Intrauterine growth retardation, body size, body composition and physical performance in adolescence

Introduction
Literature review
INCAP longitudinal study (1969-77) and its follow-up study (1988-89)
References
Discussion

R Martorell¹, U Ramakrishnan¹, DG Schroeder¹, P Melgar² and L Neufeld²

Correspondence: Dr R Martorell

¹Department of International Health, The Rollins School of Public Health of Emory University, 1518 Clifton Road, N.E., Atlanta, Georgia 30322;
²Instituto de Nutrición de Centro América y Panamá, Carretera Roosevelt, Zona 11, Apartado Postal 1188, Guatemala City, Guatemala, Central America

Studies about effects of IUGR on growth in childhood as well as on body size, body composition and physical performance in adolescence and adulthood are reviewed. The review is based on 12 studies that distinguished IUGR from other types of low birthweight and compared outcomes of IUGR cases with those of non-IUGR controls. This information is complemented by results of a follow-up study of IUGR cases and controls carried out in Guatemalan adolescents and young adults.

In Guatemala as well as in other countries, IUGR newborns showed partial catch-up growth during the first one or two years of life, and then maintained their achieved place in the growth distribution. Guatemalan IUGR cases were shorter, lighter and weaker than non-IUGR controls as adolescents and young adults. The differences in adult body size observed in Guatemala between cases and controls are similar to those found in more affluent countries (i.e. about 5 cm in height and 5 kg in weight).

Introduction

Intrauterine growth retardation (IUGR) is said to occur when infants are born full-term (usually ³37 weeks of gestation) but small (< 2500 g or other similar criteria, sometimes adjusted for gestational age using an external reference population). This paper focuses on the consequences of IUGR for later growth in childhood and for attained body size, composition and physical performance at adolescence and adulthood. Two types of results are presented. The first are relevant findings in the literature, whether from developed or developing countries. The second are results from a study carried out in Guatemala, the first to provide information from a developing country on consequences of IUGR for the adolescent.

Literature review

The review was limited to 12 studies which distinguished IUGR from other types of low birthweight and which compared outcomes in IUGR cases to those in non-IUGR, full-term healthy infants (referred to as cases and controls respectively). For example, studies from Finland (Nilsen et al, 1984) and the USA (Frisancho et al, 1994) which related birthweight to size at 17-18 years of age were excluded because their samples of small-for-date infants included premature infants (< 37 weeks). Also, the review was limited to those studies with follow-up to at least two years of age. This minimum limit was chosen because of findings in the literature indicating stability in the growth patterns of IUGR subjects after about 2 years of age.

The studies identified are given in Table 1, in order of age of follow-up, from 2 to 19 years. Most were prospective studies and all but three were from developed countries. No study was found from developing countries which assessed status at adolescence or adulthood. A variety of definitions of cases and controls were used as shown in Table 1; however, these differences are unlikely to influence the main conclusions of the literature review. Height, weight and head circumference were the most commonly used measures of size at follow-up and the body mass index (BMI) was the measure most often used to study body composition.

The main results of the literature review indicate that IUGR newborns catch-up partially in growth relative to controls during the first one or two years of life (Barros et al, 1992; Fitzhardinge and Inwood, 1989; Villar et al, 1984; Walther, 1988; Albertsson-Wikland and Karlberg, 1994). After about 2 years of age, IUGR subjects maintain their achieved place in the distribution and neither catch-up nor fall further behind.

The ponderal index (PI = [weight(g)/length³ (cm)] x 100) is an indicator of wasting. Newborns born with low ponderal indices (also known as "disproportionate" or thin) tend to experience more pronounced catch-up growth in childhood than those born with adequate ponderal indices, according to a study in Guatemala (Villar et al, 1984). However, a study from New Zealand found few differences between proportionate and disproportionate IUGR subjects in achieved size at 18 years of age (Williams et al, 1992).

Maturation is not delayed in adolescents born with IUGR as judged by the timing of peak height velocity (Albertsson-Wikland and Karlberg, 1994) or skeletal age and sexual maturation (Westwood et al, 1983).

Achieved size at ages 2 to 7 years was significantly different in cases compared to controls (Fritzhardinge and Inwood, 1989; Villar et al, 1984; Barros et al, 1992; Pryor, 1992; Low et al, 1981; Mata, 1978; Walther, 1988; Albertsson-Wikland and Karlberg, 1994; Williams et al, 1992). Similarly, achieved sizes were significantly less in cases than controls by about 5 cm in height and 5 kg in weight at ages 17-19 years (Westwood et al, 1983; Paz et al, 1993; Lagerström et al, 1994; Williams et al, 1992; Albertsson-Wikland and Karlberg, 1994). Findings were similar in males and females in all aspects regarding body size and maturation.

