ARTICLE |
a Departments of Psychology
d Public Health, University of Helsinki, Helsinki, Finland
b National Public Health Institute, Helsinki, Finland
c Hospital for Children and Adolescents, Helsinki University Central Hospital, Helsinki, Finland
| ABSTRACT |
|---|
|
|
|---|
METHODS. The study was a retrospective longitudinal study of 158 young adults born with very low birth weight and 169 term-born control subjects (aged 18.5–27.1 years). The principal outcome variable was sleep-disordered breathing defined as chronic snoring.
RESULTS. The crude prevalence of chronic snoring was similar in both groups: 15.8% for the very low birth weight group versus 13.6% for the control group. However, after controlling for the confounding variables in multivariate logistic regression models (age, gender, current smoking, parental education, height, BMI, and depression), chronic snoring was 2.2 times more likely in the very low birth weight group compared with the control group. In addition, maternal smoking during pregnancy was significantly and independently of very low birth weight related to risk of sleep-disordered breathing. Maternal preeclampsia, standardized birth weight, and, for very low birth weight infants, small-for-gestational-age status were not related to sleep-disordered breathing.
CONCLUSIONS. Premature infants with very low birth weight have a twofold risk of sleep-disordered breathing as young adults. In addition, maternal smoking during pregnancy increases the risk of sleep-disordered breathing by more than twofold.
Key Words: sleep-disordered breathing snoring prematurity very low birth weight
Abbreviations: OSAS—obstructive sleep apnea syndrome VLBW—very low birth weight AGA—appropriate for gestational age SGA—small for gestational age BDI—Beck Depression Inventory CI—confidence interval aOR—adjusted odds ratio
Sleep-disordered breathing is characterized by intermittent upper airway breathing obstructions that lead to hypoventilation, disrupted sleep architecture, and decrease in blood oxygen saturation.1 The term "sleep-disordered breathing" refers to deviant nocturnal respiratory functioning in general, and it includes both obstructive sleep apnea syndrome (OSAS) and upper airway resistance syndrome, but it is not limited to these conditions.2 Milder forms of disordered breathing can manifest, for example, as chronic snoring. Both snoring and objectively defined OSAS have been linked with many somatic, psychiatric, and cognitive problems, such as increased risk of cardiovascular morbidity, impaired neurocognitive performance, and psychiatric symptoms.3–8
In adulthood, sleep-disordered breathing is closely associated with obesity, whereas the most important risk factor in childhood is adenotonsillar hypertrophy.1 One previous study has suggested that preterm infants are more likely to have sleep-disordered breathing as children,9 but whether this difference is sustained in adulthood is not known. The question has recently gained actuality, because prematurely born children with sleep-disordered breathing may be at heightened risk for impairments in neurocognitive development as compared with term-born control subjects with sleep-disordered breathing.10 Interestingly, both prematurity and sleep-disordered breathing have been associated with neurocognitive impairment,4,11 depression,7,12,13 higher blood pressure,6,14–16 and impaired glucose regulation.5,16,17 Sleep-disordered breathing might, thus, mediate or moderate the effects of prematurity on health if prematurity can be confirmed to constitute an independent risk factor for sleep-breathing disorders.
The main objective of the present study was to test whether very low birth weight (VLBW; <1500 g) is related to increased prevalence of sleep-disordered breathing in young adulthood. The second objective of the study was to investigate whether pregnancy-related risk factors, such as maternal smoking during pregnancy, preeclampsia, intrauterine growth retardation, and mechanical ventilation, are related to sleep-disordered breathing. Characterization of risk factors for SBD may facilitate new alternatives to early intervention, as well as provide insights to our understanding of pathophysiological mechanisms of development of sleeping problems.
