Early Human Development 89 (2013) 721–726

Contents lists available at SciVerse ScienceDirect

Early Human Development

journal homepage: www.elsevier.com/locate/earlhumdev

Early vocabulary in full term and preterm Estonian children

Astra Schults a,b,⁎, Tiia Tulviste a,b, Elis Haan a
a
University of Tartu, Estonia
b
Centre of Behavioral and Health Sciences, Estonia

a r t i c l e i n f o a b s t r a c t

Article history: Background: Preterm birth is considered to be a high risk factor for child development and early vocabulary
Received 13 November 2012 can be used as an indicator for later development.
Received in revised form 14 April 2013 Aims: The aim of the present study is to compare the size of early vocabulary, proportional use of different
Accepted 15 May 2013 word categories, and mean length of utterance (henceforth MLU) of preterm and full term children.
Method: The sample consisted of 40 preterm (corrected ages 16–25 months) and two matched groups of full
Keywords:
term children. First full term group consisted of 120 children who were matched by age and gender. Second
Preterm children
Early vocabulary
full term group consisted of 109 children who were matched by age, gender and size of productive vocabu-
Language development lary. The data for this study were gathered using the Estonian adaptation of MacArthur–Bates Communicative
CDI Development Inventory: Words and Sentences.
Estonian language Results: Full term children who were matched by age and gender had larger vocabulary as compared to
Parental report the preterm children's vocabulary (U = 1758.5, p = 0.01). Poisson regression yielded that age, gender,
and preterm birth explained significantly the variance in the vocabulary size. Poisson regressions showed
that all three variables explained significantly variance in proportional use of social terms and predicates.
Age had significant effect for proportional use of common nouns. Age and preterm birth had a significant
effect on the proportional use of function words. MLU was shorter in preterm than in full term children
(U = 1125.0, p = 0.002).
Conclusions: Estonian preterm children's vocabulary is slightly smaller than full term children's vocabulary.
There is a difference in the proportions of word categories used, as preterm children use more social terms,
and less predicates, and function words.
© 2013 Elsevier Ireland Ltd. All rights reserved.

1. Introduction who have found to be healthy, other studies have allowed for some
medical problems, and still other studies have allowed for severe
Early vocabulary has been demonstrated to be a good predictor of medical problems. The gestational ages and birth weights of the pre-
later development. Preterm birth is considered to be a high risk factor term children have varied from study to study as well. Keeping these
for the child development, including language development. Early confounding factors in mind some differences in preterm and full
diagnosis for language problems in preterm children and enrollment term children language development are brought out next. In the
in language-focused interventions permits to anticipate and alleviate beginning of productive vocabulary development starting around
some of the future developmental problems these children might the age of 10 months, preterm born children have been shown to pro-
face. The aim of the present study is to compare the size and compo- duce fewer canonical syllables (well formed syllables with consonant
sition of early vocabulary and MLU of preterm and full term children. vowel structure) than the full term children [30]. At 24 months pre-
Preterm children's language development has been shown to dif- term born children have smaller productive vocabulary [9,10] and
fer from full term children's language development starting as early more social terms (words for games, routines and animal noises)
as ten months of age and continuing to the school years. Those differ- compared to the full term children [9]. At the same age preterm chil-
ences tend to be slight, not reaching statistical significance on every dren have been found to use less complex morphology and syntax
occasion, but the differences are still consistently found. As different [10]. Preterm born children have been shown to have fewer verbs
researchers have had different criteria for including preterm born in their vocabulary at the age of 42 and 60 months [34]. At this age
children in the studies these criteria probably have an effect on the they also produce fewer and shorter sentences [34]. At school age
results obtained. Some studies have included those preterm children preterm born children have been shown to have smaller receptive
vocabulary, deficits in sentence comprehension [33] and naming dif-
⁎ Corresponding author at: Institute of Psychology, University of Tartu, Tiigi 78-336,
ficulties [33,39]. Even at the age of ten years preterm born children
50410 Tartu, Estonia. Tel.: +372 7 375906; fax: +372 7 375900. had smaller receptive and productive vocabularies compared to the
E-mail address: [email protected] (A. Schults). full term born children [36]. According to these previous results we

