Abstract
Objectives: The goal of this study was to compare aerobic capacity and exercise performance of children and adolescents born extremely preterm and at term, and to relate findings to medical history and lifestyle factors. Potential cohort effects were assessed by studying subjects born in different decades.
Methods: Two area-based cohorts of subjects born with gestational age ≤28 weeks or birth weight ≤1000 g in 1982–1985 and 1991–1992 and matched control subjects born at term were compared by using standardized maximal treadmill exercise and pulmonary function tests. Background data were collected from questionnaires and medical records.
Results: Seventy-five of 86 eligible preterm subjects (87%) and 75 control subjects were assessed at mean ages of 17.6 years (n = 40 + 40) and 10.6 years (n = 35 + 35). At average, measures of aerobic capacity for subjects born preterm and at term were in the same range, whereas average running distance was modestly reduced for those born preterm. Leisure-time physical activity was similarly and positively associated with exercise capacity in preterm and term-born adolescents alike, although participation was lower among those born preterm. Neonatal bronchopulmonary dysplasia and current forced expiratory vol in 1 second was unrelated to exercise capacity. Differences between subjects born preterm and at term had not changed over the 2 decades studied.
Conclusion: Despite their high-risk start to life and a series of potential shortcomings, subjects born preterm may achieve normal exercise capacity, and their response to physical training seems comparable to peers born at term.
- BPD —
- bronchopulmonary dysplasia
- FEV1 —
- forced expiratory vol in 1 second
- VCo2 —
- carbon dioxide production
- Vemax —
- maximally attainable minute ventilation
- peak V̇o2 —
- maximally attainable oxygen consumption
What’s Known On This Subject:
Extreme preterm birth is associated with developmental shortcomings that may reduce exercise capacity and participation in physical activities in later life. The number of studies addressing these issues in adolescent populations is limited, test methods differ, and results are diverging.
What This Study Adds:
Exercise capacity after preterm birth was in the same range as in term-born control subjects. Participation in physical activity was lower in preterm subjects compared with control subjects; however, the response to exercise in terms of increased aerobic capacity was similar.
Although survival after preterm birth has increased considerably over the past decades, lifelong health expectancies have yet to be described in this special population.1–3 Survival rates have particularly increased for the most immature infants, opening for long-term cohort effects because of the extreme vulnerability of these subjects.
Play and sports are fundamental elements of a normal childhood and are also important for subsequent long-term development.4,5 Individual involvement and success in this arena are influenced by the ability to endure physical activity, often referred to as the exercise capacity of the child. A generally accepted and stringent definition of exercise capacity is difficult to extract from the literature but may be described as the amount of physical exertion someone can sustain in a given activity. Exercise capacity may be limited by the aerobic capacity of the individual, commonly expressed as the maximally attainable oxygen consumption (peak V̇o2), which is the highest rate at which someone can consume oxygen and produce aerobic energy during exercise.6,7 Individual exercise capacity also depends on factors other than the aerobic capacity per se, however, such as neuromotor skills, attitudes, endurance, and sensory and cognitive functions as well as experience with the particular activity in question.8
In subjects born preterm, one may envision several developmental shortcomings that may interfere with exercise capacity, ranging from airway obstruction and disturbed alveolar development9,10 to neuromotor, muscular, sensory, or cognitive deficiencies.3,11–14 Physical inactivity due to real or perceived limitations or weaknesses may by itself reinforce the negative attitude toward exercise, with long-lasting negative effects on exercise capacity.15–17
The purpose of this study was to assess exercise capacity in 2 birth cohorts of subjects born extremely preterm in the past 2 decades and in comparable subjects born at term. Findings were related to selected perinatal and medical variables and to lifestyle factors. Potential cohort effects were assessed by studying effects of decade of birth.
Methods
Subjects and Background Data
Two population-based cohorts of young people born at gestational age ≤28 weeks or with birth weight ≤1000 g in the years 1982–1985 and 1991–1992 in Western Norway were examined in 2002. Detailed information on these subjects has been reported elsewhere18 and is summarized in Table 1. Medical care had been provided at the only neonatal intensive care unit in the region (Haukeland University Hospital), the senior staff being largely the same in both periods. For the periods 1982–1985 and 1991–1992, the mortality rates of infants admitted to the neonatal intensive care unit were 39% versus 27% (P = .157), respectively; treatment rates with antenatal steroids were 38% versus 44% (P = .642), respectively; and treatment rates with postnatal steroids for chronic lung disease were 8% versus 29% (P = .019), respectively. Surfactant was available only to the 1991–1992 cohort and was administered to 18 survivors (51%). Self-reported smoking was similarly distributed between preterm and term-born adolescents, that is, 33% versus 25%, respectively (P = .622). Bronchopulmonary dysplasia (BPD) was defined as mild if oxygen treatment was required at 28 days of age and as moderate/severe if oxygen treatment was still required at a postmenstrual age of 36 weeks, giving 3 strata of BPD: none, mild, and moderate/severe.19
Neonatal Characteristics of the Preterm Subjects Completing the Exercise Test
For each child born preterm, the subsequent birth of the same gender and with a birth weight between 3 and 4 kg (Norwegian 10–90 centiles) was selected as a term-born control subject. If that subject declined participation in the study, the next-born subject was asked, and so on.
