|
Development From Infancy to Late Childhood in Children Who Are Blind
A Follow Up to the Bielefeld Longitudinal Study
Birgit Spohn and Michael Brambring
University of Bielefeld, Germany
Faculty of Psychology
University of Bielefeld
P.B. 10 01 31
33501 Bielefeld
Germany
05 21/106-4347
birgit.spohn@uni-bielefeld.de
A study is presented that documents the development of 10 children who are blind from infancy to late childhood. The Bielefeld longitudinal study on early intervention and family counseling for blind infants and preschoolers (Brambring et al., 1995) assessed all areas of development. A follow-up study in 2000-2001 documented the state of development at the age of 11-12 years. This showed that four of the children were mentally retarded; the cognitive ability of the remaining six children was in the normal range. The present study retrospectively assessed the differences in the longitudinal course of cognitive development in the mentally retarded and mentally nonretarded children at the age of 12-48 months. Correlations between the state of cognitive development in infancy and preschool age (longitudinal study) and at the age of 11-12 years (follow-up study) are discussed. The study also examines how well a mental retardation in late childhood can be predicted from the state of cognitive development in infancy and preschool age. Finally, the influence of birth status on the course of development is considered.
Effective intervention requires knowledge about outcome and a model of both what is desirable and what would be risked by nonintervention. It is often unclear whether observed developmental discrepancies or delays in blind children are deviant or merely different convergent paths to eventual adaptation (Freeman et al., 1989). Only longitudinal data provide answers to these questions (Rutter, 1988).
(Freeman, Goetz, Richards & Groenveld, 1991, p. 365).
Longitudinal studies make it possible to plot courses of development directly, giving them a decisive advantage over cross-sectional designs. However, few longitudinal studies have been published on development in children who are blind from infancy or preschool age to school age or adulthood (see Cohen, 1966; Freeman, Goetz, Richards & Groenveld, 1991; Gillman & Goddard, 1974). Furthermore, these studies are all rather old, and their findings cannot be generalized unreservedly to the changing population of children who are blind. In addition, they mostly examined only a few aspects of development. The present follow-up study assessed the state of development in the 10 subjects of the Bielefeld longitudinal study at the age of 11-12 years. As a result, data are now available on their development when they were 12-48 months old and at the end of childhood. The goals of the study were (a) to describe the children's current development level, (b) to trace relations from the state of development in infancy and preschool age to that in late childhood, and (c) to examine the effects of biographical, medical, and sociodemographic data on the course of development.
Method
The Bielefeld Longitudinal Study (1990-1994) collected data on 10 children who were blind from the age of 12 to 48 months. It assessed their development at the following age intervals: 12, 15, 18, 24, 30, 36, 42, and 48 months1 with the Bielefeld Developmental Test for Blind Infants and Preschoolers (BEB-KV, Experimental Version; Brambring, Dobslaw, Klee, Obermann, & Tröster, 1987). This test covers all relevant areas of development: cognition, language, socioemotional development, basic neuromotor skills, daily living skills, and orientation and mobility. Additional information on behavior problems and developmental idiosyncrasies as well as biographical, medical, and sociodemographic data was collected with the Bielefeld Parent Questionnaire for Blind Infants and Preschoolers (Brambring et al., 1987; Brambring et al., 1989). The following criteria were used to select children for the study: (a) complete blindness (light perception at most), (b) blind at birth or during the first year of life, (c) no signs of brain damage, and (d) aged between 9-15 months. The sample contained five girls and five boys. Five of the children were extremely preterm; the other five, full-term.
The follow-up study was carried out in the years 2000-2001 when the children were aged 11-12 years. It assessed general cognitive ability, academic achievement, learning behavior, social behavior, temperament, daily living skills, orientation, idiosyncrasies in locomotion, walking style, body awareness, stereotyped movements, sleep behavior, behavior problems, and mental disturbances. Data were collected from both parents and class teachers. This article only reports the findings on cognitive development. Cognitive ability was measured with the verbal subtests in the German adaptation of the Wechsler Intelligence Scale for Children-III (HAWIK-III; Tewes, Rossmann, & Schallberger, 2000). These subtests are: Information, Comprehension, Similarities, Arithmetic, Vocabulary, and the optional Digit Span.
