 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
a Institute of Psychiatry, King's College, London, United Kingdom
b St Michael's Unit, Mercy University Hospital, Grenville Place, Cork, Ireland
c Department of Neonatal Paediatrics, Royal Free and University College Medical School, London, United Kingdom
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
METHODS. We assessed 108 VPT individuals and 67 term-born controls at ages 18 to 19 years with the Eysenck Personality Questionnaire-Revised, short form (EPQ-RS). This questionnaire rates 3 dimensions of personality: extraversion (sociability, liveliness, sensation seeking); neuroticism (anxiety, low mood, low self-esteem); and psychoticism (coldness, aggression, predisposition to antisocial behavior). A fourth scale, "lie," which measures dissimulation, is also derived.
RESULTS. VPT individuals had significantly lower extraversion scores, higher neuroticism scores, and higher lie scores than term-born controls, after controlling for age at assessment and socioeconomic status. P scores were not significantly different between the 2 groups. There was a gender difference in that the increased neuroticism and decreased extraversion scores were accounted for mainly by VPT females. Associations between EPQ-RS scores and neonatal status, adolescent behavioral ratings, and body size at 18 to 19 years were assessed by using Kendall partial correlations, correcting for age at assessment and socioeconomic status. Gestational age, indices of neonatal hypoxia, and neonatal ultrasound ratings were not correlated with EPQ-RS scores. Birth weight was weakly associated with increased lie scores. Rutter Parents' Scale score, a measure of adolescent psychopathology, was associated with an increased neuroticism score. Poor social adjustment in adolescence was associated with an increased lie score. Height and weight at 18 to 19 years were not associated with EPQ-RS, but reduced occipitofrontal circumference was associated with both decreased extraversion and increased lie scores.
CONCLUSIONS. Young adults who are born VPT have different personality styles from their term-born peers. This may be associated with an increased risk of psychiatric difficulties.
Key Words: preterm • neurodevelopment • personality
Abbreviations: VPT—very preterm • VLBW—very low birth weight • EPQ—Eysenck Personality Questionnaire • EPQ-RS—Eysenck Personality Questionnaire-Revised, short scale • df—degrees of freedom • SAS—Social Adjustment Scale
VERY PRETERM (VPT) birth (before 33 weeks' gestation) is associated with an increased likelihood of neurodevelopmental disruption, which may manifest in childhood as neurologic and neuropsychological impairments,1–3 behavior problems,4 and reduced academic attainment.5–9 There has been a relative lack of studies that follow prematurely born children beyond school age and into adulthood. However, there is much evidence that adolescence and young adulthood is a time of active development of both brain structure and function as adult skills and behaviors are acquired.10–13 One dimension of this is the acquisition and stabilization of adult personality, which can be defined as the relatively stable attributes that differentiate one individual from another and influence the formation and maintenance of relationships. There is some evidence that neurodevelopmental insults can affect adult personality; for example, Neugebauer et al14 reported that males who had been exposed in utero to the Dutch famine of 1944–1945 had an increased likelihood of acquiring a diagnosis of antisocial personality disorder in adulthood. Despite this, there have been few attempts to characterize personality patterns in those who are born preterm, partly because such groups are historically rather recent.
Hack et al15 administered self-report questionnaires that explored risk-taking, antisocial behavior and general health status to a group of 20-year-olds who were of very low birth weight (VLBW; <1500 g). Although not synonymous with VPT, the VLBW and VPT populations overlap significantly and are comparable to an extent. The VLBW group reported significantly fewer risk-taking and antisocial behaviors than the control group. Levy-Schiff et al16 found increased anxiety and depression in VLBW adolescents compared with term controls. Rickards et al17 found that VLBW adolescents had lower self-esteem and experienced higher levels of social rejection than their normal–birth weight peers.