0378-3782/$ – see front matter © 2013 Elsevier Ireland Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.earlhumdev.2013.05.004
722 A. Schults et al. / Early Human Development 89 (2013) 721–726

assume that the preterm children in our study will have smaller pro- at a constant low level [2,12,13]. Explanations for the noun bias
ductive vocabulary and shorter sentences compared to the full term include the possibility that the proportion of nouns in child's vocabu-
children. We also assume that the differences seen at the earlier age lary corresponds to the proportions of nouns in child-directed speech
(at 16 to 18 months of age) will be still present at the age of 24 and [1,21] and the possibility that different parental strategies lead chil-
25 months. dren to either produce more nouns or to produce more predicates
The differences found in language development between preterm (e.g. maternal use of directives increases the proportion of verbs in
and full term children have been explained by many authors. Three child's early vocabulary) [20,22,23]. Preterm children have been
levels of explanations can be differentiated. First, there are neurolog- shown to have more nouns and less function words in their produc-
ical deficits explanations which focus on the possible damage and the tive vocabulary compared to full term children [9,10,24,25]. It is
location of damage or on neural maturation. It has been shown that suggested that overall pattern of vocabulary development in the use
extremely low birth weight children are more prone to the damage of different semantic categories is delayed rather than deviant for
to the periventricular region of the brain in the perinatal period. the preterm children compared to the full term children [9]. In this
Damage to this region is associated with later deficiencies in working study we aim to see if the same pattern applies to our sample and
memory, difficulties with sentence comprehension, reduced recep- to show that the productive vocabulary size rather than child's age
tive vocabulary and naming difficulties [33]. Neural maturation has has an effect on the type of words belonging to child's vocabulary.
been found to have an effect on the infants' ability of speech discrim- Girls' vocabulary has been shown to be larger than boys' vocabu-
ination [37]. Second, there are explanations focusing on underlying lary during the first years [7,15,17,18]. Girls' advantage has been
cognitive processes. For example, it has been proposed that children explained by faster maturation process. It has also been explained
born premature and with very low birth weight perform poorer in by gender based communication styles [26] as parents tend to speak
language measures because of slower information processing and more with the girls and try to make girls express themselves verbally
attention deficits [32]. Very immature preterm birth has been shown more [27]. As to premature children's vocabulary development gen-
to affect phonological working memory abilities leading to persisting der difference favoring girls has been shown to exist in some research
albeit not severe difficulties in grammar development [35]. Through [28] but not in others [10,29].
school age as they grow older the preterm children can develop In this study, we assumed that older children have larger produc-
more problems like having difficulties with attention, memory and tive vocabularies and that full term born children have larger produc-
reasoning skills instead of growing out of their problems [11]. Third, tive vocabulary compared to the preterm children. We compared the
there are explanations dealing more specifically with language per- production of words belonging to different categories (social terms,
ception. These include phonetic perception and word segmentation common nouns, predicates, and function words) and expected
ability under the age of 12 months which have been shown to be preterm children to use proportionally more social terms and more
related to productive vocabulary at 24 months [38,40]. We assume nouns and fewer function words as compared to the full term chil-
that neural maturation as well as early phonetic perception and dren. We also checked if the difference was connected to smaller
