The Future of Very Preterm Infants: Title and subTitle BreakLearning From the Past

What is the long-term future of infants who are born after very short gestations (<33 weeks) compared with infants born at term? As the article by Swamy and colleagues¹ in this issue of JAMA suggests, the experience of a population of very preterm infants in Norway offers some insights. This retrospective cohort included 1.1 million singleton births in Norway from 1967 through 1988 occurring at 22 or more weeks of gestation and weighing 500 g or more. The investigators assessed the perinatal, childhood, and adolescent mortality of this cohort through 2002 and followed a subset of survivors for educational and reproductive outcomes through 2004. They found that male and female very preterm offspring (born at 22-32 weeks of gestation) had a higher risk of mortality from the perinatal period through age 5 years compared with their term counterparts (born at 37-42 weeks of gestation). The highest mortality rate occurred at delivery and at the shortest gestations. The rate of stillbirths at 22 to 27 weeks of gestation was 53.1%, whereas at term the rate was 0.38%. Survival rates increased from infancy to late childhood.

The investigators also analyzed adult educational attainment and reproduction in this cohort. Compared with term infants, preterm infants had lower educational attainment: the lower the gestational age, the lower the average educational attainment. Men and women in the study who were born very preterm (22-32 weeks) were less likely to reproduce. Among adults who were preterm at birth, only women had an increased risk of having a preterm infant compared with women who were term at birth.

Unlike other studies among preterm infants, Swamy et al used gestational age rather than birth weight to assess long-term and intergenerational outcomes. Using a detailed and complete vital records registry, the authors demonstrate the utility of longitudinal data in understanding life-course influences on the health of Norwegian men and women born in the late 1960s to 1980s.

Despite these strengths, the findings should be interpreted with caution. The Norwegian environment differs from that of the United States. In many respects, Norway is a best-case scenario. The Norwegian cohort and their reproduction occurred in a relatively homogeneous population with universal access to medical care and a well-developed social safety net.

Another consideration is that maternal and neonatal treatment given during the 1970s and 1980s differed from that of the past 2 decades. Any long-term follow-up study, however, will have this limitation. To compensate for it, the potential influence of differences in clinical course and treatment must be considered. One difference between the births occurring during this study and births within the past 2 decades in the United States relates to the proportion of preterm deliveries that were induced and the proportion that were spontaneous. The rate of medically induced preterm deliveries has increased in the United States during the past decade² and in Norway during the 1990s.³ Spontaneous deliveries probably accounted for most of the very preterm births in Norway throughout the 1960s and 1970s. But the cause of these preterm births is unclear. Preterm rupture of membranes can be an important causal mechanism. A portion of preterm deliveries following rupture of membranes is associated with intrauterine infection, which involves a potential risk of sepsis and adverse neurological outcomes.⁴ Compared with infants whose deliveries were induced after 28 to 31 weeks of gestation, infants of comparable gestational age who were delivered spontaneously were more likely to have cerebral palsy.⁵

Swamy et al do not report rates of induction or cesarean delivery among the index participants by gestational length. A prior Norwegian study using births for 1970 through 1988 from the same registry reported a cesarean delivery rate of 6%.⁶ In the 1970s in the United States, the rate of cesarean delivery was 5% to 6%.⁷ By 2005, it was 30.3%.⁸ In the United States, Norway, and other developed countries, preterm deliveries due to induction or cesarean delivery are increasing.^{3 ,9} Reasons for these increases are not completely known but include fetal monitoring, delayed childbearing, assisted reproductive technologies, multiple births, litigation concerns, and, possibly, maternal requests.^{10 - 13} These changes and improving neonatal survival during and after the 1990s likely explain clinicians' increased willingness to induce delivery despite very short gestations.¹⁴

Another difference in clinical practice and outcomes relates to substantial improvements during the past 2 decades in the clinical care of very preterm infants. During the 1970s and 1980s, perinatal regionalization improved the referral and transport of women with imminent high-risk deliveries to appropriate perinatal centers.¹⁵ Improvements in the 1990s included antenatal steroids and neonatal surfactant therapy, which increased the survival of very preterm infants and expanded the limits of viability to 22 to 24 weeks of gestation. Whether presurfactant-era preterm survivors in Norway experienced morbidity similar to preterm infants born after these perinatal advances is unknown. Studies comparing neonatal outcomes between the presurfactant and postsurfactant eras in the United States have not observed improved outcomes among children who were born very preterm and with extremely low birth weight.^{16 - 17} Despite advances in the postsurfactant era, a greater proportion of preterm infants have morbidity compared with term infants.^{16 ,18}

Measuring educational achievement among adults who had been born very preterm only crudely reflects the spectrum of potential neurodevelopmental outcomes noted in recent decades. Investigators examining the neurodevelopmental status of children who had extremely low birth weight observed that 39% had minor impairment and 18% had major impairment.¹⁹ These impairments subsequently may influence the ability to socialize and find a mate, with consequent reduced reproduction.^{20 - 21} The reduced reproduction of very preterm male adults also may be caused partly by their high prevalence of cryptorchidism, a condition that often results in reduced fertility.²²

Many questions remain. Researchers should assess preterm birth more accurately by improved reporting and measurement of gestational age. Although the easiest and most common measure of fetal growth is birth weight, biological maturity needs to be incorporated into research on mortality and morbidity. With advances in ultrasound technology, other measures may improve the understanding of fetal maturity. These measures include biparietal diameter and neck, head, or abdominal circumference, as well as humerus and femur length.²³ Accurate and complete reporting of stillbirths in the United States is lacking and limits the ability to understand population-based life-course outcomes. Future researchers will want to examine long-term influences on children's health. Longitudinal studies, such as the National Children's Study (http://www.nationalchildrensstudy.gov/), will need to understand the influence of the in utero environment. Data linkage throughout the life course is needed. Data on early life events permit investigators to understand the clinical circumstances of a child's delivery, neonatal treatment, and subsequent health services. These issues will remain important in understanding the relationship of preterm birth with development and functioning across the life span and into future generations.

At present, clinicians can extend guarded optimism to the families of children who are born very preterm. The findings of Swamy et al illustrate that the survival of preterm infants—although lower than that of their term peers—improves to adulthood. Nonetheless, compared with their adult term peers, fewer adult preterm survivors reproduce. These risks should be interpreted cautiously because the majority of preterm infants have good health and good reproduction. Norway demonstrates better outcomes than the United States, which has persistent, stark racial disparities.^{24 - 25} Because lifetime risk of poor health is increased among individuals who were born preterm, patients should inform their clinicians about their history of preterm birth. This information may help clinicians identify and manage childhood and adult chronic conditions. Clearly, population-based data on preterm delivery and its long-term consequences may be pertinent medical history for the nation's future health.