As part of a follow-up study of the long-term effects of preterm birth, we administered a self-report measure of personality, the Eysenck Personality Questionnaire-Revised, short form (EPQ-RS)18,19 to a group of VPT individuals and a term-born comparison group at ages 18 to 19 years. The EPQ-RS can be used to derive 4 measures: 3 dimensions of personality (according to Eysenck's personality theory) and a "lie" scale, which is a measure of how honestly the questionnaire is completed (eg, it can detect whether a subject is "faking good"). The 3 personality dimensions are extraversion, neuroticism, and psychoticism. The construct validity of the EPQ-RS has been tested across populations,20 and the short form has been shown to be comparable to the original full-length questionnaire.21 We hypothesized that VPT young adults would show increased neuroticism, reduced extraversion, and increased psychoticism scores compared with term-born young adults.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
2 = 3.13; df = 5; P = .679) from those who were not followed-up. Fifty-three infants who were delivered at term (38–42 weeks' gestation) at University College Hospital in 1979–1980 were enrolled as age-matched controls for assessments made on the cohort at 4 years. Ten members of this group were recruited again at ages 18 to 19. Another 57 individuals were recruited by advertisement in the local (South London) and national press. Exclusion criteria were history of low birth weight or preterm birth and history of neurologic illness (including significant head injury, meningitis, and epilepsy). The term comparison group thus comprised 67 individuals.
EPQ Assessments
At assessment, each participant filled in the EPQ-RS, a 48-item self-report instrument.19 It consists of 48 questions (eg, "Does your mood ever go up and down?"; "Do you enjoy cooperating with others?") to which the respondent is asked to answer "yes" or "no." The participants were asked to answer in accordance with their first impulse and not to consider the questions in depth. Four factors were derived: extraversion includes such attributes as sociability, liveliness, assertiveness, being active, and sensation seeking; neuroticism includes anxiety, low mood, feelings of guilt, and low self-esteem: psychoticism describes characteristics of coldness, aggression, cruelty, and predisposition to antisocial behavior rather than the current concept of the disturbance of mind found in disorders such as schizophrenia. The fourth scale, lie, measures dissimulation and the tendency for respondents to adjust their responses so that they are more socially acceptable. More details of the EPQ-RS dimensions are described in Table 1. The EPQ-RS may overlap to some extent with childhood behavior–rating scales such as the Child Behavior Checklist. However, only the broad distinction between "externalizing" and "internalizing" behaviors may be directly comparable. For example, higher extraversion and psychoticism might be expected to relate to externalizing problems measured by the Child Behavior Checklist.24
|
Behavioral Assessments in Adolescence
In the assessment at 14 to 15 years, 2 measures of psychopathology and social functioning were administered:
Anthropometric Data
As part of the assessment at 18 to 19 years, individuals had their height, weight, and occipitofrontal circumference recorded.
Statistical Analysis
Statistical analysis was performed by using SPSS 10.0 (SPSS Inc, Chicago, IL). Demographic details were examined by t test or 2 analyses as appropriate. Between-group analyses of covariance were used to control for the effects of age at assessment and social class on the EPQ-RS. Relationships between perinatal, behavioral, and anthropometric data were examined by using Kendall partial correlations, controlling for age at assessment and socioeconomic status.
Ethics
Approval was obtained from the Joint University College London/University College Hospital Committee on the Ethics of Human Research and the Joint Medical Ethical Committee of the Institute of Psychiatry. Informed, written consent was obtained from all participants.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
2 = 16.26; df = 5; P = .006). We therefore controlled for these 2 variables in the following analyses. VPT individuals scored significantly higher on the SAS at 14 years than term individuals (t = –2.72; df = 107; P = .008) and were also rated as having more behavior problems on the Rutter Parents' Scale, although this did not reach statistical significance (t = –1.71; df = 104; P = .091). Significantly more VPT individuals than term individuals had required some degree of extra educational assistance (2 = 7.61; df = 1; P = .006). Similar proportions of the 2 groups were at school, in further education, employed, and unemployed (see Table 2).
|
|
|
Females
VPT females had significantly reduced extraversion scores (F = 6.76; df = 3; P < .001) and increased neuroticism scores (F = 3.49; df = 3; P = .019) compared with term females. There were no significant differences between VPT and term females in psychoticism (F = 0.39; df = 3; P = .763) or lie (F = 1.06; df = 3; P = .373) scores.
Perinatal, Behavioral, and Anthropometric Correlates of EPQ-RS in the VPT Group
Kendall partial correlation analyses, controlling for age at assessment and socioeconomic status, were used to examine these relationships. Gestational age was not significantly related to EPQ-RS scores. Birth weight was negatively correlated with lie score (r = –0.23; P = .051) but not with any other EPQ-RS subscales. There were no significant relationships between EPQ-RS scores and other neonatal data (Apgar score, pH, time to spontaneous respiration) and no correlation with neonatal ultrasound appearances.