word segmentation play part in our sample's results as we have not vocabulary size. We expected full term children to have higher MLU
included any preterm children with perinatal brain damage or later compared to the preterm children's MLU. We expected the full term
medical complications. girls to have more words in their productive vocabulary compared
As to productive vocabulary development around 16 to 18 months to the full term boys and checked for a gender difference in the pre-
of age the size of children's productive vocabulary grows significantly term group. In addition to that we ran regression analyses to show
so that instead of some words there will be dozens of words in possible age, gender, and being born preterm effects on the vocabu-
children's active vocabulary [1,2]. As the productive vocabulary size lary size as well as on the use of words belonging to different
reaches 20–50 words rapid word acquisition follows in many categories.
cases [3–5]. This more rapid growth of productive vocabulary starts
oftentimes during the second year of life of the child [6]. The rapid 2. Method
increase in the size of productive vocabulary has been explained
in several ways. One explanation brings forth the idea that vocabulary 2.1. Participants
starts to grow fast as the children discover that things can be named
and understand the concept of naming [7]. Another explanation con- The sample consisted of 40 preterm children (16 boys and 24
nects the growth in productive vocabulary to the development of girls) and their matched controls. Preterm children were born before
concepts and the ability to categorize. Growth in vocabulary follows the 36th week of pregnancy (M = 30.6, SD = 0 2.3, range of 24–35).
as the concepts become more detailed and easier to separate from Mean weight of preterm children at birth was 1618.1 g (SD = 388.8,
one another [8]. It has been shown that productive vocabulary devel- range of 840–2500). Preterm children were recruited by medical per-
opment is slower for the children born prematurely [9,10]. Children sonnel. The selection criteria were set so that only children with no
born preterm have also been found to produce shorter and simpler congenital, physical or severe neurological anomalies were recruited.
utterances compared to full term children [10,29]. All of the preterm children were described as healthy by their par-
Research has shown that some word categories are used propor- ents. Corrected age was used for the preterm children [24,31].
tionally more by the children according to the size of their productive Full term children were recruited by pediatricians, play-group
vocabulary. When the vocabulary is less than 50 words social terms teachers, and those parents who had previously completed the inven-
form the largest proportion [4,12]. The category of social terms refers tory about their own child's vocabulary. Full term children were
to words for sound effects and animal sounds, people, and games and matched for preterm children from a larger sample (N = 752). We
routines. When the vocabulary grows over 50 words common nouns formed two control groups. The first full term group (henceforth
(e.g., words for animals, vehicles, toys, food and drink, clothing, body FT1) was formed to check for the effects of prematurity to productive
parts, furniture and rooms, small household items) form the largest vocabulary size and composition. It included 120 children (47 boys
proportion of vocabulary in several languages (e.g., Danish, Dutch, and 73 girls). Match criteria included gender and age in months (in
English, Estonian, Finnish, French, Italian, Spanish) [2,4,12–17,19] case of preterm children we used their corrected age). The second
but not in others (e.g., Mandarin) [20]. At the same time the use of full term group (henceforth FT2) was formed to check for the effect
predicates (including verbs and adjectives) shows a slow and steady of productive vocabulary size to the productive vocabulary composi-
pattern of growth and the use of function words (e.g., pronouns, tion. It included 109 children (48 boys and 61 girls). The match
question words, prepositions and locations, and quantifiers) remains criteria for the FT2 included overall vocabulary size in addition to
 A. Schults et al. / Early Human Development 89 (2013) 721–726 723