Medical history and past and current diagnoses and treatments were obtained at a clinical examination performed by a senior pediatrician (T.H.) and from hospital records. Blindness, deafness, quadriplegic cerebral palsy, or serious psychiatric disorders were classified as major disabilities. Disabilities not preventing attendance at regular schools were classified as minor.
The regional ethics committee approved the study. Informed written consent was obtained from participating subjects and parents.
Demographic Variables
A questionnaire was used to register general background variables. Two validated questions served to determine the level of leisure-time physical activity20–22: (1) Apart from at school, how often do you usually exercise so much that you get out of breath or sweat? and (2) apart from at school, how many hours a week do you usually exercise so much that you get out of breath or sweat? Parental answers were used in the youngest cohort.
Pulmonary Function
Pulmonary function was measured with SensorMedics Vmax 22 spirometer (Yorba Linda, CA), applying standard quality criteria.23 Forced expiratory vol in 1 second (FEV1) expressed as percentage of predicted24 was used to characterize airway function and to calculate maximal voluntary ventilation.
Exercise Test
An incremental treadmill (ELG 70; Woodway, Weil am Rhein, Germany) test was applied, using a modified Bruce protocol, identical in all subjects.25 Speed and elevation were gradually increased every 90 seconds from an initial slow-walking phase. After baseline variables had been established, subjects ran to exhaustion wearing a face mask connected to a Vmax 29 cardiopulmonary exercise unit (SensorMedics, Yorba Linda, CA). The test was stopped when the subject indicated exhaustion, preferably supported by a plateau in oxygen consumption or heart rate response.6,26 Variables of gas exchange and airflow were measured breath-by-breath, averaged over 10 seconds, and the highest values determined during the last 60 seconds were taken as maximal values. Oxygen uptake (peak V̇o2), carbon dioxide production (peak VCo2), tidal vol, respiratory rate, and heart rate were measured directly, whereas maximum minute ventilation (Vemax) was calculated from tidal vol and respiratory rate. Peak respiratory exchange ratio (respiratory quotient) was calculated from peak VCo2/peak V̇o2. Anaerobic threshold was obtained from the inflection on the VCo2 / Vo2 plot, with support from the Ve/VCo2 and Ve/V̇o2 plots.6 Calculation of ventilatory reserve at maximum exercise was based on the assumption that FEV1 × 35 was the Vemax.27 Exercise performance in the test situation was described by the “distance completed on the treadmill” (meters), measured directly through the computerized protocol.
Statistical Methods
Means and SDs were estimated. Group comparisons were performed with paired-samples t tests for preterm subjects versus matched control subjects; otherwise, comparisons were performed with unpaired-samples t tests, one-way analyses of variance tests, or χ2 tests, as appropriate. The main outcome variable for aerobic capacity was peak V̇o2 (mL ⋅ kg−1 ⋅ min−1), and for exercise performance, the “distance completed on the treadmill” (meters). Adjusted for gender, effects from the following prenatal and postnatal variables were assessed, one by one, using linear regression: gestational age at birth, birth weight, duration of ventilator treatment (days), duration of neonatal oxygen treatment (days), antenatal maternal treatment with corticosteroids, postnatal treatment with corticosteroids, postnatal septicemia, and surfactant treatment (only 1991–1992 cohort). A multiple-linear regression model containing all variables (except surfactant treatment) was constructed. Participation in leisure-time physical activities was handled as ordinal categorical variables and related to measures of exercise capacity with linear regression models (adjusted for gender). Effects from explanatory variables in various subgroups were assessed with interaction terms,28 using the mixed-linear model of SPSS (SPSS, Chicago, IL) if the matched structure of the study could be used (paired tests of interaction) and multiple-linear regression if it could not be used. Based on previous data from unpaired groups,29 the study had 80% power to detect differences between preterm and control groups for peak V̇o2 (mL⋅ kg−1 ⋅ min−1) of approximately 4, providing a 5% significance level.