There were no significant relationships between educational performance (numbers of General Certificate of Secondary Education or "A"-level examination passes graded A–C) and EPQ-RS scores.
Increased SAS scores (at 14–15 years), which indicate poorer social functioning, were significantly correlated with increased lie scores. Increased Rutter Parents' Scale scores (at 14–15 years) were significantly correlated with increased neuroticism scores. Smaller head circumference at 18 to 19 years was associated with lower extraversion and increased lie scores. Height and weight at 18 to 19 years were not significantly associated with EPQ-RS scores. These results are shown in Table 5.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Our results are reminiscent of those of Hack et al,15 who administered self-report questionnaires that explored risk-taking, antisocial behavior and general health status to a group of VLBW 20-year-olds and a matched comparison group. The VLBW group reported significantly fewer risk-taking and antisocial behaviors than the control group. Our results are also consistent with Hack et al15 in that the differences between the groups are almost all accounted for by the females. Young adult females in the VLBW group studied by Hack et al admitted significantly less alcohol and marijuana use than their term peers. VLBW females were less likely than controls to have had sexual intercourse, to have become pregnant, or to have given birth by the age of 20. VLBW males reported significantly less truancy and alcohol or drug offenses and also less contact with police than their normal–birth weight peers. We found an increase in lie scores in the VPT group, which suggests that their responses were being adjusted to produce a more favorable answer. This was most prominent among the males, and such an effect on responses might potentially bias the findings of Hack et al, tending to lead to an underestimate of police contact and illegal behavior in this group.
The etiology of personality differences is likely to be complex and multigenic, with interactions between genetic and environmental influences. Potential influences include not only early brain injury but also genetic makeup, rearing, parental style, and peer relations. Hack et al suggested that the behavioral "cautiousness" that they found in VLBW 20-year-olds was partly a result of increased parental monitoring, which they postulated to be experienced by children who were born prematurely. Reti et al32 showed that individuals who experienced reduced levels of parental care and higher parental behavioral restrictiveness have alterations of adult personality trait scores, including increased neuroticism, decreased conscientiousness, reduced self-directedness, and increased harm avoidance. However, this study relied on the memory of the subjects (mean age: 51 years), which made it vulnerable to the possibility of recall bias. In the case of VPT infants, it is possible that early worry about the health of a very immature infant could develop into a longer-term pattern of increased parental concern. However, the contribution of such "environmental" factors as parenting style to adult personality may be small: studies of twins reared apart suggest that parenting may only account for a relatively small amount of the variance in personality.33
Normal personality has a heritable component: for example, extraversion and neuroticism traits are estimated to have a heritability of between 35% and 50%.34 Macaskill et al,35 who studied Australian twins, found evidence for a genetic component in psychoticism and neuroticism. Several genetic polymorphisms have been shown to be associated with adult personality. For example, Jorm et al36 showed that a polymorphism of the promoter region of the serotonin transporter gene is associated with anxiety traits in adolescents. Complex traits such as personality are likely to arise from an interaction between multiple genetic and environmental factors.37 There have been some recent demonstrations of such gene-environment interactions. For example, Caspi et al38 found that an interaction between a particular monoamine oxidase A genotype and childhood maltreatment contributes to development of antisocial behavior. Little is known about the possibility of such interactions in VPT individuals.
There is also a link between maternal antenatal well-being and pregnancy outcome. A recent study by Dole et al39 explored a variety of psychosocial factors in a large-population cohort of healthy mothers-to-be. They reported that preterm birth was more frequent in women with high levels of pregnancy-associated anxiety and in those who had experienced recent negative life events or racial discrimination. They did not find that a supportive social environment mitigated these effects. Because anxiety proneness is related to personality traits (such as Eysenck's neuroticism), the transmission of such traits might contribute to the increased neuroticism scores that we demonstrate in preterm individuals. In contrast, Schatz et al40 did not find maternal personality to be significantly related to the occurrence of preterm birth or low birth weight in their study.