| METHODS |
|---|
|
|
|---|
37 weeks) and appropriate for gestational age (AGA; birth weight in relation to gestational age and gender by Finnish standards18 more than –2 SD) so that the next eligible infant with the same gender was picked up as a control for each VLBW survivor. Nearly all (95.1%) of the VLBW survivors and control subjects (96.8%) were traced, and those who lived in the greater Helsinki area were invited to participate in the study. There were 314 control and 255 VLBW subjects who fulfilled the inclusion criteria, and of these, 166 VLBW subjects (65.1%) and 172 control subjects (54.8%) were willing to participate. In addition, we excluded 1 VLBW subject with panhypopituitarism, 1 control subject with type 1 diabetes mellitus, and 1 control subject and 1 VLBW subject because of pregnancy. In addition, 6 VLBW subjects and 1 control subject did not fill out the sleep-related questions. Thus, the final sample consisted of 158 young adults with VLBW and 169 term-born young adults. There were 69 men (43.7%) in the VLBW group and 69 men (40.8%) in the control group.
Measurements
Data on infant (birth weight and gestational age) and maternal (smoking during pregnancy and preeclampsia) characteristics, as well as duration of mechanical ventilation while at the perinatal and neonatal wards, were collected from hospital records. Birth weight was transformed into standardized birth weight score according to normal distribution of birth weight in relation to gestational age and gender by Finnish standards.18 The VLBW group was subdivided into 2 groups, AGA and small for gestational age (SGA; birth weight: at or less than –2.0 SD) according to standardized birth weight. Maternal smoking during pregnancy was categorized as no versus yes, and preeclampsia was defined according to the criteria by the National High Pressure Education Program Working Group on High Blood Pressure in Pregnancy.19
Sleep-related breathing disorders were measured with the following 3 questions derived from the Basic Nordic Sleep Questionnaire.20 The first question focused on chronic snoring, "Do you snore while sleeping (ask other people if you are not sure)?" and was rated on a 5-point scale reflecting the frequency and, thus, severity of the disorder, ranging from "never or less than once per month" to "every night or almost every night." Chronic snoring was defined as snoring
1 to 2 times per week versus less. The other questions concerned the type or quality of snoring, "How do you snore (ask other people about the quality of snoring)?" and the frequency of apneas, "Have you had breathing pauses (sleep apnea) at sleep (have other people noticed that you have pauses in respiration when you sleep)?"
The subjects also filled out a questionnaire that covered current smoking status (no versus yes) and maternal and paternal educational attainment of which the higher one was used (1: elementary school, 8.6%; 2: high school, 21.2%; 3: intermediate education, 34.8%; 4: university degree, 35.4%). Depressive symptoms were assessed by using the Beck Depression Inventory (BDI) consisting of 21 items.21 Each item contains 4 statements reflecting varying degrees of symptom severity the past 2 weeks, ranging from 0 to 3, indicating increasing severity. Weight and height were measured, and the questionnaire data were filled out in conjunction with a clinical examination. BMI was calculated as kg/m2.
Statistical Analysis
First, we compared the crude frequency and mean differences of various clinical, prenatal, and background variables between the control group and the VLBW group, as well as participants without and with chronic snoring using the
2 test or the t test. We calculated 95% confidence intervals (CIs) for the main comparisons. Second, we performed a series of logistic regression models to assess whether prematurity is related to chronic snoring in young adulthood while controlling for significant background factors and comorbidity. We controlled for age, gender, current smoking, parental education, height, BMI, and depression measured as logarithmically transformed BDI score, ln(BDI score + 1). The effects of maternal smoking during pregnancy, standardized birth weight, and preeclampsia were assessed in separate models while controlling for the confounders. The effects of intrauterine growth restriction (AGA versus SGA status) and mechanical ventilation were assessed separately among the VLBW subjects only. We also constructed a full model including all of the explaining variables of interest and assessed all of the 2-way interactions between the main explanatory variables and between these variables and gender. Finally we studied whether snoring and VLBW status are related to an elevated depressive symptom score using the same set of confounders as defined previously. All of the significance tests were 2-sided, and the significance level was set at a P value of <.05.
| RESULTS |
|---|
|
|
|---|
|
|
Three subjects in the control group (1.8%) and 4 subjects (2.5%) in the VLBW group reported snoring always or almost always (P = .64). Seven subjects (4.1%) in the control group and 12 subjects (7.6%) in the VLBW group reported snoring more often than 3 days per week (P = .18).
