Of the approximately 4 million annual live births in the United States, 7.5% are classified as low birth weight (LBW, <=2,500 grams) and 1.5% as very low birth weight (VLBW, <=1,500 grams). These figures translate to about 300,000 LBW infants born each year in this country. Unfortunately, while many public health statistics, for example, infant mortality rate, have shown steady improvement over the past quarter century, the LBW rate has been gradually increasing because of such factors as prenatal substance exposure, fertility technologies, and advanced maternal age.

Thanks to state-of-the-art perinatal services and incredible advances in neonatal intensive care technology (approximately 700 tertiary neonatal intensive care units [NICUs] in the United States), the overwhelming majority of even the smallest and sickest LBW infants now survive to hospital discharge and well beyond. The regionalization of perinatal care with maternal transport to highly specialized centers, frequently referred to as Level III facilities, has had a major impact on supporting the survival of LBW, premature infants. Ninety percent and higher survival rates are currently the expectation down to 1,000 grams (28 weeks gestation) with rates for extremely low birth weight infants (ELBW, <=1,000 grams) overall between 75 and 85% and almost 50% as low as 500-600 grams (24 weeks gestation). These are almost unbelievable survival rates and dramatically higher than 25 years ago. They are largely the result of contemporary, sophisticated, aggressive neonatal intensive care.

The increasing survival of fragile premature infants has, fortunately, produced a strong interest in defining their long-term outcomes, identifying their "quality of life," and investigating intervention strategies for improving their developmental/behavioral prognosis. The neurodevelopmental outcome of LBW infants can be conveniently divided into issues of so-called "major" and "minor" impairments. Major impairments include such disabilities as cerebral palsy, global developmental delay (mental retardation over time), epilepsy, hearing and/or visual loss, and progressive hydrocephalus. These impairments tend to be more severe (although they certainly present in milder forms as well), identified in the first two years of life, and caused by substantial brain insults such as hemorrhage or hypoxia/ischemia. They often occur more frequently at lower birth weights and gestational ages, and have identifiable markers on neuroimaging.

Approximately 15-20% of VLBW survivors and 20-25% of ELBW survivors will develop one or more major neurodevelopmental impairments. Attempts at preventing or significantly reducing these outcomes following substantial brain injury by means of a variety of early intervention approaches have, for the most part, proved disappointing. Physical therapy following NICU discharge does not prevent cerebral palsy; more global developmental intervention in early infancy does not prevent mental retardation. These major outcomes will probably be significantly reduced only by the primary prevention of prematurity itself. However, it should be noted that, with modern neonatal intensive care, the great majority of VLBW and even ELBW survivors are not left with a major neurodevelopmental disability. And, most importantly, increased survival in recent years has not produced an increased rate of major impairments even at the lowest birth weights.

In contrast, "minor" neurodevelopmental and neurobehavioral impairments remain prevalent among survivors and become increasingly apparent as follow-up populations age. Because these multiple brain dysfunctions increase in incidence as birth weight and gestational age decrease, they have become especially prevalent with the increased survival of the smallest and sickest "micro-preemies." Forty to 50 percent of VLBW survivors and 60-70% of ELBW survivors will experience varying combinations of temperament problems, language delays, fine motor deficits, visual-motor dysfunction, sensory integration dysfunction, social incompetence, emotional immaturity, borderline intelligence, attention deficits, learning disorders, and ultimately diminished school performance. In the hundreds of long-term follow-up studies performed across multiple developmental domains, a consistent theme emerges: groups of VLBW, premature children score significantly lower and perform more poorly than socioeconomically matched groups of normal birth weight, full-term children. These regular differences persist even when examining relatively "clean" groups of VLBW children who experienced mild neonatal illness, are free of major or moderate impairments, and reside in adequate parenting environments. It definitely appears that, in addition to the more obvious brain insults of hemorrhage, hypoxia/ischemia, and infection, there is often an interference with the complex processes of early brain development and a resulting neurodevelopmental price to pay associated with prematurity.

These common "minor" neurodevelopmental impairments are, of course, not necessarily minor in terms of their adverse impact on functional life skills including learning, socializing, working, and living independently. As the high prevalence of these brain dysfunctions among VLBW survivors has become widely realized, a corresponding interest in potential preventive or ameliorative interventions has naturally grown. The first place to look, logically, has been the NICU environment itself. The history of develop mental intervention provided in the special care nursery is one of initial enthusiasm, gradual evidence-based physiological understanding, and eventual complete reversal of approach. At first, the LBW, premature infant was essentially viewed as a deficient full-term infant in need of "infant stimulation" activities to bridge the developmental gap. Accordingly, early intervention efforts tended to consist of a multimodal variety of manipulations of the nursery environment including combinations of tactile (massage, brushing), kinesthetic-vestibular (rocking, oscillating beds), auditory (talking, music boxes), and visual (mobiles, bright decorations) stimulations. Since small controlled studies generally failed to convincingly identify an optimal approach, nurseries were left to pick and choose for themselves.

Fifteen to 20 years ago important data began to emerge regarding the fragility of the recovering premature infant and its susceptibility to the stresses of neonatal intensive care. It became apparent that the intrusive procedures, repetitive handling, and other "overstimulations" frequently resulted in interference with autoregulation and autonomic nervous system stability to the point of apnea, bradycardia, hypotension, and reduced cerebral blood flow. This new information clearly indicated that the existing conceptualization of the needs of the premature infant must be reevaluated. Could the standard practices of the life-saving NICU with its ecology of continual disruption of infant diurnal rhythms and sleep-wake cycles actually be detrimental to long-term brain function?