1. Cognitive Ability at the Time of the Follow-Up Study
Table 1 lists the children's IQ scores (HAWIK-III, Verbal IQ) at the time of the follow-up study. Because the HAWIK-III does not discriminate sufficiently in the well-below-average achievement domain, it also reports raw scores for children whose IQ scores were below 70.
Table 1
|
|
Child 1 |
Child 2 |
Child 3 |
Child 4 |
Child 5 |
Child 6 |
Child 7 |
Child 8 |
Child 9 |
Child 10 |
|
Verbal-IQ |
121 |
108 |
101 |
98 |
96 |
90 |
48 |
46 |
46 |
46 |
|
Raw Score |
|
|
|
|
|
|
22 |
13 |
11 |
1 |
Note:
Verbal-IQ from HAWIK-III and raw scores for children with IQs under 70.
Regarding the presence of signs of mental retardation, the children could be split into two groups:
1. One group of six children with an average (90-109) or even well-above-average IQ.
2. A second group of four children with IQs well below 70.
According to DSM-IV (APA, 1994), the essential features of a mental retardation are an IQ of about 70 or below and significant impairments of adaptive functioning. Hence, the second group definitely met one of the criteria for diagnosing mental retardation. Their scores in performance areas permitting statements on adaptive functioning also indicated a mental retardation. Because the children in both groups were blind, in the following they shall be labeled the mentally retarded group and the mentally nonretarded group.
2. Cognitive Development in Infancy and Preschool Age
The cognitive development of the four mentally retarded and the six mentally nonretarded children when they were infants and preschoolers was examined retrospectively in terms of their raw scores on the cognition scale of the BEB-KV. We analyzed the course of cognitive development, how strongly it differed in the two groups, and what progress the children had made. These issues were examined with analyses of variance (ANOVA) and Mann-Whitney U Tests. Only the measurement waves from 18-48 months could be entered into the ANOVAs, because scores were not available for all children at 12 and 15 months.
Figure 1
Courses of development in infancy to preschool age in the mentally retarded and nonretarded groups
Notes
1. Bivariate ANOVA with repeated measures. First variable: mental retardation vs. no mental retardation; second variable (repeated measure): measurement times; corrected according to Greenhouse and Geisser's formula. Results: significant main effect of first variable, F = 12.515, p < .01, and second variable, F = 11.134, p < .001, and significant interaction, F = 5.424, p < .05.
2. Univariate ANOVAs with measurement times as repeated measure, calculated separately for the mentally retarded and mentally nonretarded groups. Results: (a) mentally nonretarded children: significant main effect of measurement times, F = 8.072, p < .05; (b) mentally retarded children: no significant main effect of measurement times, F = 3.350, ns.
3. Mann-Whitney U Tests of test score differences between the two groups at the individual measurement times. Results: significant differences at 18- to 48-month measurement times, U = 1.00, p < .05; U = 2.00, p < .05; U = 1.00, p < .05; U = 3.00, p < .05; U = 2.00, p < .05; U = 0, p < .01; no significant differences at 12- and 15-month measurement times, U = 1.50, ns.; U = 4.5, ns, one-tailed.
* p < .05. ** p < .01.
Figure 1 illustrates the different courses of development in the mentally retarded and the mentally nonretarded children. The initial bivariate ANOVA for repeated measures revealed significant main effects of mental retardation versus no mental retardation and of the repeated-measures variable measurement times. The significant interaction between the two variables reduced the two main effects, and indicated significantly different courses of development in the two groups. Separate univariate ANOVAs for repeated measures computed for each group showed that only the mentally nonhandicapped group made significant progress during this age interval. No statistically significant developmental progress could be observed in the mentally retarded group. Results of the Mann-Whitney U Tests showed that the scores of the two groups differed significantly at all measurement times between the ages of 18 and 48 months in favor of the mentally nonretarded group. However, there were no significant differences between the two groups at ages of 12 and 15 months.
