Reduced Time in Bed and Obstructive Sleep-Disordered Breathing in Children Are Associated With Cognitive Impairment
OBJECTIVE. The purpose of this study was to determine if reduced time in bed as well as the degree of obstructive sleep-disordered breathing predicted the risk of impaired cognitive function in children with adenotonsillar hypertrophy suspected of having obstructive sleep-disordered breathing.
DESIGN. We studied 56 children, aged 6 to 12 years, with adenotonsillar hypertrophy referred for suspected obstructive sleep-disordered breathing. Children were given a sleep diary and underwent wrist actigraphy for 6 consecutive days and nights. On day 7, the children were given general cognitive tests, memory tests, and continuous performance tests followed by attended polysomnography that night. Parents completed snoring and behavior questionnaires.
RESULTS. Shorter mean time in bed for 6 nights and a history of nightly snoring were highly predictive of lower scores for the vocabulary and similarities cognitive function tests. Children who had a mean time in bed of 557 minutes and did not snore nightly were predicted to have vocabulary and similarities scores more than 1 standard deviation higher than children who had a mean time in bed of 521 minutes and snored nightly. Shorter mean time in bed and the log of the apnea hypopnea index also predicted lower vocabulary and similarities scores. Greater night to night variability in time in bed was significantly predictive of lower vocabulary and similarities scores, but variability was not as predictive as mean time in bed. Neither mean time in bed nor the coefficient of variation of time in bed predicted other cognitive or behavioral scores.
CONCLUSIONS. Short or variable time in bed and nightly snoring or higher apnea hypopnea index predicted impaired vocabulary and similarities scores in children with adenotonsillar hypertrophy suspected of having obstructive sleep-disordered breathing. The degree of cognitive impairment attributable to short time in bed and obstructive sleep-disordered breathing is clinically very significant.
- time in bed
- sleep restriction
- sleep disordered breathing
- sleep apnea
- adenotonsillar hypertrophy
- cognitive function
- cognitive impairment
Children with habitual snoring or obstructive sleep-disordered breathing (OSDB) are reported to have impaired academic and cognitive performance.1–11 Reduced time in bed (TIB) has also been shown to impair cognitive and academic performance.12–14 Recent studies have indicated that reduced TIB is a common occurrence in children.15–17 Studies on the effect of reduced TIB have been conducted in normal children but, to our knowledge, have not been performed in children with OSDB.
The purpose of this study was to determine whether cognitive performance in children with adenotonsillar hypertrophy suspected of having OSDB would be related to time they spend in bed at night, as well as the degree of OSDB. We postulated that children with short TIB and those with more OSDB would have more impaired cognitive performance. We studied children with adenotonsillar hypertrophy suspected of having OSDB with both sleep diaries and wrist actigraphy for 6 nights before cognitive testing.
These studies were approved by the University of Virginia Human Investigation Committee. All of the participants were volunteers who signed informed assent or consent and whose parents signed informed consent for the study. We studied children 6 to 12 years old. These children were consecutive participants in a National Institutes of Health-sponsored study on the effect of OSDB because of adenotonsillar hypertrophy on behavior, cognitive performance, and growth. Because subjects in this larger study had to hold still for 5 minutes in a magnetic resonance scanner, subjects <6 years old were not included. All of the children with adenotonsillar hypertrophy who were thought to have OSDB and were seen in the authors' clinics at the University of Virginia and at Pediatrics Associates of Charlottesville (Charlottesville, VA) were offered participation in this study. Children were excluded if they had a pulmonary, cardiac, neurologic, or auditory disorder or if they were on any medications including those for attention-deficit/hyperactivity disorder. There was no BMI exclusion criteria.
Actigraphy monitoring was initiated on July 7, 2000, after the larger study had started. Actigraphy monitoring was planned as part of the larger study but could not be initiated at the start of the larger study because of logistic and technical problems. Subjects in this study are, thus, a subset of the larger study. This subset includes all of the subjects in the larger study enrolled after the above date. Cognitive findings in the larger study set have been published7 but did not include actigraphy or sleep diary data, which were only collected in this smaller subset. For 6 nights, subjects kept a sleep diary18 and were monitored with actigraphy using a Mini Motion Logger Actigraph (Ambulatory Monitoring Inc, Ardsley, NY) worn on the nondominant wrist. The Actigraph operated in “0 crossing mode” and recorded the number of movements in 1-minute intervals. Actigraphy studies started Tuesdays through Fridays throughout the year and occurred when some children were in school and others were not in school.