There were no significant differences in chronic snoring between those adults with VLBW born AGA or SGA: the prevalence of chronic snoring was 15.9% (95% CI: 9.0–22.8; n = 17 of 107) in the subjects with VLBW who were born AGA and 15.7% (95% CI: 5.7–25.7; n = 8 of 51) in those born SGA (P = .85). There were no differences in quality or type of snoring between the 2 groups, and sleep apneas were uncommon in both groups: 97.6% (n = 166) of the control subjects and 97.5% (n = 155) of the VLBW subjects reported no sleep apneas.
Furthermore, a history of preeclampsia was not significantly related to chronic snoring (18.8% in those with positive history of preeclampsia versus 14.0% in others; P = .39) in the control group (15.4% vs 13.5%; P = .85) or the VLBW group (20.0% vs 14.6%; P = .44). Chronic snoring was, however, more prevalent in those whose mothers reported smoking during pregnancy compared with those whose mothers did not (26.3% vs 11.3%; P = .003). The effect concerned especially the VLBW group (30.0% vs 11.9%; P = .01), whereas it did not reach statistical significance in the control group (22.2% vs 10.9%; P = .11).
Table 3 shows the adjusted odds ratios (aORs) of chronic snoring in relation to VLBW, preeclampsia, maternal smoking during pregnancy, and standardized birth weight. Thus, when controlling for gender, age, current smoking, parental educational attainment, height, BMI, and current depressive symptoms, the odds for chronic snoring in young adulthood among those born with VLBW were 2.2-fold (95% CI: 1.1–4.5) when compared with the term-born control subjects. The adjusted models also showed that infants with VLBW born AGA or SGA did not differ from each other in chronic snoring when assessed separately in the preterm group; there were no SGA cases in the control group. The aOR for snoring was also significantly increased for those whose mothers smoked during pregnancy (aOR: 2.6; 95% CI: 1.2–6.0). Mechanical ventilation and preeclampsia continued to be unrelated with chronic snoring.
|
Snoring was associated with depressive symptoms when individuals reporting no depressive symptoms (BDI: <10 points) were contrasted with those who reported at least mild depressive symptoms (BDI:
10 points): the aOR in a model including the major confounding variables was 2.91 (95% CI: 1.27–6.65). VLBW status was not related to depressive symptoms (P = .81), and there was no significant interaction between snoring and VLBW status in relation to depressive symptoms (P = .32).
| DISCUSSION |
|---|
|
|
|---|
Previously, only a limited number of studies have assessed the effects of prenatal and perinatal factors on sleep in infancy22 or early childhood.23 Shang et al23 reported that prenatal or perinatal exposure to alcohol, caffeine, or medication was related to increased frequency of various sleeping problems at the age of 4 to 9 years. Moreover, certain pregnancy- or labor-related factors, such as vaginal bleeding and artificial delivery, were related to sleeping difficulties later on. Hoppenbrouwers et al22 reported that premature infants in general do not have alternations in the sleep architecture at the age of 33 to 58 weeks. Specifically, prematurity was not related to increased predisposition to sleep apneas. However, a history of assisted ventilation and being born SGA was related to a delay in the maturation of the sleep-wake architecture and increased proportion of active sleep, respectively.22
Rosen et al9 assessed OSAS in 8- to 11-year-old children and found that the effect of prematurity on OSAS varied depending on the criteria used. The aORs were 5.0 (95% CI: 1.6–20.1) and 2.7 (95% CI: 1.2–6.5) using more (>5 hypopnea events per hour) and less stringent criteria (>1 apnea event per hour), respectively. Thus, these aORs were slightly higher than the aORs that we found for chronic snoring but not beyond the reported CIs. Rosen et al9 also reported that the crude prevalence of chronic snoring, applying exactly the same criterion as in the present study, was higher in the preterm group than in the control group (21% vs 14%, respectively), whereas in our study the difference became evident only after adjustment for confounding factors. Possible explanations for the lower risk ratio that we found in adults include enlargement of airways during growth and active treatment of OSAS. Rosen et al9 reported increased rates of adenotonsillectomy in 8- to 11-year-old preterm children as compared with control subjects (13% vs 3%, respectively).