Results
Subjects
Of 130 preterm subjects admitted to the neonatal department during the 2 inclusion periods, 86 (66%) were alive at follow-up. Five declined participation, and 81 (94%) were included; 46 (90%) from the 1982–1985 cohort and 35 (100%) from the 1991–1992 cohort. All but 2 were white. On average, 1.3 potential control children were approached to find 1 willing match for each preterm child. Average age at examination in preterm and control subjects was 17.5 (SD: 1.1) and 17.7 (SD: 1.2) years, respectively, versus 10.5 (SD: 0.4) and 10.6 (SD: 0.4) years in the 1982–1985 and 1991–1992 birth cohorts, respectively.
None of the control subjects had disabilities. In the 1982–1985 preterm cohort, 6 (13%) had minor and 8 (17%) had major disabilities (4 subjects had disabling cerebral palsy, 3 were blind, 3 were deaf, and 2 had severe psychiatric disorders). In the 1991–1992 cohort, 7 (20%) had minor disabilities, and none had major disabilities. All control subjects completed all tests. All preterm subjects from the 1982–1985 cohort produced satisfactory spirometry test results, whereas 5 were unable to run: 4 had disabling cerebral palsy and 1 was diagnosed with a psychiatric disorder. Additionally, 1 preterm subject from the 1982–1985 cohort was excluded because of a submaximal test result (respiratory quotient = 0.86 and peak V̇o2 = 11.8 mL⋅ kg−1⋅ min−1). Only complete preterm-control pairs were analyzed, so 150 subjects (75 + 75) contributed, representing 87% of eligible and 93% of participating preterm subjects. Neonatal characteristics are presented in Table 1.
Exercise Capacity
All subjects ran to perceived exhaustion. Respiratory quotient ≥1.05 or maximal heart rate ≥95% predicted was achieved by all control subjects and by 88% of preterm subjects in the 1982–1985 cohort and also by 91% of control subjects and 77% of preterm subjects in the 1991–1992 cohort. Differences in aerobic capacity between the preterm and term-born cohorts were minor and not statistically significant for variables adjusted for body weight (Table 2). There were no significant decreases for average peak V̇o2 (mL ⋅ kg−1⋅ min−1) or anaerobic threshold. Average treadmill distance was ∼10% shorter in the preterm compared with the term-born cohorts, and these differences were statistically significant. Differences in treadmill distance increased if adjusted for variations in body size (weight, height, and BMI), except for height in the 1982–1985 cohort, which reduced the difference from 97 to 58 m (linear regression, P = .232).
Variables of Exercise Capacity in Preterm Subjects and Matched Term-Born Control Subjects
Excluding subjects with respiratory quotient <1.05 or maximal heart rate <95% predicted, average peak V̇o2 (mL ⋅ kg−1⋅ min−1) was 47.1 versus 48.3 (P = .406) for the preterm versus the term-born cohort. In the 1982–1985 preterm cohort, peak V̇o2 (mL ⋅ kg−1⋅ min−1) was 47.6 for subjects without disabilities (n = 32) compared with 43.0 (P = .831) for subjects with minor disabilities (n = 5) and 47.1 (P > .999) for subjects with major disabilities (3 male subjects who were deaf). In the 1991–1992 cohort, corresponding figures were 44.5 and 39.3 for subjects without (n = 28) and with (n = 7) minor disabilities, respectively (P = .135).
Exercise Capacity in Relation to Birth Cohort and Gender
For all assessed measures of exercise capacity, differences between preterm and matched term-born control subjects were similar for the 2 birth cohorts and also when tested relative to cohort-related differences in outcome measures (tests of interaction and independent sample t tests, data not shown). Male subjects born at term had significantly higher mean peak V̇o2 (mL ⋅ kg−1⋅ min−1) than male subjects born preterm (53.7 [SD: 8.4] versus 49.3 [SD: 8.5], P = .043), whereas there was no such difference for female subjects (42.9 [SD: 6.2] versus 42.0 [SD: 7.2], P = .555). This apparent effect from gender was not statistically significant, however (paired test of interaction, P = .164).
Exercise Capacity in Relation to Neonatal Variables
For subjects born preterm, peak V̇o2 (mL ⋅ kg−1 ⋅ min−1) was unrelated to all examined neonatal variables when assessed separately by linear regression (adjusted for gender) and in a multiple-linear regression model. The distance completed (meters) was positively associated with gestational age (P = .016) and negatively associated with postnatal treatment with corticosteroids for BPD (P = .002) in separate linear regression analyses (adjusted for gender). Only male gender (β = 247 m, P < .001) and postnatal treatment with corticosteroids (β = −187 m, P = .013) remained significant in a multiple-regression model including all assessed neonatal variables.