Actigraph and sleep diary data were analyzed to determine TIB. TIB was defined as the time from when the child first went to bed as recorded on their sleep diary and as indicated by a decrease in wrist activity until they got out of bed as recorded in their diary and as indicated by an increase in wrist activity level. Discrepancies between diaries and Actigraph recordings were resolved by contacting parents and or by consensus review among the investigators who did not have knowledge of cognitive scores or other results. Subjects were excluded if they did not wear the Actigraph for all 6 nights or if there were irreconcilable differences between their sleep diary and their Actigraph recordings. TIB was averaged for the 6 nights before cognitive function studies (TIB1–6).
After wearing the Actigraph for 6 consecutive nights, subjects were admitted to the University of Virginia General Clinical Research Center. At ∼1:30 pm, children had cognitive function studies. These consisted of 3 subtests of the Wechsler Intelligence Scale for Children, Third Edition,19 vocabulary, similarities, and block design. The vocabulary subtest is a measure of word knowledge and verbal expression. The child is asked to define specific words presented in order of increasing difficulty. In the similarities subtest, the child is presented with pairs of words and asked to explain the basic similarity/relationship of the 2 items. Thus, it is a measure of verbal conceptualization, which provides unique measures of verbal reasoning and concept formation. The block design subtest consists of items of increasing difficulty in which the child is shown a picture of an abstract design then asked to replicate the design using blocks that have either solid red, solid white, or diagonally split red and white sides. The subtest requires the abilities to perceive and analyze a form, break it down into components, then assemble the block components into a replication of that form. It involves visual organization, visual motor coordination, nonverbal logic, and reasoning. A difference of 3 points on each of these tests represents 1 SD. Additional cognitive testing included the Wide Range Assessment of Memory and Learning test.20 We also administered the Connors Continuous Performance tests,21 a computer test designed to measure sustained attention. Testing was performed by 2 experienced neuropsychology technicians. The parent accompanying the child to the General Clinical Research Center completed the Connor's Parent Behavior Rating Scale long form,21 which is used to detect behavioral problems in children.
A sleep questionnaire was administered to all of the parents to solicit information about snoring and breathing during sleep.22 Parents were asked to describe the frequency of their children's snoring as: (1) never, (2) rarely (less than once month), (3) occasionally (1–4 times a month), (4) frequently (more than once a week), or (5) most nights.
In the evening, subjects had overnight polysomnography with a Sandman sleep system (Sandman Sleep Diagnostics, Kanata, Ontario, Canada) using conventional acquisition and analysis techniques as described previously.7,23 Briefly, this included electroencephalograms, electro-oculograms, submental electromyograms, nasal and oral airflow, pulse oximetry, and chest and abdominal movement. Apneas were characterized by reductions in flow to <20% of normal for ≥6 seconds or 2 respiratory cycles. Hypopneas were characterized by reductions in flow to <50% of normal for ≥2 respiratory cycles and were often accompanied by hypoxemia or arousal.23 Apneas and hypopneas were reported to be obstructive when the chest and abdomen moved and central when chest and abdominal movements were absent. We calculated the number of obstructive apneas and hypopneas per hour of sleep and reported it as the obstructive apnea hypopnea index (OAHI). Because criteria for arousals have not yet been developed for children,24 arousals were defined as recommended by the American Sleep Disorders Association Task Force report25 and included respiratory-related, technician-induced, and spontaneous arousals. They were expressed as total arousals per hour of sleep. All of the studies were scored using the above criteria by a single technician blinded to the results of actigraphy and cognitive tests. All of the studies were reviewed by Dr D'Andrea or Suratt, who were also blinded to results of the actigraphy and cognitive studies to insure the criteria were being followed. Problematic studies were reviewed by both Dr D'Andrea and Suratt. Because of errors in the commercial software calculations, oxyhemoglobin saturation (Sao2) values were extracted from the Sandman event files into ASCII files. Using a previously described program,26 we calculated the mean Sao2 and the lowest Sao2 values during sleep.
