mean,
cious puberty or hyperandrogenism, and its study may as previously described [10–12]. The percentage of CAH subjects
highlight HRQOL aspects uniquely affected in children with scores below the cutoff point for the ‘at risk’ status was calcu-
with CAH. lated for each domain score.
198.143.32.65 - 1/26/2016 7:36:14 PM
CAH Hypothyroidism
(n = 33) (n = 14)
The two-sided independent t test was used to compare the visits was the same between subjects with classic and non-
PedsQL domain scores between the two groups, i.e. CAH versus classic CAH, at an average of 3.7 visits per year. There was
hypothyroid subjects and subcategories within CAH based on the an equal distribution of males and females. Of the subjects
classification of CAH (classic vs. nonclassic). The concordance be-
tween child self-report and parent self-report was determined with hypothyroidism, 50% had acquired disease and 50%
through Pearson’s correlation coefficients. p values <0.05 were congenital hypothyroidism. As a group, they were older
considered statistically significant. Statistical analyses were per- than the CAH patients at diagnosis and primarily female.
formed using SAS and Microsoft Excel. Visits were less frequent compared to the CAH subjects.
DOI: 10.1159/000435855
Downloaded by:
Table 2. HRQOL results for children with CAH using the PedsQL 4.0 generic core scales
Child report
Total 82.9 ± 13.2 78.1 ± 18.9 – 4.8 vs. 4.36* 23.3**, †
Physical 86.9 ± 13.9 86.8 ± 18.2 – 0.1 vs. 6.66 23.3**, †
Emotional 78.2 ± 18.6 78.6 ± 20.3 0.4 vs. 8.94 23.3**, †
Social 84.0 ± 17.4 84.8 ± 23.4 – 0.8 vs. 8.36 23.3**, †
School 79.9 ± 16.9 74.8 ± 23.8 – 5.1 vs. 9.12 33.3**
Parent report
Total 81.3 ± 15.9 73.1 ± 20.4 – 8.2 vs. 4.5* 26.7**
Physical 83.3 ± 20.0 80.0 ± 21.3 – 3.3 vs. 6.92 16.7
Emotional 80.3 ± 17.0 66.7 ± 21.4 –16.6 vs. 7.79* 33.3**, †
Social 82.2 ± 20.1 74.6 ± 25.6 – 7.6 vs. 8.98 20*
School 76.9 ± 20.2 67.3 ± 25.5 – 9.6 vs. 9.67 33.3**, †
Scores are presented as the mean ± SD. Population mean scores were obtained from Varni and colleagues [10
– 12].
* The difference between the CAH mean and the healthy population mean is >1 SEM and constitutes an
MCID. Abnormal results are set in bold.
** Subjects with scores 1 SD below the population mean are considered to have impaired HRQOL or to be at
risk. The percentage of subjects whose scores fell below the cutoff point for each domain are reported; those
greater than expected in the general population are set in bold.
† Although percentages are the same in these domains, the individual subjects within each domain are different.
80
dren with CAH and their parents compared to the healthy
population are shown in figure 1.
The ‘at risk’ status for impaired HRQOL was also de-
60
termined by calculating the percentage of children with
CAH whose scores fell below 1 SD of the general pediatric
40 mean. The expected percentage of a population that is
Children with CAH found 1 SD below the mean is about 16%. The percentage
Children from the general population of children with CAH who reported scores in the im-
20
Parents of children with CAH
Parents from the general population
paired or ‘at risk’ state ranged from 23.3 to 33.3%, which
0
are all greater than the 16% that is expected for the gen-
Total Physical Emotional Social School eral population (table 2, last column). These results sug-
gest that children with CAH perceive themselves as hav-
ing impaired HRQOL more frequently than unaffected
Fig. 1. HRQOL results for children with CAH using the PedsQL children. Parents similarly reported greater percentages
4.0 generic core scales compared to the healthy population. For of CAH children with scores in the ‘at risk’ range in all
each domain, the mean scores with SD reported by children with
CAH and their parents are compared to scores reported by healthy domains except for the physical domain (table 2).