age and gender. In addition, all of the full term children had to be distribution of ECDI2 results. To determine possible statistically
healthy (without significant birth trauma, and without serious medi- significant differences between preterm and full term children's pro-
cal problems according to parental report). See Table 1 for age and duction vocabulary size and composition we compared preterm
gender distribution for the entire sample. group to FT1. To determine if the composition of productive vocabu-
All children chosen for this study were acquiring Estonian as their lary depends on vocabulary size we compared preterm group to
first language. Most of the children (96.3%) were from higher or mid- FT2. We compared boys' and girls' word production in preterm
dle SES homes (mothers having at least 12 years of schooling). It group and FT1. Analyses of variance and Scheffé test as post hoc
had to be the first time the parent completed the MacArthur–Bates were used for multiple group comparisons (to determine statistically
Communicative Development Inventory: Words and Sentences significant differences in word production between age groups both
about the child participating in this study. in preterm group and FT1). Poisson regression model of factors
There were neither statistically significant differences in distribu- predicting the production of words is presented.
tion of gender nor in distribution of corrected age in months between
preterm children group and either of full term groups. 3. Results

2.2. Materials and procedure 3.1. Age difference in word production

The data for this study were gathered using the Estonian adapta- In FT1 older children had more words in their vocabularies as
tion of MacArthur–Bates Communicative Development Inventory: compared to younger children (F = 9.3, df = 9, p b 0.00001, partial
Words and Sentences (henceforth ECDI 2, (see [18])). The ECDI 2 eta-squared 0.4). As to specific age differences Scheffé test showed
consists of two main parts. The first part contains a list of 680 that 22–25 month old children had larger vocabulary (accordingly
words, organized into semantic categories (e.g., words for toys, 22 month old M = 189, SD = 167.5, range of 2–594; 23 month old
words for people etc.). The number of total words produced was M = 200, SD = 113.3, range of 5–411; 24 month old M = 208,
established for each child according to the total score in ECDI 2. We SD = 168.3, range of 23–447; 25 month old M = 332, SD = 158.0,
organized the list of words into four categories (see also Bates et al. range of 54–552) as compared to 16 month old children's vocabulary
[13], Caselli et al. [12,16], Stolt et al. [4]). The categories included (M = 45, SD = 53.3, range of 5–265). Children who were 25 months
social terms (including sound effects and animal sounds, people, of age also had larger vocabulary as compared to the 17 months of
and games and routines), common nouns (including animals, vehicles, age children's vocabulary (M = 40, SD = 38.4, range of 2–127).
toys, food and drink, clothing, body parts, small household items, The same age difference in the vocabulary size was seen in preterm
furniture and rooms, outside things, places to go), predicates (includ- group (F = 5.1, df = 9, p = 0.0003, partial eta-squared = 0.6).
ing action words and descriptive words), and function words (includ- Scheffé test showed that 25 month old children had larger vocabulary
ing words about time, pronouns, question words, prepositions and (M = 300, SD = 172.5, range of 138–523) as compared to 16 month
locations, quantifiers, and connecting words). We established for old children's vocabulary (M = 23, SD = 19.1, range of 5–63).
each child how many of the words s/he produced belonged to each
of these categories and found the proportion scores of social terms, 3.2. Difference between full term and preterm children in word
common nouns, predicates, and function words. production
The second part consists of questions about sentences in child's
speech. Parents were asked if the child had already started to com- FT1 had larger vocabulary (M = 168, SD = 152.6, range of
bine words in utterances and to provide samples of three longest 2–594) as compared to the preterm children vocabulary (M = 99.6,
utterances their child had produced. SD = 109.6, range of 2–523) (U = 1758.5, p = 0.01) (see Fig. 1).
Parents, who agreed to participate, were handed out or sent copies
of ECDI 2. Parents were asked to check the words that their children 3.3. Different word categories
produced. A subject-information sheet that contained questions
about the child's developmental history, language(s) spoken in the As to different word categories there was a significant difference
home, and the parents' education was also included. between groups as preterm children used proportionally more

2.3. Statistical analyses 400

Statistical analyses were done using Statistica 8. The significance 350 344

level was set at 5%. Mann Whitney U-tests were used for two group
300
comparisons. Nonparametric tests were used because of the skewed 300
Vocabulary size

250
223
Table 1 209 208
200 191
Gender and age distribution in the sample.
172
Child's age in months 150 144 140
120 124 118 120
16 17 18 19 20 21 22 23 24 25

Boys 100 80 74
Preterm N 5 0 0 2 0 3 2 1 1 2 61 56
FT1 N 9 5 0 7 0 7 7 2 4 6 50 44
FT2 N 10 6 0 6 0 9 6 3 3 5 24
15
Girls 0
Preterm N 8 2 1 1 1 2 3 3 1 2 16 17 18 19 20 21 22 23 24 25
FT1 N 9 9 4 4 3 10 11 12 5 6
Corrected age in months
FT2 N 7 12 6 3 3 4 9 9 2 6
Fullterm Preterm
Note: Corrected age is used for the preterm children, N—number of children, FT1—first
full term group, FT2—second full term group. Fig. 1. Vocabulary size in full term (FT1) and preterm infants at different ages.
724 A. Schults et al. / Early Human Development 89 (2013) 721–726