Rosen et al9 speculated that formerly preterm children may be at increased risk for sleep-disordered breathing because of various adverse exposures during the perinatal period, which may be related to craniofacial development, respiratory control mechanisms, or the upper airway size. Prematurity has, for example, been linked with facial asymmetry and sagittal occlusal relationships, which can alter the airway dimensions.24 On the other hand, preterm birth seems to predispose to components of the metabolic syndrome, such as insulin resistance16,25 and higher blood pressure later in life.14,16 Because it has been proposed that sleep-disordered breathing may not be only an independent disorder but also a manifestation of metabolic syndrome,26 our finding is also in line with this hypothesis. Additional studies are needed to enlighten the underlying mechanisms and to seek other biological explanations.
Preeclampsia is an important risk factor for prematurity, and, thus, it might be independently related to adverse health outcomes in the offspring. In this study, no such effect could be demonstrated. Maternal smoking during pregnancy was, however, independently related to the risk of sleep-disordered breathing. Similar findings were made by OBrien et al,27 who reported that parental smoking is related to snoring at the age of 6 years. Hoppenbrouwers et al22 reported that maternal smoking was related to increased postnatal wake time at 33 to 58 weeks of age in premature born infants. Whether the association reflects environmental or biological factors remains to be evaluated in additional studies. It may constitute an independent risk factor for sleep-disordered breathing if smoking can biologically alter the metabolism of the fetus. Maternal smoking during pregnancy can also indicate a risk factor that continues to postnatal life, that is, exposure to passive smoking, which can predispose to snoring in later life.28 Whether this explains the observed relationship between maternal smoking during pregnancy and snoring in the offspring remains an unanswered question.
In childhood, sleep-disordered breathing constitutes a significant risk factor for impaired neurocognitive performance, behavioral problems, externalizing symptoms, and inattention.3,29–33 The most severe forms of sleep-disordered breathing in childhood can, if untreated, lead to pulmonary hypertension, growth retardation, and failure to thrive. Among adults, sleep-disordered breathing has also been related to impairments in executive functioning, decreased attention, and memory functioning, as well as depression, cardiovascular disease, hypertension, and impaired glucose metabolism.4–7 Part of the changes produced by sleep-disordered breathing may be irreversible, and, thus, early intervention is warranted.3,30 For example, among children with sleep-disordered breathing, adenotonsillectomy reduces behavioral and attentional problems and improves cognitive performance.34,35 The American Academy of Pediatrics suggests that all children should be screened for snoring, and high-risk patients should always be referred to a specialist.36 Because former premature children and adults may be at increased risk for adverse symptoms related to sleep-disordered breathing,10 patients with suggestive symptoms should be raising vigilance to screen for the diagnosis. Moreover, given the widespread secondary consequences of sleep-disordered breathing, studies assessing the effects of prematurity in adulthood, especially with regard to neurocognitive development and cardiovascular morbidity, should take into account the possibility of sleep-disordered breathing. The presence of an underlying and possibly unrecognized sleep-disordered breathing may significantly modify the relationship between prematurity and secondary morbidity.
| LIMITATIONS |
|---|
|
|
|---|
| CONCLUSIONS |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
Address correspondence to E. Juulia Paavonen, MD, PhD, BsocSc, Department of Psychology, PO Box 9, 00014 University of Helsinki, Helsinki, Finland. E-mail: juulia.paavonen{at}helsinki.fi
The authors have indicated they have no financial relationships relevant to this article to disclose.
The sponsors had no role in the design or conduct of the study; collection, management, analysis, or interpretation of the data; or preparation, review, or approval of the article.
| REFERENCES |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||