Intrauterine growth restriction (IUGR) is commonly defined as a birth weight less than the 10th percentile at a given gestational age (Table 123-1). It has also been defined as a fetus that has not reached its growth potential at a given gestational age, because of one or more causative factors (Lin and Santolaya-Forgas, 1998). The population from which expected fetal growth standards are derived to define the 10th percentile is of major importance. A review of all publications in the English language literature since 1963 that presented values for the 10th percentile of birth weight for gestational age has been published (Goldenberg et al., 1989). It was found that studies differed in how gestational age was determined, types of infants excluded, populations studied, and whether they were controlled for the sex of the infant as well as the race and parity of the mother. The 10th percentile birth weights at each gestational age from these various published standards differed substantially.

Different standards for fetal growth throughout gestation have been reported. These standards set the normal range of fetal growth as being between 2 SD of the mean (2.5th to 97.5th percentile) or between the 10th and 90th percentiles for gestational age. Fetal growth curves plotting fetal biometric values against gestational age have been published from a variety of populations. One of the earliest set of growth curves was derived from a population in Denver, Colorado (Lubchenco et al., 1966). These Denver curves soon became the standard fetal growth curves used throughout North America. However, these curves underestimate the incidence of small for gestational age (SGA) infants at sea level, especially during the third trimester, and also do not take into account the increase in birth weight noted during the past 30 years. Although many centers report the continued use of the Denver curves, more contemporary standards are available. These include growth standards from the state of California, based on data from more than 2 million singleton births occurring between 1970 and 1976, and also growth standards from Canada, based on data from more than 1 million singleton births and more than 10,000 twin births occurring between 1986 and 1988, as well as from national US standards, based on data from more than 3 million births in 1991 (Williams et al., 1982; Arbuckle et al., 1993; Alexander et al., 1996). Commonly used cutoffs for defining IUGR in male and female fetuses at various gestational ages are shown in Table 123-1. A further problem with the use of such cutoff values for diagnosing IUGR is that they cannot identify a fetus that has failed to reach its own growth potential, but whose weight is not yet below the usual cutoff value, such as the 10th percentile. Such a fetus may be identified through serial sonography, in which sequential weight estimates are associated with decreasing percentile values for gestational age.

The term intrauterine growth restriction is frequently and erroneously used as a substitute for the original term small for gestational age. SGA describes a population of fetuses with a weight below the 10th percentile without reference to the cause. Most SGA fetuses are normal except the small fetuses that simply reflect the normal weight distribution within a population. The use of a particular cutoff value such as the 10th percentile is clearly arbitrary, and will inevitably include a large number of fetuses that are constitutionally small, without any evidence of pathology. Up to 70% of SGA infants are small because of such constitutional reasons as maternal ethnicity, parity, or body mass index (Lin and Santolaya-Forgas, 1998). IUGR describes a subset of these SGA fetuses whose weight is below the 10th percentile as a result of a pathologic process that is due to a diverse group of disorders. The term intrauterine growth restriction is preferred to intrauterine growth retardation, as the word retardation has a tremendously negative influence on patients.

IUGR has classically been subdivided into two patterns: asymmetric and symmetric IUGR. This provides further information on fetal body size and length, rather than a simple reliance on fetal weight. With symmetric IUGR, both the head and abdomen are decreased proportionately, while asymmetric IUGR refers to a greater decrease in abdominal size, which is also referred to as the head-sparing effect (Lin and Santolaya-Forgas, 1998). Approximately 70% to 80% of cases of IUGR are asymmetric, with the remaining 20% to 30% being symmetric. It was commonly believed that asymmetric IUGR represented placental insufficiency, while symmetric IUGR was more likely to be associated with constitutional problems, such as aneuploidy. However, it is now recognized that the timing of the pathologic insult is of more importance than the actual nature of the underlying pathology in determining the pattern of IUGR, thereby calling into question the clinical utility of subdividing IUGR into such patterns (Lin and Santolaya-Forgas, 1998). Symmetric IUGR can be caused by placental insufficiency occurring early in gestation, so that by the time the fetus is examined it has evolved from an initial asymmetric pattern to a pattern of symmetric IUGR. Overall infant body proportions can be described also by using the ponderal index, which is the birth weight in grams divided by the crown-to-heel length in cubic centimeters.

The list of possible causes of IUGR is extensive and is summarized in Table 123-2. These causes can be conveniently divided into fetal, placental, and maternal factors (Lin and Santolaya-Forgas, 1998). The most common fetal factors associated with IUGR include fetal chromosomal abnormalities, structural fetal malformations, fetal infections, and complications related to multiple gestations (Figure 123-1). Chromosomal abnormalities are a major cause of IUGR (ACOG, 2000). Confined placental mosaicism is three times more common in placentas of IUGR fetuses compared with those from appropriately grown fetuses (Wilkins-Haug et al., 1995). Up to one-fourth of all infants with congenital structural malformations will have IUGR, and the incidence of growth restriction increases significantly as the number of different malformations per infant increases (Khoury et al., 1988). The number of infectious agents proven to cause IUGR is limited; they include rubella and cytomegalovirus, although no known bacterial infections have been linked to IUGR. The importance of defining a subpopulation of fetuses with IUGR lies in its association with adverse pregnancy outcome. The likelihood of perinatal morbidity and perinatal mortality increases significantly as the birth weight percentile decreases, so that once below the third to fifth percentiles, the chances of fetal death increase by as much as 20-fold (Scott and Usher, 1966). For infants weighing less than 1500 grams at term, the perinatal mortality rate is increased at least 70-fold as compared with appropriately grown term infants (Williams et al., 1982). Much of the increased perinatal morbidity and mortality in IUGR fetuses is due to the strong association between aneuploidy and structural fetal malformations with IUGR (Scott and Usher, 1966).

The incidence of IUGR varies depending on the population examined and the standard growth curves used to make the diagnosis (Goldenberg et al., 1989). Using the commonly quoted cutoff of the 10th percentile for defining pregnancies at risk for IUGR implies that at least 10% of the entire obstetric population will be labeled as being IUGR. In Europe, the commonly used cutoff for defining IUGR of 2 SD below the mean will include 5% of the total population. Approximately one third of all infants weighing less than 2500 grams at birth are not just small for gestational age, but have sustained IUGR (Lin and Santolaya-Forgas, 1998). Approximately 4% to 7% of all infants born in developed countries and 6% to 30% in developing countries are classified as growth-restricted (Scott and Usher, 1966; Lugo and Cassady, 1971; Galbraith et al., 1979).