social terms (M = 36%, SD = 24.3, range of 11–100) as compared for proportional use of function words age and preterm birth had a
to the FT1 (M = 30%, SD = 22.7, range of 9–100) (U = 1865.5, significant effect (see Table 3).
p = 0.04). There was also a significant difference in the use of
words belonging to function words category showing that preterm 3.6. Mean length of utterances
children used them less (M = 3%, SD = 3.5, range of 0–13) com-
pared to the FT1 (M = 5%, SD = 4.2, range of 0–25) (U = 1706.0, was shorter in preterm group (M = 1.45, SD = 1.5, range of 0–5)
p = 0.006). compared to the FT1 (M = 2.73, SD = 2.29, range of 0–12) (U =
There were no differences between FT1 and preterm children's 1125.0, p = 0.002).
use of words belonging to the categories of common nouns and
predicates. 4. Discussion
There were no differences between FT2 and preterm children's
use of words belonging to any of the categories. See Table 2 for The aims of the present study were to compare the size of early
descriptive statistics for vocabulary size and composition in preterm, vocabulary in preterm and full term children, to look for differences
FT1 and FT2 children. in proportions of the use of words belonging to different categories
in these two groups, to see if the differences in the proportions of cat-
egories of words used are due to prematurity or smaller vocabulary,
3.4. Gender difference in word production
and to see if the MLU differs in preterm and full term children.
As expected, older children had more words in their productive
In FT1 there was a significant difference in word production favor-
vocabulary regardless of being born full or preterm. During the sec-
ing girls (U = 2192.5, p = 0.003). On average girls had 117 words
ond year of life rapid increase in productive vocabulary has been
(SD = 152.5, range of 2–594) in their vocabulary and boys had on
shown to take place [1–6]. In our sample preterm and FT1 children
average 110 words (SD = 125.6, range of 2–552) in their vocabulary
had around 30 to 50 words (accordingly) at the age of 16 months
as checked in ECDI 2 by their parents.
and their vocabulary increased to around 300 to 340 words (accord-
No 'gender differences were found in preterm children's vocabu-
ingly) within next 9 months. Thus the pattern of vocabulary growth
lary size.
was found to be similar in FT1 and preterm group.
Full term born children had larger productive vocabulary com-
3.5. Predictive value of corrected age, gender, and birth status. pared to the preterm children as has also been observed in previous
research [9,10]. The difference between the two groups seemed to
Poisson regression analyses were run including preterm group be quite stable over the age span leading us to conclude that the
and FT1. Poisson regressions were computed for vocabulary size as speed of the vocabulary growth is similar in both groups. As the pre-
dependent variable. Age in months, gender (a boy or a girl), and term children included in this study were healthy according to their
birth status (full term or preterm) were used as predictive variables. parents this finding seems to indicate that developmental delay rath-
The regression yielded that all three predictive variables were signif- er than deviant development can be found in healthy preterms. At the
icantly explaining the variance in the vocabulary size (see Table 3). same time it can be concluded that preterm born children continue to
Poisson regression analyses were also run separately for pro- have less words in their vocabulary compared to the full term born
portional use of different word categories. In these analyses the children at the age of 25 months. This finding shows that at least up
dependent variables were the percentages of the words of overall to their 2nd birthday preterm born children have not grown out of
productive vocabulary that belonged to each of the categories. Age, their problems but the gap compared to the vocabulary size of full
gender, and birth status were again used as predictive variables. For term children has not grown bigger either.
proportional use of social terms and predicates all of the three pre- In addition to that it is important to note that when the compari-
dictive variables played a significant part in explaining the variance sons were made within age groups the differences between preterm
in vocabulary size (see Table 3). For proportional use of common and FT1 children's vocabulary size were not statistically significant.
nouns only the age had significant effect on the vocabulary size and Thus, the preterm children's productive vocabulary is smaller than
the FT1 children's productive vocabulary but the difference does not
always reach statistical significance. This may be due to the small
Table 2 number of preterm children in this study. Another possible explana-
Descriptive statistics for productive vocabulary and proportional use of different word
categories.
tion is that as the vocabulary size varies a lot in full term children
this variability covers the low range where preterm children's vocab-
M SD Range ulary size tends to fall.
Preterm group 97 105.9 2–499 Preterm born children used proportionally more social terms as
Productive vocabulary compared to the FT1 children. At the same time there were propor-
Social terms % 36.7 24.3 11–100
tionally less function words in their productive vocabulary compared
Common nouns % 49.5 21.0 0–83
Predicates % 10.6 8.7 0–33 to the FT1 children. We did not find any differences in the proportion-
Function words % 3.2 3.5 0–14 al use of common nouns and predicates in FT1 and preterm group.
FT1 162 145.1 2–550 Preterm children's vocabulary development seems to stay at a lower
Productive vocabulary level within this age group as indicated by the earlier mentioned
Social terms % 30.6 22.7 9–100
Common nouns % 50.6 16.7 0–79
smaller number of words in their vocabulary and also by their propor-
Predicates % 13.6 8.7 0–35 tional use of different types of words. In earlier studies the pattern of
Function words % 5.1 4.2 0–25 having more social terms and less function words has been associated
FT2 102 109.0 4–534 with younger children with smaller vocabularies [2,4,12,13,25]. The
Productive vocabulary
same pattern seems to apply to the preterm children's vocabulary in
Social terms % 37.3 21.9 10–100
Common nouns % 47.2 16.6 0–74 this study pointing once more towards slightly delayed rather than
Predicates % 11.2 8.2 0–27 deviant development.
Function words % 4.2 4.2 0–25 As to the use of words belonging to different categories it is known
Note: FT1—first full term group matched for age and gender, FT2—second full term that the younger the children and the smaller the vocabulary the
group matched for age, gender, and vocabulary size. more social terms and the less function words they tend to have
 A. Schults et al. / Early Human Development 89 (2013) 721–726 725