The prevailing approach to the NICU environment question in recent years has thus evolved to a dramatically different orientation than the original one. Through the work of Dr. Heidi Als and others, the concept of individualized developmental care has been defined and is increasingly applied in intensive care nurseries across the United States and globally. At the core of this philosophy is the somewhat revolutionary idea that the unstable premature infant primarily needs time and space to recover and to be handled with care. This "infant protection" approach stands in marked contrast to previous stimulatory attitudes. Individualized developmental care requires a thorough appreciation of brain development by gestational age as well as assessment techniques for determining the neurological readiness for interactions and interventions of each infant. It includes such specific details as a "minimum handling" protocol timed to the infant's receptivity cues, facilitative positioning, skin-skin contact (kangaroo care), and radical restructuring of the NICU to reduce excessive light and noise pollution. It sensitively attempts to regard each infant as developmentally unique and in need of careful "brain care." Developmental care has been evaluated in several studies with preliminarily positive short-term outcomes including shortened respiratory course, improved weight gain, and reduced duration of hospitalization. At the least, it makes good sense and humane care for sick VLBW infants, their families, and their NICU providers. However, even if optimally applied, developmental care will probably not eliminate the many neurodevelopmental dysfunctions associated with prematurity.

Developmental interventions for LBW, premature infants after nursery discharge have also been investigated. By far the most comprehensive and ambitious of these is the Infant Health and Development Program (IHDP). The IHDP is a national, multisite randomized clinical trial of a related group of early interventions designed to optimize the long-term school outcomes and minimize the developmental and behavioral sequelae of prematurity. In the mid-1980s approximately 1,000 LBW infants at eight university sites were randomized to either an intervention or follow-up group prior to leaving the hospital. The study population was stratified by birth weight with one-third of the infants (Heavier group) having a birth weight between 2,001 and 2,500 grams and two-thirds (Lighter group) weighing 2,000 grams or less at birth. The intervention "package" was guided by a written curriculum of early developmental activities and included regular home visitation, daily attendance at a special developmental center, and topic-focused parent group meetings over the first three years of life. Both groups received close follow-up assessment, social service support as necessary, and referral for specific therapies if required. While the formal set of interventions concluded at age 3, the study population has continued to be followed and assessed in multiple areas through 8 years of age. Over 90% of study families are still available for long-term follow-up.

Initial results upon completion of the study interventions appeared quite positive. Intervention group children had significantly higher intelligence quotient (IQ) scores and significantly fewer behavior problems. Obviously, these early results were hailed as lessening the impact of low birth weight and prematurity. But just as with early interventions for socioeconomically disadvantaged children, the large IQ differences diminished rapidly over time. Most disconcerting of all was the observation that, as birth weight and gestational age decreased, so did the difference in measured IQ between the intervention and follow-up groups. The Lighter birth weight children showed no group differences in any developmental or behavioral parameter, including early academic skills, at ages five and eight. Heavier birth weight children retained small, significant group differences in IQ and language at these later ages, based mostly on the greater intervention impact on children and families at high environmental risk. In marked contrast, even these extensive, high-quality interventions seem incapable of producing significant developmental gain at early school age or preventing impairment in those premature children at the greatest biomedical risk of brain dysfunction. A final follow-up evaluation of the IHDP population in late high school is currently planned to investigate the possibility of long-term intervention effects on academic and social function.

So, what have we really learned and where do we go from here? It should be clear to all by now that prematurity, especially extreme prematurity, results in a substantial intrinsic risk to the developing brain even in the absence of overt neuroimaging abnormalities. This risk is multifactorial and frequently subtle but is manifested in long-term deficits in various life skills. If this fragile infant was also prenatally exposed to potentially deleterious substances and eventually goes home to a vulnerable, chaotic environment, then the risks to competent development and behavior are multiplied.

We are gradually learning ways to interrupt this cycle. We need to more fully apply what we already know. Public health efforts to reduce the incidence of low birth weight and prematurity should be supported conceptually and financially. The vascular and neurochemical risks of tobacco, alcohol, cocaine, marijuana, methamphetamine, and other adverse substances to the fetal brain should be publicly emphasized rather than exonerated. Neonatologists must give at least as much attention to the premature brain as to the premature lung and increase applicable knowledge in such areas as cerebral blood flow and neonatal nutrition. The philosophy and practices of individualized developmental care should be widely applied in NICUs just as the birthing and care of full-term infants were humanized several decades ago.

Finally, while early intervention efforts directed at LBW, premature infants in the early years of life have not proven as beneficial as hoped in preventing the negative outcomes associated with biomedical risk, it is increasingly evident that these comprehensive interventions can significantly improve the language, learning, and behavioral outcomes associated with environmental risk. However, to be maximally effective these interventions must have a high level of intensity and duration (ie, begin early in life and continue into early school years), actively involve families and assist them in solving overwhelming life needs, and be conducted by competent, committed individuals who are adequately compensated for this important work. This is a tall order and unfulfilled challenge to policy makers and to all of us. Are we as a society, both collectively and individually, up to it?