Statistical analysis was performed with S-Plus software (Insightful Corp, Seattle, WA). Multiple linear regression analysis was used to compare cognitive and behavioral tests to TIB, to sleep and breathing variables shown previously to be important by other studies and our previous study, to variables that we thought were plausibly related to outcomes, and to potential confounders. Sleep variables included the following: sleep latency, sleep efficiency, percentage of sleep in nonrapid eye movement and in rapid eye movement sleep, arousals, and sleep stage shifts per hour of sleep. Breathing variables included sleep questionnaire data, the OAHI, and minimum Sao2. Potential confounders included gender, age, race, health insurance status, and type of school attended (private versus public). Variables were included in the multiple regression analysis models if their coefficients were significant at the P ≤ .05 level. BMI was not included in the regression analysis, because it was thought to be less likely to independently influence outcomes in this population than other variables. Snoring frequency was analyzed as a dichotomous variable with a high snoring category (“1”) including all of the subjects whose parents described their snoring frequency as a 5 (snoring most nights) and a low snoring category (“0”) including all of the other subjects (snoring less than most nights). Race was also analyzed as a dichotomous variable with black children in 1 group (“0”) and “other” children consisting of white and Asian children in the other group (“1”). Comparisons between 2 groups were performed with the 2-sample Wilcoxon rank-sum test.
Actigraphy monitoring was initiated on July 7, 2000, after the larger study had started. Because of logistic problems in getting the device to children 6 days before their cognitive and polysomnography testing, it was not given to 7 subjects, 1 subject refused to wear it, the device malfunctioned in 11 subjects, and 8 subjects did not use it on all 6 nights, leaving 56 children who were monitored with it for 6 consecutive nights before cognitive function testing. Comparing the 56 subjects enrolled in the study with the subjects who were not enrolled for reasons listed above, there was no difference in age, gender, BMI z score, OAHI, Sao2 nadir, snore group, sleep efficiency, or cognitive function tests. There was, however, a higher proportion of black children enrolled in the study than not enrolled (enrolled subjects: 34% black; not enrolled subjects: 11% black; P = .0288). Comparing the 56 subjects included in this study with the 114 subjects in the larger previously published study, there were no significant differences in age, gender, race, BMI z score, OAHI, Sao2 nadir, snore group, sleep efficiency, or cognitive function tests. In the current study, 34% of children were black, and in the larger study, 30% were black.
Subject characteristics are shown in Table 1 and breathing and sleep variables in Fig 1. A typical Actigraph record is shown in Fig 2. Figure 3 (left) shows the raw TIB values for each of 6 consecutive nights. Figure 3 (right) shows the mean bed and rise times ±1 SD by night of the week.
Analysis Based on TIB for 6 Nights Before Cognitive Testing
The cognitive tests with the strongest relationships to TIB1–6 were the Wechsler subtests vocabulary, similarities, and, to a lesser extent, block design (Table 2). Figure 4 shows scatter plots of vocabulary and similarities scores versus TIB1–6. It illustrates evidence of an increasing trend in vocabulary and similarities scores as TIB1–6 increases.
With multiple regression analysis, cognitive tests with the strongest relationships to both TIB1–6 and to sleep and breathing variables were 2 of the Wechsler subtests, vocabulary and similarities, as shown in Table 3 and as described by the following well-fitting linear regression models: (1)(2) The term “snore group” is a dichotomous indicator of snoring, which equals 0 for subjects in the low-snoring group and 1 for subjects in the high-snoring group (a parental report of nightly snoring). Its negative coefficients (P = .0094 and .0069, respectively, for models 1 and 2) reflects the negative impact of high snoring levels on predicted scores. After accounting for the snoring terms, the term TIB1–6 has positive coefficients (P = .0055 and .0015, respectively, for models 1 and 2), providing an indication that, within a given snoring group, higher vocabulary and similarities scores should be predicted for children with longer TIB1–6.
Replacing snoring group with the log of the apnea hypopnea index (log10[OAHI + 1]) yielded similar results. We used a logarithmic transformation of OAHI to correct for the nonlinear distribution of OAHI values (Fig 1, left): (3)(4) The negative coefficients for log10(OAHI + 1; P = .012 and.067, respectively, for models 3 and 4) reflect the negative impact of log10(OAHI + 1) on predicted vocabulary and similarities scores. Because the P for log10(OAHI + 1) for model 4 does not quite reach significance, snoring group is superior to log10(OAHI + 1) as a predictor of similarities score.
In a previous report involving all 114 subjects in the large cohort,7 we found that a model using snoring group, race, and sleep efficiency (measured during polysomnography the night after cognitive testing) predicted vocabulary and similarities scores. When we applied that model to the current study (a smaller subset of the larger study), we found that that the general model for vocabulary was still applicable with an r2 of 23.1% (P = .0032). The P for snore group remains highly significant; however, the P for race (P = .056) and the P for sleep efficiency (P = .089) no longer reached significance, reflecting a potential type 2 error. Similar results were seen with the similarities score. When TIB1–6 was added into the model containing snoring group, race, and sleep efficiency, only the snore group and TIB1–6 were significant predictors of both vocabulary and similarities scores as described above.