Differences between preterm and matched control subjects for peak V̇o2 (mL ⋅ kg−1 ⋅ min−1) and “distance completed” (meters) did not increase over the 3 BPD strata (paired tests of interaction, P = .502 and P = .887) (Fig 1). For peak V̇o2 (% predicted), mean values over the BPD strata of none, mild, and moderate/severe were 110%, 104%, and 95%, respectively. This numerical trend was not significant, however (paired test of interaction, P = .188).
Exercise capacity in relation to severity of neonatal BPD. The figure presents error bars, describing mean values and 95% confidence intervals. M/S, moderate/severe.
Aerobic Capacity in Relation to Level of Activity
Extracurricular physical activity 2 to 3 times per week or more was reported by 34% of preterm subjects and 72% of control subjects (χ2 test, P < .001), and extracurricular physical activity 2 to 3 hours per week or more was reported by 36% of preterm subjects and 59% of control subjects (P = .004) (Tables 3 and 4). Adjusted for gender, increased activity on these 2 measures of physical activity was positively associated with peak V̇o2 (mL ⋅ kg−1 ⋅ min−1) and the number of meters completed on the treadmill in the 1982-1985 cohort (linear regression analyses, P < .001) but not in the 1991–1992 cohort. As illustrated in Fig 2, these relations were similar in preterm and control subjects (tests of interaction [adjusted for gender]; hours per week versus peak V̇o2, P = .748, and hours per week versus number of meters, P = .271). Neonatal history, FEV1, and ventilatory reserve capacity were not significantly different over the different levels of physical activity (Table 5).
Number of Times Per Week With Extracurricular Physical Activity
Number of Hours Per Week With Extracurricular Physical Activity
Exercise capacity in relation to the number of hours per week of extracurricular physical activity. Estimated marginal means are mean values at each point, adjusted for gender differences. The number of subjects at each level of exercise was (left to right, x axis): 18, 17, 18, 10, 11, and 6 (1982–1985 cohort) and 5, 4, 17, 23, 12, and 9 (1991–1992 cohort).
Neonatal Characteristics and Current Lung Function for Subjects Born Preterm, Listed According to Participation in Physical Activity
Aerobic Capacity in Relation to FEV1
FEV1 (% predicted) was not significantly related to peak V̇o2 (mL ⋅ kg−1 ⋅ min−1), that is, R2 = 0.022 and P = .070. Split into quartiles, no associations with peak V̇o2 could be demonstrated (Fig 3).
FEV1% predicted by quartiles versus aerobic capacity in subjects born preterm. The figure presents error bars, describing mean values and 95% confidence intervals for peak V̇o2 by quartiles for FEV1% predicted for preterm participants of the study. The first (lowest) quartile covers subjects with values of FEV1 classified as subnormal (below the 5th centile) according to Quanjer et al,24 who published the applied reference equation.
Discussion
In this area-based study, exercise capacity of subjects born extremely preterm was normal and in the same range as matched control subjects born at term. Leisure-time physical activity was similarly and positively associated with exercise capacity in preterm and term-born adolescents alike, although participation was lower among those born preterm. Neonatal BPD and current airway obstruction were unrelated to exercise capacity. Differences between subjects born preterm and at term had not changed over the 2 decades studied.
The major strengths of this study were the area-based design, the nearly complete participation, and the high rate of successful exercise tests. Importantly, control subjects were individually matched with each preterm subject with respect to age and gender but otherwise randomly selected, and with few exceptions, the first subject approached agreed to participate. The risk of systematic bias in the control population was low, and also, paired statistical analyses could be applied. These features are not attended to in some previous publications.14,29–31 The same team conducted all parts of the study, limiting interobserver variability. A limitation of the study was the overall number of participants; however, statistical power was regarded as acceptable.
Previous studies on exercise capacity after preterm birth report diverging results.14,29–36 Such disparity may certainly reflect real differences between the populations studied; however, the test set-ups applied also may be of relevance, because children born preterm differ from those born at term on a range of abilities that may influence results.12–14 For example, cognitive or sensory impairments may strongly influence results from complicated test set-ups,7,31 and not all children are familiar with cycling. Walking and running are familiar to most people and also highly relevant for daily life.8 Treadmill exercise, which most participants completed satisfactorily, was therefore chosen as the mode of exercise in this study.
The only neonatal exposure with significant impact on current exercise performance was postnatal treatment with corticosteroids, a treatment form that has been discussed also in relation to other unfavorable outcomes in this study group.37 Interestingly, neonatal BPD was unrelated to all measures of current exercise capacity, contrasting previous reports on lung function9,18; however, studies of this size, and particularly subgroup analyses, are prone to statistical errors and must be interpreted with caution. Although mean FEV1 was reduced in subjects born preterm, it was unrelated to exercise capacity. Airflow limitation did not seem to be a limiting factor for exercise capacity.
We know that generally, aerobic capacity is related strongly to age, maturity, gender, and body size, although these relationships are complex.7 Conceivably, large bodies should have larger muscle mass and greater cardiac output and ventilatory capacities than small bodies; however, individually these relationships may not be optimally balanced. In subjects born preterm, these issues may be further complicated by neuromuscular, cardiopulmonary, cognitive, or regulatory alterations, potentially influencing and jeopardizing aerobic capacity and exercise performance. In the current study, most weight-adjusted measures of aerobic capacity were normal in preterm subjects and not significantly different from those of term-born control subjects. Average treadmill distance, however, was reduced slightly in preterm subjects. This reduction was not explained by variations in body size per se, although height seemed to attenuate the difference in the oldest birth cohort, possibly related to a relatively large age span in this group. The complexity of these issues and numerous possible interactions prevent a more thorough discussion within the frame of this study.
Our findings are in line with some studies30,32,35,38 but at variance with 3 particular recent studies.29,31,36 Smith et al31 reported that exercise capacity in 10.1-year-old preterm subjects was reduced to less than half of that of a control group. The 20-m shuttle-run test applied in that study is demanding regarding a range of abilities typically influenced by preterm birth.3,7,12,39 Results may therefore be misleading and conclusions limited to this particular test scenario.7 Further uncertainty was introduced by an increased age (1.5 years) and a male preponderance in the control group, and a low rate of participation in the preterm group (21% of those invited). The 2 other studies showing major decreases involved children who were born particularly immature, that is, with birth weight <801 g29 or gestational age <26 weeks.36 Thus, children born at the very threshold of viability may be at particular risk for reduced exercise capacity, possibly due to obstructive airway disease, acinar simplification, or alveolar hypoplasia, features that have been identified as concerns in this population.9,10
Inconsistent findings between studies also may be related to variations in lifestyle, particularly among adolescents. In the current study, leisure-time physical activity was positively associated with exercise capacity in adolescents but not in the 10-year-olds. All participating subjects attended public schools with mandatory programs for physical education, so variations in leisure-time physical activity probably reflected total variation in physical activity reasonably well.22 Most young children in our country participate in some sport or leisure-time physical activity, reducing interindividual differences in that age group. Drop-out rates from these activities start to increase in late childhood, increasing differences between individuals.40 A possible relationship between physical activity and exercise capacity may be more difficult to detect in young children than in adolescents. This line of reasoning fits well with what was actually observed in the current study. Regarding subjects being born preterm, at least 2 scenarios are possible: (1) only those who are able to participate in sports continue doing so into adolescent life, thereby improving their exercise capacity, or (2) alternatively, all children born preterm may improve their exercise capacity, provided they continue to exercise. Our data support the latter theory because the neonatal history and current lung function was similar over the various levels of physical activity, and ventilatory reserve capacity did not seem to be a limiting factor. Supporting Kajantie et al,17 we found that subjects born preterm exercised less than peers born at term; however, the response to physical activity in terms of an increasing peak V̇o2 seemed to be similar. Indeed, active preterm adolescents performed better than sedentary control subjects born at term. These results regarding the effects of continued and adequate physical activity for these vulnerable children are promising.
Physically active adolescents tend to become physically active adults, which is positively associated with health. Preterm children are faced with a multitude of factors that may jeopardize lifelong health and well-being. We argue that an active lifestyle should be particularly encouraged in this group.
Conclusions
Despite their high-risk start to life and a series of potential shortcomings, subjects born preterm may achieve normal exercise capacity. Leisure-time physical activity was similarly and positively associated with exercise capacity in adolescent preterm and term-born subjects alike, although participation was lower in those born preterm. Parents of “neonatal intensive care unit graduates” should actively encourage their children to participate in regular childhood physical activities and sports.
Acknowledgments
Major funding for this article was received from Haukeland University Hospital and the University of Bergen. Minor support was received from the Pediatric Lung Research Fund, Bergen.
Footnotes
- Accepted September 14, 2011.
- Address correspondence to Thomas Halvorsen, Department of Pediatrics, Haukeland University Hospital, N-5021, Bergen, Norway. E-mail: thomas.halvorsen{at}helse-bergen.no
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
References
- Copyright © 2012 by the American Academy of Pediatrics