Table 3
Poisson regressions to predict vocabulary size.

Variables Production of words

All words % of social terms % of common nouns % of predicates % of function words

Wald Statistic p Wald Statistic p Wald Statistic p Wald Statistic p Wald Statistic p

Age 6479.8 .0000 654.1 .0000 145.5 .0000 196.6 .0000 34.7 .0000
Gender 667.9 .0000 27.0 .0000 0.0 ns 42.1 .0000 1.96 ns
Preterm birth 502.2 .0000 4.1 .04 1.6 ns 14.3 .0002 11.1 .001

Note: Corrected age in months was used for the preterm children. Preterm group and first full term group were included in the regression analyses.

[2,4,12,13,25]. To our knowledge the two factors—younger age and to stem from the overall vocabulary size difference as more social
smaller vocabulary—had not been separated in earlier studies. Thus terms and common nouns and less predicates and function words
the question remained if different proportions of word categories in have been shown to be used by the children with smaller vocabular-
productive vocabulary were due to vocabulary size or due to age dif- ies. Thus, there is a small but significant lag in preterm children's
ferences. To answer that question we formed a second control group vocabulary development. However, preterm children's vocabulary
of full term children (FT2) matching them to the preterms not only by development follows the same pattern as full term children's vocabu-
age and gender but also by their productive vocabulary size. 'We did lary development.
not find any differences in proportions of word categories used by
preterm and FT2 children. This finding indicates in the direction Conflict of interest statement
that the proportional use of different word categories is dependent
on vocabulary size. None declared.
Girls were found to produce more words in their as compared
to the boys in the FT1 as it has been found in previous studies Acknowledgments
[7,15,17,18]. In preterm born children's group no gender differences
were found. This finding is in accordance with some of the previous Research for this article was supported by the Estonian Research
studies [10,29] but contradicts some others [28]. One possible expla- Competency Council (grant no. SF0180025s08) and ETF9033 from
nation is, that as girl's advantage has been explained by faster matu- the Estonian Science Foundation.
ration it seems not to apply in preterm children. Another explanation The authors would like to thank the parents participating in the
as to why we did not find the gender differences in preterm group is study and medical personnel for helping to contact the parents.
the small number of preterm children in this study and wide variation
in vocabulary size at this age. References
Regression analyses for the factors affecting productive vocabulary
size showed that child's age in months had an effect on the whole [1] Hoff E. Language experience and language milestones during early childhood. In:
McCartney K, Phillips D, editors. Blackwell handbook of early childhood develop-
vocabulary size (see also [1–6]) as well as on the each of the words ment. Malden MA: Blackwell; 2006. p. 233–51.
categories. The older the children the more words they have. Age dif- [2] Schults A, Tulviste T, Konstabel K. Early vocabulary and gestures in Estonian chil-
ferences alone did not explain the whole variance. Birth status in dren. J Child Lang 2012;39:664–86.
[3] Ganger J, Brent MR. Reexamining the vocabulary spurt. Dev Psychol 2004;40:
addition to age had an effect of its own. Being born preterm had an 621–32.
effect on the vocabulary size as a whole and on all of the word catego- [4] Stolt S, Haataja L, Lapinleimu H, Lehtonen L. Early lexical development of Finnish
ries with the exception of common nouns. Preterm children's produc- children: a longitudinal study. First Lang 2008;28:259–79.