Many studies have shown reduced academic performance5–9 and clumsiness and motor problems1–3 in VPT and VLBW children and adults.15 Such deficits may make premature children more likely to experience bullying at school, which could be associated with reduced self-esteem and self-confidence. This possibility is supported by the finding that VPT adults tend to be of shorter stature than their term-born peers.15 In addition, Saigal et al41 reported that extremely low birth weight adolescents did not differ greatly from term-born controls in self-esteem, but heavier individuals reported closer friendships. There is also evidence that men who were small at birth are less likely to marry, although the mechanism underlying this effect remains unclear.42 Our findings are consistent with this in that we find reduced head size to be related to lower extraversion and higher lie scores, although we found no relationship with short stature in adulthood. Saigal et al43 found that extremely low birth weight teenagers rate their health-related quality of life as significantly lower than do their term-born peers. Our results are compatible with this finding, although we did not use a health-related quality-of-life measure.
Outcomes that will be of particular interest to VPT young adults and their parents include mood disorders, risk-taking behaviors, and academic underperformance. Our results and those of Hack et al15 would suggest that VPT teenagers and young adults are more likely to develop personality styles that are less associated with risk taking. However, their risk of mood disorders (particularly anxiety and depression) may be increased by personality factors such as neuroticism. The EPQ does not explicitly include information about unsafe sexual practices, drug use, or alcohol abuse, so our conclusions about this are necessarily speculative.
The unequal socioeconomic status distribution of the 2 groups is a possible confounding factor in this study. We attempted to control for this statistically in the analyses. The different social-class distribution highlights some of the difficulties that are common to research of this type. It is known that VPT birth is more common in groups at socioeconomic disadvantage. It is also possible that controls recruited via newspaper advertisements are a biased sample, being more likely to have a higher level of education and to come from relatively more affluent backgrounds. Another limitation of this study is the different numbers in the VPT and term groups, which may have limited the power of the study. With the unequal sample sizes (108 in the VPT group and 67 in the term group), the power to detect a difference in EPQ-RS psychoticism scores at a significance level of P < .05 was 80%. If equal numbers of VPT and term individuals were assumed, the estimated power to detect this difference rose to 88%. It therefore seems unlikely that the unequal numbers in the study groups has significantly limited the power of this study. Another potential source of bias is the fact that the VPT group is likely to have excluded the individuals with the most impairments, who are more likely to drop out of follow-up (although we attempted to minimize this effect by offering appointments in the subjects' own homes whenever requested). However, the effect of any such bias ought to have been to diminish the strength of our findings rather than cause false-positives. The relationships between EPQ-RS scores and other variables are correlational only and do not prove a causal link. We did not correct for multiple comparisons. These results, therefore, should be interpreted with caution and regarded as preliminary.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
We thank all the young people who generously gave their time to take part in this study.
|
FOOTNOTES |
|---|
Address correspondence to Matthew Allin, MRCPsych, Section of General Psychiatry, Division of Psychological Medicine, Box 063, Institute of Psychiatry, De Crespigny Park, London SE5 8AF, United Kingdom. E-mail: matthew.allin{at}iop.kcl.ac.uk
This work was presented in part in abstract form at the Biennial Winter Workshop on Schizophrenia; February 3–7, 2004; Davos, Switzerland.
The authors have indicated they have no financial relationships relevant to this article to disclose.
This work was presented in part in abstract form at the Biennial Winter Workshop on Schizophrenia; February 3–7, 2004; Davos, Switzerland.
The authors have indicated they have no financial relationships relevant to this article to disclose.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
32 weeks of gestation: a 5-year prospective follow-up.
Dev Med Child Neurol. 1998;40
:21
–30[ISI][Medline]This article has been cited by other articles:
 |
|
 |
|
  D. Moster, R. T. Lie, and T. Markestad Long-Term Medical and Social Consequences of Preterm Birth N. Engl. J. Med., July 17, 2008; 359(3): 262 - 273. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Kajantie, P. Hovi, K. Raikkonen, A.-K. Pesonen, K. Heinonen, A.-L. Jarvenpaa, J. G Eriksson, S. Strang-Karlsson, and S. Andersson Young Adults With Very Low Birth Weight: Leaving the Parental Home and Sexual Relationships--Helsinki Study of Very Low Birth Weight Adults Pediatrics, July 1, 2008; 122(1): e62 - e72. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. A. Schmidt, V. Miskovic, M. H. Boyle, and S. Saigal Shyness and Timidity in Young Adults Who Were Born at Extremely Low Birth Weight Pediatrics, July 1, 2008; 122(1): e181 - e187. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2006 American Academy of Pediatrics. All rights reserved. |
||
|
||