Analysis Based on the Variability of TIB 6 Nights Before Cognitive Testing
The left panel of Fig 2 illustrates that there is considerable variability in the night-to-night TIB in many subjects. Therefore, we expressed the night-to-night variability of TIB in each subject as a coefficient of variation. There was a modest negative relationship between the coefficient of variation TIB1–6 and TIB1–6 (r2 = 11.5%; P = .011). Higher coefficients of variation TIB1–6 predicted shorter TIB1–6.
There were also modest significant negative linear relationships between the coefficient of variation of TIB1–6 and both the vocabulary score and similarities scores as seen in Fig 5. The best-fitting linear regression models predicting the relationship between vocabulary and similarities scores and TIB1–6 and sleep and breathing variables are as follows: (5)(6) The negative coefficient for coefficient of variation TIB1–6 (P = .013 and .015, respectively, for models 5 and 6) reflects its negative impact on vocabulary and similarities scores. For the vocabulary score, the log10(OAHI + 1) also had a negative impact on the predicted value (P = .0157) with higher log10(OAHI + 1) values predicting lower vocabulary scores. Adding the snore group instead of, or in place of, log10(OAHI + 1) did not add to the predictive power of the coefficient of TIB1–6 in predicting vocabulary or similarities scores.
Analysis Based on Weeknights and Weekends
The right panel of Fig 2 suggests that bed and rise times were earlier on Sunday through Thursday nights (weeknights) than on Friday and Saturday nights (weekends). This was confirmed by multiple comparisons testing as shown in Table 4. TIB also tended to be shorter on weeknights than on weekends (Table 4). Because of these differences, we analyzed the relationships between cognitive tests and TIBweeknight separately from TIBweekend. Significant results are shown in Tables 2 and 3.
TIBweeknight predicted vocabulary, similarities, and block design scores with correlation coefficients similar to that found for TIB1–6 (Tables 2 and 3). TIBweekend, however, did not predict any cognitive score. The only cognitive test predicted by bed or rise time was vocabulary; it was predicted only by bedtime on the weekend with later bedtimes predicting lower vocabulary scores (data not shown).
TIB1–6 did not significantly correlate with other cognitive or behavioral tests or with any breathing variable from the polysomnography study including OAHI and its log transformation. The coefficient of variation of TIB also did not correlate with other cognitive or behavioral test scores. Potential confounders, including age, gender, health insurance statues, and type of school attended, did not contribute to the predictions of any cognitive or behavioral test.
Although the median TIB1–6 was somewhat lower in the black group than in the other race group, the differences were not significant (531 vs 549; P = .11). The coefficient of variation TIBn1–6 was, however, significantly larger in the black group compared with the other race group (median: 0.13 and 0.083; P = .0081). Significance did not change when the 1 Asian child was removed from the other group. The coefficient of variation TIB1–6 was also higher in the high snoring group compared with the low snoring group (0.096 vs 0.066; P = .010). TIB1–6 correlated negatively with age (r2 = 15.91%; P < .0023).
To determine whether high sleep efficiencies and short sleep latencies that we observed previously during polysomnographic testing on night 7 were related to short TIB, we compared TIB to these factors. TIB1–6 did not correlate with sleep efficiency and sleep latency. Although the TIB for the 3 nights before polysomnography (nights 4–6) correlated with sleep efficiency and sleep latency (r2 = 13%, P = .0069 and r2 = 8.7%, P = .032, respectively), the significance was driven by 1 outlier, and when this outlier was removed, the relationships were no longer significant (r2 = 6.3%, P = .065 and r2 = 0.95%, P = .49, respectively).