[5] Barrett M. Early lexical development. In: Fetcher P, MacWhinney B, editors. The
tive vocabulary was smaller compared to the full term children's
handbook of child language. Oxford: Blackwell; 1995. p. 96–151.
vocabulary (see also [9,10]). There were proportionally more social [6] Dapretto M, Bjork E. The development of word retrieval abilities in the second
terms and less predicates and function words in the preterm year and its relation to early vocabulary growth. Child Dev 2000;71:635–48.
[7] Reznick JS, Goldfield BA. Rapid change in lexical development in comprehension
children's vocabularies. Also, gender had an effect of its own on the
and production. Dev Psychol 1992;28:406–13.
whole vocabulary size with girls having larger vocabulary compared [8] Lifter K, Bloom L. Object knowledge and the emergence of language. Infant Behav
to the boys (see also [7,15,17,18]). Gender also had an effect on Dev 1989;12:395–423.
the proportion of social terms and predicates used by the children [9] Kern S, Gayraud F. Influence of preterm birth on early lexical and grammatical
acquisition. First Lang 2007;27:159–73.
as girls used less social terms and more predicates compared to the [10] Foster-Cohen S, Edgin JO, Champion PR, Woodward LJ. Early delayed language
boys. In conclusion it can be said that when we include the three development in very preterm infants: evidence from the MacArthur–Bates CDI.
variables that have been separately shown to have an effect on vocab- J Child Lang 2007;34:655–75.
[11] Luciana M. Cognitive development in children born preterm: implications for the-
ulary development (age, gender, and prematurity) into one regres- ories of brain plasticity following early injury. Dev Psychopathol 2003;15:
sion model all of those variables remain significant in explaining of 1017–47.
the differences in the size of productive vocabulary and in the compo- [12] Caselli C, Cassadio P, Bates E. A comparison of the transition from first words to
grammar in English and Italian. J Child Lang 1999;26:69–111.
sition of productive vocabulary. [13] Bates E, Marchman V, Thal D, Fenson L, Dale P, Reznick JS, et al. Development and
Full term children's MLU was longer than preterm children's MLU stylistic variation in the composition of early vocabulary. J Child Lang 1994;21:
(see also [10,29]). MLU is connected to the vocabulary size thus the 85–123.
[14] Nelson K. Structure and strategy in learning to talk. Monogr Soc Res Child Dev
fact that preterm children have smaller vocabulary effects the length
1973;38:1–137.
of the utterances. [15] Bornstein MH, Cote LR, Maital S, Painter K, Park SY, Pascual L, et al. Cross-linguistic
In conclusion, Estonian preterm children's vocabulary develop- analysis of vocabulary in young children: Spanish, Dutch, French, Hebrew, Italian,
Korean, and American English. Child Dev 2004;75:1115–39.
ment seems to follow generally the same pattern as full term born
[16] Caselli MC, Bates E, Casadio P, Fenson J, Fenson L, Sanderl L, et al. A cross-linguistic
children's vocabulary development, major difference being a small study of early lexical development. Cogn Dev 1995;10:159–99.
lag in the vocabulary size. There seems to be a difference in the pro- [17] Fenson L, Dale P, Reznick JS, Bates E, Thal DJ, Pethick S. Variability in early commu-
portions of word categories used, as preterm children use more social nicative development. Monogr Soc Res Child Dev 1994;59:1–185.
[18] Eriksson M, Marschik PB, Tulviste T, Almgren M, Pérez Pereira M, Wehberg S, et al.
terms, less predicates, and less function words even as the age as an Differences between girls and boys in emerging language skills: evidence from 10
affective factor has been taken into account. This difference seems language communities. Br J Dev Psychol 2012;30:326–43.
726 A. Schults et al. / Early Human Development 89 (2013) 721–726