3. Relation Between Cognitive Ability at the Follow Up and State of Cognitive Development at Infancy to Preschool Age
We studied how strongly cognitive ability at the time of the follow up related to the state of cognitive development when the children were infants and preschoolers. The IQ scores at the follow up were correlated with raw scores on the BEB-KV cognition scale at the ages of 12-48 months. Each relationship was measured with Kendall's rank-correlation coefficient (Kendall's Tau).
Table 2 lists the empirical relations. From the age of 15 months onward, there were significant and substantial relations between the infant and preschool-age scores and those in late childhood. Values tended to increase as a function of age. After the age of 30 months, all coefficients were above .70.
Table 2
Relations Between Intelligence Scores at the Follow Up at Age 11-12 Years (HAWIK-III, Verbal IQ) and Cognition Scores on the BEB-KV at Ages 12-48 Months
|
Age in months (Longitudinal study) |
12
|
15
|
18
|
24
|
30
|
36
|
42
|
48
|
|
Correlation with follow-up test score |
.30 |
.63*
|
.57* |
.60* |
.75** |
.76** |
.79** |
.74** |
Notes. Kendall's rank-correlation coefficients, tested for significance with Fisher's Exact Probability Test. * p < .05. ** p < .01.
4. State of Cognitive Development in Infancy to Preschool Age as Predictor of Cognitive Ability at Follow Up
We also studied how well mental retardation or normal cognitive ability (as given by the IQ scores at the follow up) could be predicted from the level of cognitive development in infancy and preschool age. The following procedure was chosen: For each measurement time in infancy and preschool age, the children's raw scores in the BEB-KV cognition scale were placed in a rank order. Then the size of the raw score was used to split the group into two subgroups whose proportions of the total sample corresponded to the proportion of mentally retarded or mentally nonretarded children at the time of the follow up. For each measurement time, the four children with the lowest scores were assigned to one group (Group II) and the remaining six to the other (Group I).
We then tested how well a mental retardation or a cognitive ability lying in the normal range in the follow-up study could be predicted on the basis of membership of Groups I or II. Because of ranking ties, it was impossible to ascertain the four lowest scores unequivocally up to the age of 30 months. As a result, only the 36 - 48-month measurement times were entered into this analysis.
The procedure chosen permitted a good prediction of mental retardation in late childhood. The error rate was low: A false classification was made in a maximum of two cases per measurement time (at 36 months). A completely correct classification was possible at 48 months. The procedure exhibited a high sensitivity (i.e., children with lowest raw scores were mostly classified correctly as "mentally retarded") and a high specificity (i.e., children with highest raw scores were mostly classified correctly as "mentally nonretarded"). The group classification assigned a child with lowest raw scores (Group II) incorrectly to the group of mentally nonretarded children at only two measurement times. A child with highest raw scores (Group I) was classified incorrectly as mentally retarded at only one measurement time. Furthermore, a false prediction was made at only one measurement time per child.
5. The Impact of Birth Status on the Course of Development
Five of the children were preterm; five, full-term. All preterm children were born before the 30th week of gestation. Three of the preterm children had extremely low birthweights (<1,000 g); two, birthweights slightly above 1,000 g. Blindness was due to a retinopathy of prematurity (ROP) in all preterm children.
An inspection of the children's development as a function of birth status revealed that all four mentally retarded children were preterm. Only one preterm child had a cognitive ability in the normal range as defined by IQ at the time of the follow-up study. Fisher's Exact Probability Test confirmed that the above-random frequency of a mental retardation in the preterm children was statistically significant, c2(1) = 6.67, p < .05. An inspection of birthweight revealed that all three children with extremely low birthweights had a mental retardation, but only one of the two children with very low birthweights. The child with the lowest birthweight (620 g) had the most severe mental retardation. However, there were no significant correlations between birthweight and intelligence scores (HAWIK-III, Verbal IQ and HAWIK-III raw score) at the time of the follow-up study.[1]
The follow-up study shows that 4 of the 10 children are mentally retarded (when tested at the age of 11-12 years) as well as blind. Considering that no brain damage was apparent at the beginning of the longitudinal study, the proportion of mental retardation seems to be relatively high. However, 5 of the 10 participants in the longitudinal study are preterm children with very low or even extremely low birthweights. All the mentally retarded children belong to this group. This finding is in line with studies suggesting a high risk of developing a mental retardation in children with ROP. For example, 8 of the 14 children with ROP examined by Vohr and Garcia-Coll (1985) were mentally retarded. Teplin (1988) found a mental retardation in 7 out of 12 children with ROP. Hecker (1998) also found clear signs of a higher risk of developmental delays in preterm blind children. Recent studies of sighted preterm children point to the significance of birthweight, and there are indications of a higher incidence of developmental impairments in extremely premature sighted children (Wolke & Meyer, 1999). In contrast, findings on less extreme prematurity are mixed.
The literature offers two potential explanations for the increased incidence of developmental impairments in premature children who are blind: The first assumes that ROP may frequently occur in combination with a brain damage that has so far evaded diagnosis. The second assumes that extreme prematurity leads to a heightened biological vulnerability and extremely restricted learning opportunities (see Wolke & Meyer, 1999), and that children who are blind are additionally exposed to a higher risk of developmental impairments than their sighted peers because their blindness restricts the opportunities for compensation (see, also, Hecker, 1998).
The courses of development already differ clearly between mentally retarded and mentally nonretarded children who are blind in infancy and preschool age. Even in early childhood, the former already exhibit a markedly lower performance in the cognitive domain than the latter. Whereas the mentally nonretarded children show clear developmental progress, their mentally retarded peers show only slight, statistically nonsignificant steps. The gap between the two groups gets even larger with increasing age.
There is a clear relation between the state of cognitive development in infancy to preschool age and cognitive ability in late childhood. From the age of 15 months onward, there are significant and substantial correlations ranging from .57 to .79. From a methodological perspective, it nonetheless has to be admitted that the heterogeneity of the sample facilitates such a close relationship. In nonretarded children, the relations between cognitive ability in infancy to preschool age and late childhood are generally lower. The literature reports correlations between intelligence test scores at the age of 24 or 36 months and the age of 10-13 years of .40 (24 months) and .50 (36 months; see, e.g., Bloom, 1974). Under the age of 24 months, correlations are even lower (see, e.g., McCall, 1979). It is only at about the age of 48 months that correlations rise to .70 or .80 (see, e.g., Bloom, 1974; Sameroff, Seifer, Baldwin, & Baldwin, 1993). In retarded children and also in premature sighted children, correlations between scores obtained in infancy to preschool age and those obtained in school age are higher than in nonretarded children. In the Zurich Longitudinal Studies, Largo and von Siebenthal (1998) found correlations ranging from .58 to .80 between EQ scores in the age range of 9-24 months and the IQ at the age of 7 years in mentally retarded children. Although the coefficients are lower in premature sighted children (between .22 and .58), they are still higher than those in full-term sighted children (between .09 and .30).
In the present study, a mental retardation or a cognitive ability in the normal range (according to IQ scores at the time of the follow up) can be predicted relatively accurately from cognitive performance at the age of 36 months and older. The sensitivity and the specificity of this approach is also very high. This could make it possible to detect children at risk and provide them with appropriate interventions. However, ranking ties in the data prevent predictions before the age of 36 months, and further studies will be needed to examine whether the present approach will also work in other samples.
The generalizability of the present findings to the population of blind children as a whole is limited for two reasons: First, the sample only contains children who have been completely blind since birth or the first year of life and had no known brain damage at the beginning of the study, and second, the sample is very small in size.
References
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders: DSM-IV (4th ed.). Washington, DC: Author.
Bloom, B.S. (1974). Stabilität und Veränderung menschlicher Merkmale. Weinheim: Beltz.
Brambring, M., Beelmann, A., Buitenhuis, S., Hecker, W., Kurp, C., Licher-Eversmann, G., & Müller, A. (1995). Frühförderung blinder Kinder. Konzeption und Hauptergebnisse des Bielefelder Projektes. Kindheit und Entwicklung, 4, 149-156.
Brambring, M., Dobslaw, G., Hauptmeier, M., & Hecker, W. (1989). Bielefelder Elternfragebogen für blinde und sehende Klein- und Vorschulkinder - revidierte Fassung. Unveröffentlichter Fragebogen. Bielefeld: Universität Bielefeld, SFB 227, Teilprojekt A3.
Brambring, M., Dobslaw, G., Klee, K., Obermann, S., & Tröster, H. (1987). Bielefelder Elternfragebogen für blinde und sehende Klein- und Vorschulkinder. Unveröffentlichter Fragebogen. Bielefeld: Universität Bielefeld, SFB 227, Teilprojekt A3.
Brambring, M., Dobslaw, G., Klee, K., Obermann, S., & Tröster, H. (1987). Bielefelder Entwicklungstest für blinde Klein- und Vorschulkinder (BEB-KV). 1. Revision. Unveröffentlichtes Testmanual. Bielefeld: Universität Bielefeld, SFB 227, Teilprojekt A3.
Cohen, J. (1966). The effects of blindness on children's development. The New Outlook for the Blind, 60, 150-154.
Freeman, R.D., Goetz, E., Richards, D.P., & Groenveld, M. (1991). Defiers of negative prediction: A 14-year follow-up study of legally blind children. Journal of Visual Impairment and Blindness, 85, 365-370.
Gillman, A.E., & Goddard, D.R. (1974). The 20-year outcome of blind children two years old and younger: A preliminary survey. The New Outlook for the Blind, 68, 1-7.
Hecker, W. (1998). Entwicklungsunterschiede zwischen früh- und reifgeborenen blinden Kleinkindern. Ein Vergleich der Entwicklung und Verhaltensbesonderheiten im ersten bis dritten Lebensjahr. Regensburg: Roderer.
Largo, R.H., & von Siebenthal, K. (1998). Prognostische Aussagekraft von Entwicklungsuntersuchungen im 1. Lebensjahr. In H.G. Schlack (Hrsg.), Welche Behandlung nützt behinderten Kindern. Mainz: Kirchheim.
McCall, R.B. (1979). The development of intelligence functioning in infancy and the prediction of later IQ. In J. Osofsky (Ed.), Handbook of Infant Development. New York: Wiley.
Sameroff, A.J., Seifer, R., Baldwin, A., & Baldwin, C. (1993). Stability of intelligence from preschool to adolescence: The influence of social and family risk factors. Child Development, 64, 80-97.
Teplin, S.W. (1988). Development of the blind infant and child with retinopathy of prematurity: The physician's role in intervention. Birth Defects: Original Article Series, 24, 301-323.
Tewes, U., Rossmann, P., & Schallberger, U. (2000). Hamburg- Wechsler- Intelligenztest für Kinder - Dritte Auflage. HAWIK-III. Bern: Huber
Vohr, B.R., & Garcia-Coll, C. (1985). Neurodevelopment and school performance of very low birthweight infants: A seven year long study. Pediatrics, 76, 345-350.
Wolke, D., & Meyer, R. (1999). Cognitive status, language attainment, and prereading skills of 6-year-old very preterm children and their peers: The Bavarian Longitudinal Study. Developmental Medicine & Child Neurology, 41 (2), 94-109.
[1] This was assessed with Kendall's rank-correlation coefficients. Results: correlation of r = .12, ns., between birthweight and HAWIK-III, Verbal IQ; r = .20, ns., between birthweight and the raw score in HAWIK-III.
|
Please send comments or questions to webmaster@icevi.org.