In this study, children with adenotonsillar hypertrophy suspected of having OSDB were more likely to have impaired cognitive function if they had short or more variable TIB and if they snored every night (snore group) or had high OAHIs. The mean degree of impairment in these tests attributable to short TIB and OSDB is very significant and, as noted in our previous study,7 is similar to the effects of lead exposure in children.27,28 Model 1 estimates that children who do not snore every night and have a TIB1–6 of 557 minutes (boundary between third and fourth interquartile range of our data) would have a mean vocabulary score of 13.2 (95% confidence interval [CI]: 11.9–14.5). Children who do snore every night and have a TIB1–6 of 521 minutes (boundary between first and second interquartile range of our data) would have a mean vocabulary score of only 9.9 (CI: 8.9–10.9). The 3.3 mean standard score difference represents a reduction in vocabulary scores equal to 1.1 SDs of the normative score distribution for the general population. Similarly, model 3 estimates that children with a TIB1–6 of 557 minutes and an OAHI of 0 would have a mean vocabulary score of 13.3 (CI: 11.9–14.7), whereas children with a TIB1–6 of 521 minutes and an OAHI of 10 would have a mean vocabulary score of 9.7 (CI: 8.4–11.0), a reduction of 3.6 standard score points equal to 1.2 SDs. For the Wechsler Intelligence Scale tests, a difference of 1 SD in scores from the population mean is considered both statistically significant and clinically important.19 A subtest score change of 3 points (1 SD) is equivalent to a 15-point change on the full scale Wechsler intelligence quotient (IQ) scale. As we noted previously,7 studies of lead exposure in children have reported that children with a lifetime average blood lead concentration of ≤10 μg/dL had a 7.4-point reduction in IQ, a decline thought to be important.29
For the individual Wechsler subtests, the vocabulary subtest correlates most highly with the overall Wechsler IQ score and is viewed as the best single predictor of general cognitive functioning. The vocabulary subtest is also considered a strong predictor of academic success.30 The similarities subtest measures conceptual grouping and verbal abstract reasoning skills and is recognized as a measure of skills important in problem solving and new learning.30
The lower mean vocabulary and similarities scores noted above represent weaknesses that could put the subpopulation of children with OSBD and reduced TIB at a serious disadvantage in terms of scholastic performance.31 Both vocabulary and similarities are relatively stable over time, so that the deficits noted in these results are likely to be long-standing. Furthermore, the deficits are severe enough that the possibility of interaction with other deficits leading to a downward academic spiral must be considered likely.
The vocabulary and similarities subtests are relatively nonengaging tasks that require a combination of listening, sustained attention, extended cognitive effort without environmental stimulation, and internal manipulation of information. These factors stand in contrast to the other tests in our battery that include visual stimuli and manipulative materials as cuing components. The vocabulary subtest consists of presenting the child with a single word and requiring the formulation and expression of an acceptable definition. Similarities involves listening to pairs of words and then reporting their most essential relationship. Thus, it seems that children with OSDB in this study have particular difficulty when confronted with nonengaging complex tasks requiring listening, sustained attention, and internalized manipulation of information. These are all skills that are essential to academic success. The difficulty that children with OSDB have with these tasks may explain their poor performance at school.
Restricted TIB could lead to sleep deprivation and impaired ability to learn and perform. Fallone et al12 sleep-restricted 6- to 12-year-old children for 1 week and found that teacher rating questionnaires indicated that the children had developed academic and attention problems. Other studies have compared sleep habits to performance and found that short sleep times, erratic sleep/wake cycles, late bed and rise times, and poor sleep quality were associated with impaired academic performance.32,33 To our knowledge, the current study is 1 of the few to use measured rather that self-reported criteria for both TIB and performance.
Our study cannot directly establish causation. We suspect that disorganized families with increased socioeconomic and psychological stress may not ensure adequate TIB for their children, resulting in poor cognitive development. Alternatively, it is also possible that a process that causes reduced performance on cognitive tests could also cause abnormal sleep schedules with inadequate and variable TIB. We think the former is more likely. It is unlikely that short TIB in our population is because of hyperactivity in children with OSDB, because we found no relationship between TIB1–6 and hyperactivity scores.
More variable TIB also predicted lower cognitive scores. This seems to confirm that maintaining regular bed and wake times is important for children. Our observation that the coefficient of variability of TIB1–6 was greater in black than in other children was also reported by Spilsbury et al,15 who studied sleep behavior in 755 school-aged children using sleep journals and questionnaires.
On weeknights, Sunday night through Friday morning, children retired earlier and arose earlier than on weekends. Our data suggest that TIBweeknight was the major predictor of cognitive tests, because TIBweeknight predicted cognitive scores, whereas TIBweekend did not. Because weekend data reflect only 1 or 2 nights of recording (compared with 4 or 5 recorded weeknights), however, we cannot be sure that this restricted weekend data set has enough information to reliably exclude a weekend influence. In future studies, longer recording times collecting data over several weeks would improve our ability to confirm this observation. Our analysis did not take into account differences in schedules between school and vacation periods; these differences could potentially be a source of variability.