children and their parents. The population mean scores were ob- On item analysis, both parents and subjects with CAH
tained from Varni and colleagues [10–12]. more frequently reported absences from school due to
198.143.32.65 - 1/26/2016 7:36:14 PM
DOI: 10.1159/000435855
Downloaded by:
In terms of physical function (i.e. sports performance, en- the MCID has been defined as the smallest difference in
ergy level, feeling pain, etc.), parents felt that the CAH a score of a domain of interest that patients perceive ben-
children were normal. In social functioning, both CAH eficial and that would mandate a change in patient man-
subjects and their parents more frequently reported other agement. Using this index, children with CAH and their
children not wanting to be friends. parents report an overall decrease in QOL, while parents
In children with type 1 and 2 diabetes mellitus studied also report lower functioning in the emotional domain.
by Varni et al. [13] using the same PedsQL 4.0 generic core In this study, we also compared the HRQOL outcomes
scales, there was a significant difference between healthy of children with hypothyroidism to those of children with
children and children with diabetes for all scales except in CAH using the same PedsQL scales. We detected lower
the case of physical functioning and social functioning scores in school performance by self-report in children
based on self-report. The difference between the scale with CAH compared to those with hypothyroidism. Again,
scores of the 191 subjects with type 1 diabetes mellitus and the increased number of school absences in CAH subjects
the general population mean scores were not greater than may be behind these findings, as subjects with CAH and
the MCID [10–13]. There was a significant difference be- their parents reported missing school for both illness and
tween healthy children and children with diabetes in all doctor’s appointments more frequently that children with
scales based on parent proxy report. Based on parent proxy hypothyroidism. No other differences were documented
report, the differences between scale scores of children with between CAH and hypothyroid children, though we ob-
type 1 diabetes and the general population mean scores served a general trend towards lower scores in CAH pa-
were greater than the MCID for total QOL and the emo- tients. It is possible that we were unable to detect small
tional domain [10–13]. These results in subjects with type differences between groups because of the small sample
1 diabetes are similar to those found in our subjects with size, especially among the hypothyroid participants.
CAH, suggesting a similar chronic disease burden on QOL. Previous studies have noted a difference in HRQOL in
Previous research on HRQOL in CAH patients has pri- adult females; however, our small sample size and high
marily focused on adults [2–7, 14], whereas pediatric data percentage of female subjects did not allow for sufficient
have been limited. A recent study from the Netherlands power to detect differences based on gender. Another
documented few negative effects of the disease on physical, limitation of our study is the use of a generalized ques-
social, and societal functioning [8]. However, the study tionnaire rather than one targeted for children with CAH.
used self-designed questionnaires, which hinders compar- The questionnaire also lacked open-ended questions that
isons to other studies. Another study that included both may have been able to provide more qualitative data.
children and adults with CAH reported normal psycho-
logical adjustment [7]. Differences in methodology may
explain some of these differences in outcomes. Finally, the Conclusion
PedsQL questionnaire was used in an Australian study of
54 children with various disorders of sex development, in- Children with CAH have an overall decrease in
cluding 16 girls with CAH [15]. In this mixed cohort, a HRQOL compared to healthy children as well as lower
decrease in psychosocial aspects of HRQOL was observed, HRQOL compared to hypothyroid children in school
while physical function was close to normal [15]. functioning. Future studies employing more descriptive
There is no HRQOL scale designed and validated spe- data may aid in our understanding of social functioning
cifically for children with CAH. In this study, we used the in children with CAH and allow for more specific thera-
generic core scales of the PedsQL 4.0 measurement mod- peutic measures to be developed. Additional investiga-
el [10–12]. The PedsQL 4.0 is a validated generic HRQOL tion may also provide insight into issues unique to chil-
measurement instrument developed to examine a broad dren growing up with CAH, thus guiding the develop-
age range. It has been used extensively in numerous clin- ment of behavioral approaches to assist families and
ical and research settings that evaluate various pediatric further improve QOL.
chronic health disorders. The PedsQL 4.0 generic core
scales distinguish between healthy children and children
with chronic health conditions, demonstrating sensitivity Acknowledgement
for disease severity and responsiveness. Results have been
The authors would like to acknowledge the CARES Foundation
shown to predict prospective pediatric health care utiliza-
for their help with patient recruitment.
tion and to impact clinical decision-making. Specifically,
198.143.32.65 - 1/26/2016 7:36:14 PM
DOI: 10.1159/000435855
Downloaded by: