Professional Documents
Culture Documents
To cite this article: Marjo–Riitta Järvelin, Elina Mäki–torkko, Martti J. Sorri & Paula T. Rantakallio
(1997) Effect of Hearing Impairment on Educational Outcomes and Employment up to
the Age of 25 Years in Northern Finland, British Journal of Audiology, 31:3, 165-175, DOI:
10.3109/03005364000000019
Article views: 95
Abstract
The association between hearing impairment in adolescence and school performance and the outcome of
education was studied among 25-year-old subjects followed since pregnancy in the Northern Finland
birth cohort. The series, 395 subjects with abnormal hearing and 977 randomly selected controls, was
based on a questionnaire on hearing and school achievement sent to 11780 members of the cohort alive at
the age of 14 years, and on audiometric screening test requested from health centres. Hearing loss was
defined as ‘clinically significant’ if the pure tone average (PTA; mean of the thresholds at 0.5, 1 and
2 kHz) exceeded 25 dB in the better ear; a threshold of 2 30 dB at 4 kHz and a PTA of I 25 dB as ‘4kHz
loss’; and as ‘slightly abnormal’ if any of the thresholds exceeded 20 dB at any frequency and the case did
not belong to the above two categories. The more severe the hearing impairment, the poorer was the
child’s performance a t elementary school. Those with normal hearing and those with a slightly abnormal
or 4 kHz loss were equally often accepted for intermediate education (88%), while those with a clinically
significant loss had the lowest acceptance figures (64%). When adjusting for neurological and social con-
founders, excluding mental disability, the risk of not qualifying from intermediate or higher education at
all was twice as high among those with a clinically significant loss as among the controls (OR 2.1,95% CI
1.13-3.8), and was still elevated after adjustment had been made for all the relevant perinatal, neurologi-
cal and social factors (OR 1.9,95% CI 1.02-3.6). 14% of those with a clinically significant hearing loss,
9% of the subjects with a 4 kHz loss and 7% of those with normal hearing were unemployed at the age of
25 years. Hearing impairment appears to have effects on both the outcome of education and employment
status.
Number of
school
Yam
Downloaded by [] at 19:15 15 March 2016
18
16
17
5
9 Universities
7
15 3
High
14 2
grade
13 I
(Tertiary
education)
t I
Polytechnic, 4
12
grade
I1 2 secondary
schools equivalent (Upper
10 schools secondary
education)
I .....
9 9 Secondary
school -
8 B Elementary
schools
1 7 Primary
6 6
stage
and
5 5 special (Basic
schools
4 4 education)
3 3 Primary -
school
2 2
1 1
.....
F're-primary education in daycare centres and primary Prcprimary
school stage
Fig. 1. The Finnish educational system. The corresponding nomenclature according to the International
Standard Classification of Education is given in brackets. The boxes indicate the number of years taken
to complete each stage. (Modifiedfrom Statistics Finland, 1992)
Hearing impairment and educational outcomes 167
ernmost provinces of Finland in 1966 and from the child guidance centres and hospi-
(Rantakallio, 1969). This cohort accounts for tals. Only a few of those in the class appropriate
96% of all pregnancies in the area with an for their age were tested, while a half of those in a
expected date of delivery in that year. Of these lower class and 95% of those in special schools or
children, 11780 were alive at the age of 14 years, not at school at all were tested (Rantakallio and
when a questionnaire was sent out to each child von Wendt, 1986).
or family, dealing with social standing, schooling All applications and preferences for intermedi-
and hearing. This achieved an exceptionally high ate education are stored in national registers
response rate of 97%. maintained by the Ministry of Education, data
The questionnaire replies indicated that 425 from which were added to the records of the birth
children were suffering from hearing loss. The cohort. Of the adolescents in this cohort, 95%
latest audiograhs were requested for these chil- took part in the national application system,
dren from primary care, with a 97% response almost half of them more than once.
rate, and for a random sample of 1000 children The highest qualification obtained was ascer-
reported to have normal hearing, the response tained from Statistics Finland, as was employ-
rate being 96% (Sorri and Rantakallio, 1985). ment status.
The correspondence between audiometry and
reported hearing was good, and the audiometric Specifications of the outcomes variables
Downloaded by [] at 19:15 15 March 2016
level comprising about 10-11 years’ education PTA was >25 dB in the better ear. As the last
and the upper level about 12years (Fig. 1). group was small (n = 55), they were analysed
mostly as one entity. The impaired hearing in
Highest level of education (Finnish Standard Table 1 comprises the above three categories
Classification of Education, 1992): combined. Bone conduction thresholds were not
taken into account as they were not measured in
1. The lowest level of tertiary education, about many cases. This classification differs from the
13-14 years of total schooling (vocational conventional grading, but the intention was to
school, polytechnic or equivalent level, pro- include even minimal losses, because their possi-
grammes lasting 4-5 years at vocational ble effect on learning could not be ignored until
schools, e.g. high national certificate engineer, analysed.
HNC). Epilepsy was considered to have been present if
2. Lower basic degree in tertiary education (low- there had been at least one episode of paroxysmal
est true university level or the equivalent), disturbances of consciousness,sensation or move-
about 15 years of total schooling (e.g. BSc, ment, primarily cerebral in origin, not associated
nursery school teacher). with acute febrile episodes (von Wendt et al.,
3. Higher basic degree in tertiary education, a 1985a).The definition of cerebralpalsy formulated
total of about 16 years (e.g. honours degree, by the Little Club of London (Pharoah, 1981) was
Downloaded by [] at 19:15 15 March 2016
Table 1. Specifications of the variables and association of perinatal and certain medical and social
factors with hearing impairment. OR = odds ratio, CI = confidence interval.
Normal hearing Impaired hearing’ Fisher’s
(total 997) (total 395) exact test
n % n % O R CI P
Birth weight under 86 8.8 54 13.7 1.6 0.01
10th percentile 1.1-2.4
Birth weight <2500g 30 3.1 23 5.8 2.0 0.02
1.1-3.4
Gestational age 68 7.0 52 13.2 2.0 <0.001
<37 weeks 1.4-3.0
Epilepsy 10 1.o 11 2.8 2.8 0.02
1.2-6.6
Perinatal brain damage2 8 0.8 14 3.5 4.5 <0.001
1.9-10.7
Other neurological injury’ 24 2.5 18 4.6 1.9 0.04
1.0-3.5
Cerebral palsy 2 0.2 3 0.8 3.7 0.2
0.6-22.4
Downloaded by [] at 19:15 15 March 2016
‘Includes: (a) slightly abnormal, if any of the thresholds exceed 20 dB and the case does not belong to the
next groups; (b) 4 kHz loss, threshold 230 dB at 4 kHz and a pure tone average in the better ear 125 dB;
(c) clinically significant loss, pure tone average >25 dB in the better ear.
21ncludesneonatal asphyxia, cerebral lesion or intracranial haemorrage, Apgar 0 at 1 min or <5 at 15 min,
neonatal convulsions (Rantakallio et al., 1987).
)Cranial fractures, contusion or concussion, spinal cord and other nervous damage (excluding perinatal brain
damage).
Social group was based on the father’s occupation at the time of the child’s birth and the common apprecia-
tion of different professions (Sosiaaliryhmitys, 1954; Alestalo and Uusitalo, 1978). Children without a known
father were classified into group IV. For analysis purposes, social class was classified as (1) I or I1 versus oth-
ers or missing; (2) class IV (unskilled labourers, persons on a disability pension, farmers with less than 8
hectares of land) versus others or missing.
5Missing 113.
6Upto the child‘s 14th birthday; standard = original complete family, i.e. both parents from birth to 14 years
versus non-standard = one or no parents at the age of 14, mostly because of divorce (Isohanni et al., 1991).
170 M-R. Jarvelin
The numbers of missing values for the biological 2. Forty-seven of the children with normal hearing
and social background variables varied by up to (4.8%) were in a class lower than was appropriate
5.3%, and these were added to the reference class, for their age at a normal elementary school or were
with the exception of social mobility, for which the in a special school for disabled children, the corre-
113 missing values were excluded. Data were lack- sponding figure among those with slightly abnor-
ing in the national register on the eventual educa- mal hearing being 15out of205 (7.3%), that for the
tional qualifications of 244 subjects, mostly children with a4 kHz loss 11out of 135 (8.1%),and
because they were abroad (n = 70), but using the that for those with aclinically significantloss 18out
other available data set it was possible to find out 55 (32.7%). Likewise, Fig. 2 shows that the more
that 94% of these had completed at least elemen- severe the hearing impairment, the poorer was the
tary school and 6% had been at a special school at child's performance at elementary school, espe-
the age of 14. The other educational data sug- cially when considering poorer than average per-
gested that the most appropriate group in which to formance (mean score 7.4 or less).
include all those with missing information on their Mental subnormality or disability (IQ 5 84)
final education was the reference group in the was observed more often among those with hear-
logistics regression analysis, i.e. those who had ing impairment. Twenty-three of the 977 controls
completed elementary or special school. (2.4%) were either of subnormal intelligence or
mentally disabled, as were 20 of the 395 with a
Results hearing disorder (5.1%; P < 0.015). Only one
Downloaded by [] at 19:15 15 March 2016
Social background and perinatal outcome child with an IQ of 71 and a 4 kHz loss was inte-
The associations of perinatal and social factors grated into the class appropriate for his age, while
with hearing impairment are presented in Table 1. the rest were in a lower class or at a school for
Even a modest decrease in the birth weight to handicapped children, with the exception of one
below 2500g or intrauterine growth retardation child with and IQ of 54 and normal hearing, who
denoted a risk of hearing impairment, and perina- was not at school at all (Table 2).
tal brain damage, later brain injuries and epilepsy
were more often reported among the subjects with Application and entrance to intermediate level
hearing impairment. Of the social variables, a low education
level of education on the part of the mother had the The young people with normal hearing, slightly
most significantassociation with impaired hearing. abnormal hearing or a 4 kHz loss were equally
often accepted into intermediate education pro-
Elementary school performance (Basic education) grammes (about SSYo)), but those who had a
The type of schooling at the age of 14 years is pre- clinically significant disorder had the lowest fig-
sented in relation to hearing impairment in Table ure (64%), and a large proportion of them
Table 2. Type of schooling at the age of 14, by hearing impairment
Hearing impairment'
I I I I I I I I
i L
Mean scores on
school repmi
87.5-0.4
8 6 . 5 - 7.4
8 <6.5 (wont
Fig. 2. Elementary school performance at the age of 14 years among adolescents with normal and
impaired hearing, expressed as mean school grades in four categories. For definitions of groups, see
footnotes to Table I .
(32.7%) did not apply at all (Table 3). Twenty- ment, as was also the case with elementary
two of the 128 who did not apply (17.2%) had an school. Even the upper level of this education, i.e.
IQ of less than 85, and 9 of these had a hearing the matriculation examination or at least 3 years
disorder. Vocational school programmes were of vocational school, was more common among
chosen by the youngsters with hearing impair- the control subjects. A university degree was
ment more often than upper secondary school more common among those with normal hearing
enrolment. than with impaired hearing (4.7% versus 2.3%,
P = 0.04). To evaluate the effect of hearing on the
Outcome of education by the age of25 years outcome of education, the latter was stratified by
The lowest level of intermediate education both IQ and hearing impairment, but mental dis-
(10-11 years of schooling, i.e. vocational pro- ability did not affect the figures noticeably, so
gramme of less than 3 years) was more often the that the percentages in Table 4 are presented
final achievement of those with a hearing impair- without stratification.
Table 3. Application and entrance to intermediate level education
Hearing impairment'
Normal Slightly 4 kHz Clinically Total
hearing abnormal loss significant
loss
n % n % n % n % n
Did not apply 81 8.3 17 8.3 12 8.9 18 32.7 12802
Applied, not accepted 19 2.0 6 2.9 4 3.0 2 3.6 50
Accepted forvocational school 437 44.7 113 55.1 68 50.3 21 38.2
Accepted for upper secondary school 440 45.0 69 33.7 51 37.8 14 25.5 41
Total 977 100.0 205 100.0 135 100.0 55 100.0 1372
Hearing impairment'
Logistic regression was fitted first with hear- dently equally important for the outcome of
ing and outcome (reference: basic education, i.e. education as a hearing impairment.
elementary school or special school) in the The group with clinically significant hearing
model to explore the risk of not obtaining inter- loss was further divided into 28 (51%) with a
mediate or higher level education at all. Slightly PTA of 26-40 dB, 15 (27%) with a PTA of 41-80
abnormal hearing or a 4 kHz loss did not signifi- and into 12 (22%) with a PTA of 281. Among
cantly affect the outcome of education, but a them there was a trend for a worse educational
clinically significant impairment did have an outcome associated with the increasing grade of
effect (crude OR 2.3, 95% CI 1.3-4.2). When hearing impairment after allowing for the above
allowing for social background factors, with or confounders (OR for the subgroup with a PTA of
without neurological disability but excluding 281 was 4.6, 95% CI 1.4-15.8; normal hearing as
mental disability, the risk of not obtaining an the reference and slightly impaired hearing and
intermediate or higher level education was simi- 4 kHz loss combined in the model).
larly twofold (OR 2.1,1.13-3.8). When mental
disability was also added to the model, the risk Main occupational activity and employment at 25
was about the same as in the full model, which years
showed that clinically significant hearing loss Altogether 7% of the subjects with normal hear-
still carried an elevated risk after adjustment ing were unemployed, the corresponding figures
had been made for all the significant perinatal for those with a 4 kHz loss and a clinically signifi-
and social factors (OR 1.9, 95% CI 1.02-3.5). cant impairment being 9 and 14%, respectively.
Social factors such as low social class (OR 1.5, 17% of the controls and those with a slightly
95% CI 1.l-2.1), non-standard family status abnormal hearing or a 4 kHz loss were still study-
(OR 1.6, 95% CI 1.2-2.3) or a young mother ing, compared with only 6% of those with a clini-
(OR 2.0, 1.2-3.2) were found to be indepen- cally significant hearing impairment (Fig. 3).
Hearing impairment and educational outcomes 173
mother
.Sick pension
Bunemployed
W Student
Fig. 3. Main occupational activity at the age of 24 years by hearing ability (for definitions of groups, see
footnotes to Table I ) . The group ‘other’ includes housewives, etc. and ‘unknown’ those on the register
but of unknown activity and those not on the register at all.
ence of early perinatal brain damage, and a clear The most marked difference between the cases
association was observed, which points to the and the controls was seen in the completion of
importance of checking a child’s hearing after a higher education, as also observed by Parving
brain injury. and Christensen (1993). Without a doubt, the
Our analysis shows that the young people with combined difficulties caused by a hearing impair-
a hearing impairment had both a poorer basic ment and associated disabilities have affected
school performance and a more modest later out- their performance at this level. In addition, any
come of education, which is partly connected delay in the diagnosis of hearing impairments,
with the higher incidence of associated handicaps even mild ones, may detract from language skills
such as mental disability and perinatal brain and primary school performance, and also from
damage. With few exceptions, the mentally dis- the later outcome, as has been suggested earlier
abled children, including those of subnormal (Tharpe and Bess, 1991).
intelligence, attended special schools, so that the The results show that even though the statu-
stratification by type of school reduced this con- tory schooling system offers possibilities for edu-
founding effect. After adjusting for social and cation, youngsters with hearing disorders did not
medical factors, it was also found that hearing exploit these effectively, and this may have led to
impairment had an independent untoward effect poorer occupational activity in later life. This
on the final outcome of education. It has been could also be due to attitudes of employers, espe-
Downloaded by [] at 19:15 15 March 2016
pointed out in the literature (Bess et al., 1986; cially with the present high level of unemploy-
Bovo et al., 1988) that even a mild hearing loss ment in Finland. Diagnostic and rehabilitation
may affect school ability and result in poorer per- efforts should be focused on these children and
formance by these children, as was shown in the their families as early as possible to guarantee
present analysis by the lower than average school optimal language skills in order to allow each
marks gained by all the groups with hearing dis- individual to increase his or her chances of lead-
orders, even by those with only slightly abnormal ing and independent adult life.
hearing. An effect was also seen in that the more
severe the hearing impairment was, the poorer Acknowledgements
the elementary school performance. The multi- Financial support was received from the Medical
variate model revealed that, allowing for the Research Council of the Finnish Academy.
important confounders affecting education, only
a clinically significant hearing loss (PTA >25 dB) References
had an independent adverse effect on the final Alestalo M, Uusitalo H. Occupational prestige
outcome of education, and further in this group and its determinants: The case of Finland.
there was a trend for a worse educational out- Research group for comparative sociology,
come associated with the increasing grade of University of Helsinki. Research Reports,
hearing impairment. 1978;20: 1-54.
Entrance to the intermediate level of education Anagnostakis D, Petmezakis J, Papazissis G ,
was equally common for the cases as for the con- Messaritakis J, Matsaniotis N. Hearing loss in
trols, with the exception of those with clinically sig- low-birthweight infants. Am J Dis Child 1982;
nificant hearing loss.The fact that all the case took 136:6024.
part in compulsory education reflects that it is Astbury J, Orgill AA, Bajuk B, Yu VYH. Neu-
quite possible for disabled people to achieve at rodevelopmental outcome, growth and health
least and intermediate-level education. Educabil- of extremely low-birthweight survivors: how
ity is not bound to normal elementary school per- soon can we tell? Dev Med Child Neuroll990;
formance. On the contrary, a certain number of 32: 582-9.
places in intermediate education are specifically Axelsson A, Aniansson G , Costa 0.Hearing loss
allocated for pupils coming from schools for the in school children. Scand Audiol 1987; 16:
mentally or otherwise disabled. This can be seen in 137-43.
the present study in that slightly abnormal hearing Barr B, Anderson H, Wedenberg E. Epidemiol-
or 4 kHz loss did not appreciably affect the out- ogy of hearing loss in childhood. Audiol 1973;
come of education, which shows that the education 12: 426-37.
system is relatively well able to take care of these Barr B, Stensland Junker K, Svard M. Early dis-
children and young people. covery of hearing impairment: A critical evalu-
Hearing impairment and educational outcomes 175
ation of the BOEL test. Audiol 1978,;17: 62-7. review. J R SOCMed 1981;74: 516-20.
Bergman I, Hirsch RP, FriaT, Shapiro SM, Holz- Rantakallio P. Groups at risk in low birth weight
man I, Painter MJ. Cause of hearing loss in the infants and perinatal mortality. Acta Paed
high-risk premature infant. J Pediatr 1985; 106: Scand 1969;(Suppl193): 1-71.
95-101. Rantakallio P. The longitudinal study of the
Bess FH, Klee T , Culbertson JL. Identification, Northern Finland birth cohort of 1966. Paedi-
assessment and management of children with atr Perinat Epidemioll988; 2: 59-88.
unilateral sensorineural hearing loss. Ear Hear Rantakallio P, von Wendt L. Mental retardation
1986;7: 43-51. and subnormality in a birth cohort of 12000
Bovo R, Martini A, Agnoletto M, Beghi A, children in Northern Finland. Am J Ment
Carmignoto D, Milani M, Zangaglia AM. Defic 1986;4: 380-7.
Auditory and academic performance of chil- Rantakallio P, von Wendt L, Koivu M. Prognosis
dren with unilateral hearing loss. Scand Audiol of perinatal brain damage: a prospective study
1988; (Suppl30): 71-4. of a one year birth cohort of 12000 children.
Breslow NE, Day NE. Statistical methods in can- Early Hum Dev 1987; 15:75-84.
cer research volume 1 -The analysis of case- Simmons FB. Patterns of deafness in newborns.
control studies. IARC scientific publications Laryngoscope 1980: 90: 448-53.
no. 32. International agency for research on Sorri M, Jounio-Ervasti K, Palva A. Hearing of
cancer. Lyon, 1980. 15-year-old people in Northern Finland.
Downloaded by [] at 19:15 15 March 2016