This study did not confirm our suspicion that high sleep efficiencies and short sleep latencies that predicted impaired cognitive performance in our previous study are related to short times in bed. The relationship between sleep efficiency and TIB approached significance, however, and it is possible that with a larger data set the relationship might be significant.
This article does not address sleep time calculated by actigraphy algorithms for the 6 nights before cognitive testing. Actigraphy algorithms for sleep are problematic in children with OSDB because of their frequent movements. Because of the complexity of this issue, it will be addressed in a separate article comparing these algorithms with polysomnography in children with OSDB. It will also address whether movements during sleep not processed by these algorithms relate to cognitive function and behavior.
Because only 4 of our subjects had severe sleep apnea with an OAHI >20, our study has no power to detect the relationship among TIB, OSDB, and cognitive tests in very severe OSDB. However, the OAHI range of our patients seems to be typical of OSDB in children. For example, Kaemingk et al3 reported an apnea hypopnea index >5 in only 51% of 6- to 12-year-old children with OSDB. O'Brien et al6 reported a mean apnea hypopnea index of 9.8 in 35 children with OSDB and a mean age of 6.7. These findings are similar to the mean OAHI of 6.7 in our subjects.
This study shows that short TIB and variable TIB predicted poor cognitive performance in children with adenotonsillar hypertrophy thought to have OSDB. Children who snore every night or those with high OAHIs were also predicted to have even lower cognitive scores. Although the cause of the relationship between short TIB and poor cognitive performance is not well understood, parents of children who snore or have OSDB should be encouraged to insure that their children have adequate TIB.
This work was supported by National Heart, Lung and Blood Institute grant HL-62401 (to Dr Suratt) and University of Virginia General Clinical Research Center grant M01 RR-00847.
We thank Tracy Puffenbarger for assistance with the polysomnography studies and Nancy Mays and Pam McArdle for performing the neuropsychological tests.
- Accepted October 1, 2006.
- Address correspondence to Paul M. Suratt, MD, Box 800546, University of Virginia Medical Center, Charlottesville, VA 22908. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
- ↵Kaemingk KL, Pasvogel AE, Goodwin JL, et al. Learning in children and sleep disordered breathing: findings of the Tucson Children's Assessment of Sleep Apnea (tuCASA) prospective cohort study. J Int Neuropsychol Soc.2003;97 :1016– 1026
- Gozal D. Sleep-disordered breathing and school performance in children. Pediatrics1998;102 :616– 620
- ↵Suratt PM, Peruggia M, D'Andrea L, et al. Cognitive function and behavior of children with adenotonsillar hypertrophy suspected of having obstructive sleep-disordered breathing. Pediatrics.2006;118 (3). Available at: www.pediatrics.org/cgi/content/full/118/3/e771
- Mulvaney SA, Goodwin JL, Morgan WJ, Rosen GR, Quan SF, Kaemingk KL. Behavior problems associated with sleep disordered breathing in school-aged children–the Tucson children's assessment of sleep apnea study. J Pediatr Psychol.2006;31 :322– 330
- ↵Montgomery-Downs HE, Crabtree VM, Gozal D. Cognition, sleep and respiration in at-risk children treated for obstructive sleep apnoea. Eur Respir J.2005;25 :336– 342
- ↵Wechsler D. Wechsler Intelligence Scale for Children (WISC-III). 3rd ed. San Antonio, TX: The Psychological Corporation; 1991
- ↵Adams W, Sheslow D. Wide Range Assessment of Memory and Learning (WRAML). Willmington, DE: JASTAK Associates; 1990
- ↵Connors CK. Connors Continuous Performance Test. Toronto, Ontario, Canada: Multi-Health Systems; 1992
- ↵Kordas K, Lopez P, Rosado JL, et al. Blood lead, anemia, and short stature are independently associated with cognitive performance in Mexican school children. J Nutr.2004;134 :363– 371
- ↵Lezak MD, Howieson DB, Loring DW. Neuropsychological Assessment. New York, NY: Oxford University Press; 2004
- ↵Kaufman AS. Intelligent Testing with the WISC-III. New York, NY: Wiley; 1994
- ↵Beebe DW, Lewin D, Zeller M, et al. Sleep in overweight adolescents: shorter sleep, poorer sleep quality, sleepiness, and sleep-disordered breathing. J Pediatr Psychol.2006; In press
- Copyright © 2007 by the American Academy of Pediatrics