[19] Wehberg S, Vach W, Bleses D, Thomsen P, Madsen TO, Basbøll H. Danish children's [30] Oller DK, Eilers RE, Steffens ML, Lynch MP, Urbano R. Speech-like vocalizations in
first words: analyzing longitudinal data based on monthly CDI parental reports. infancy: an evaluation of potential risk factors. J Child Lang 1994;21:609–31.
First Lang 2007;27:361–83. [31] Hindmarsh GJ, O'Callaghan MJ, Mohay HA, Rogers YM. Gender differences in cog-
[20] Tardif T. Nouns are not always learned before verbs: evidence from Mandarin nitive abilities at 2 years in ELBW infants. Early Hum Dev 2000;60:115–22.
speakers' early vocabulary. Dev Psychol 1996;32:492–504. [32] Ortiz-Mantilla S, Choudhury N, Leevers H, Benasich AA. Understanding language
[21] Hart B. What toddlers talk about. First Lang 2004;24:91–106. and cognitive deficits in very low birth weight children. Dev Psychobiol 2008;50:
[22] Fernald A, Morikawa H. Common themes and cultural variation in Japanese and 107–26.
American mothers' speech to infants. Child Dev 1993;64:637–56. [33] Frisk V, Whyte H. The long-term consequences of periventricular brain damage on
[23] Tomasello M, Kruger AC. Joint attention on actions: acquiring verbs in ostensive language and verbal memory. Dev Neuropsychol 1994;10:313–33.
and non-ostensive contexts. J Child Lang 1992;19:311–33. [34] Le Normand M-T, Cohen H. The delayed emergence of lexical morphology in
[24] Stolt S, Haataja L, Lapinheimu H, Lehtonen L. The early lexical development and its preterm children: the case of verbs. J Neurolinguistics 1999;12:235–46.
predictive value to language skills at 2 years in very-low-birth-weight children. J [35] Sansavini A, Guarini A, Alessandroni R, Faldella G, Giovanelli G, Salvioli G. Are
Commun Disord 2009;42:107–23. early grammatical and phonological working memory abilities affected by
[25] Stolt S, Klippi A, Launonen K, Munck P, Lehtonen L, Lapinheimu H, et al. The Pipari preterm birth? J Commun Disord 2007;40:239–56.
Study Group. Size and composition of the lexicon in prematurely born very-low- [36] Magill-Evans J, Harrison MJ, Van der Zalm J, Holdgrafer G. Cognitive and language
birth-weight and full-term Finnish children at two years of age. J Child Lang development of healthy preterm infants at 10 years of age. Phys Occup Ther
2007;34:283–310. Pediatr 2002;22:41–56.
[26] Wolke D, Meyer R. Cognitive status. Language attainment and prereading skills of [37] Peña M, Pittaluga E, Mehler J. Language acquisition in premature and full-term
6-year-old very preterm children and their peers: the Baravian longitudinal infants. 2010;107:3823–8.
study. Dev Med Child Neurol 1999;41:94–109. [38] Bosch L. Precursors to language in preterm infants: speech perception abilities in
[27] Leaper C, Anderson KJ, Sanders P. Moderators of gender effects on parents' talk to the first year of life. In: Braddick O, Atkinson J, Innocenti G, editors. Progress in
their children: a meta analysis. Dev Psychol 1998;34:327. Brain Research, 189. Amsterdam: Elsevier B.V; 2011. p. 239–57.
[28] Sansavini A, Guarini A, Alessandroni R, Faldella G, Giovanelli G, Salvioli G. Early [39] Saavalainen P, Luoma L, Bowler D, Timonen T, Määttä S, Laukkanen E, et al. Nam-
relations between lexical and grammatical development in very immature Italian ing skills of children born preterm in comparison with their term peers at the ages
preterms. J Child Lang 2006;33:199–216. of 9 and 16 years. 2006;48:28–32.
[29] Kern S, Gayraud F. French CDI words and sentences to assess early vocabulary and [40] Tsao F-M, Liu H-M, Kuhl P. Speech perception in infancy predicts language devel-
morphosyntax of preterm and full-term two year old children. Proceedings from opment in the second year of life: a longitudinal study. Child Dev 2004;75:
the First European Network Meeting on the Communicative Development Inven- 1067–84.
tories. Gävle: University of Gävle; 2006.