Table 1. IUGR and future growth and body size

Author(s) Country Age at follow-up

Design

Definitions

Results

Fritzhardinge and Inwood (1989) Canada 2 yrs

Prospective study of IUGR cases born at a hospital between Oct 1982 and May 1986.

IUGR was defined as ³ 36 wks of gestation and < -2 SD weight for gestational age. The IUGR sample was composed of 34 males and 44 females.

There was catch-up in growth but most occurred in the first 9 months. By 18-24 months, the percentage of cases that continued to be below the fifth percentile was 29% for weight, 29% for length and 27% for head circumference. A high ponderal index at birth was associated with shortness at 18-24 months.

Villar et al (1984) Guatemala 3 yrs

Prospective study of newborns of 4 rural villages (born 1969-77).

Cases and controls were ³ 37 wks. IUGR was defined as < 10th percentile of weight for gestational age. Controls (n = 146) had birth weights above the 10th but £ 90th. IUGR infants were divided into LPI (n=21) and API (n=38) if below or above the 10th percentile of ponderal index by gestational age, respectively.

At birth, API infants were shorter, lighter and had smaller head circumference than controls. LPI infants differed from controls only in birth weight. There was greater catch- up growth in the LPI group particularly in the first few months of life. At 3 years of age, the API cases were the shortest and lightest. LPI infants had similar lengths and nearly the same weight as controls. Head circumferences were reduced to the same extent in both LPI and API cases relative to controls.

Barrow et al (1992) Brazil 4 yrs

Prospective study of all hospital births in 1982 in the city of Pelotas.

IUGR: ³ 37 wks of gestation and < 10th percentile of weight for gestational age (n=422). Controls: ³ 37 wks and weight ³ 10 percentile (n=3958).

Some 1008 children had complete data at ages 11, 23 and 47 months. Sex adjusted mean percentiles:

IUGR

Controls

Age (m)
Wt (kg)
Ht (cm)

0
4
-

11
14
14

23
17
12

47
22
15

0
61
-

11
50
42

23
53
37

47
58
40

Pryor (1992) New Zealand 4 yrs

Cases and controls identified from maternity hospital records in 1983-84.

IUGR: 37 wks of gestation and < 3rd percentile of weight for gestational age (n=67).Controls ³ 38 wks and weight ³ 25th percentile (n=46)

Differences in head circumference between groups were twice as large birth as at follow-up.

IUGR (X ± SD)

Controls (X ± SD)

Wt (centile)
Ht (centile)
HC (cm)

26.0 ± 25.8
29.4 ± 23.8
50.2 ± 1.7

61.1 ± 27.9
62.7 ± 23.1
51.6 ± 1.3

<.0001
<.001
<.001

Low et al (1981) Canada 5 yrs

Prospective study carried out 1970-80.

IUGR: ³ 37 wks of gestation and< 10 percentile of weight for gestational age (n = 76). Controls: ³ 37 wks and weight ³ 25th percentile (n = 88)

At 60 months, differences favored controls and were 2.3 kg for weight, 3.6 cm for height, 1.2 cm for head circumference and 1.5 cm for chest circumference.

Mata (1978) Guatemala 6 yrs

Prospective study of newborns from the village of Santa María Cauqué (1964-73).

Cases and controls ³ 37 weeks. IUGR defined as < 2500g. Controls were divided into Moderate (2501-3000g) and High (³ 3001g) groups. Sample sizes were 143 cases and 199 and 43 controls respectively.

Author presents curves from birth to 6 years for length and head circumference. At 6 years, differences between IUGR and high group were about 2 kg in weight and 2 cm in head circumference.

Walther (1988) Netherlands
7 yrs

Prospective study of consecutively born term infants (38-42) weeks.

IUGR-LPI (n=24) were those with birthweights < 10th and ponderal indices £ 5th percentile for gestational age. Controls (n = 24) were between the 10 and 90th percentiles for weighs end above the 10th percentile for ponderal index.

There was some catch-up growth in the IUGR group but only in the first few years. The ponderal index increased quickly after birth in the LPI cases. Sizes (x + SD) at 7 years were as follows:

IUGR-LPI

Controls

Ht (cm)
Wt (kg)
P.I.
HC (cm)

121.5 ± 4.4
20.6 ± 1.4
1.15 ± 0.09
50.8 ± 1.6

124.0 ± 4.6
23.8 ± 1.5
1.25 ± 0.12
51.9 ± 1.4

N.S.
<.001
<.005
<.02

Paz et al (1993)
Israel
17 yrs

All full-term singletons born in a single maternity ward (Nov 70-Dec 71) and who were evaluated by the army at 17 yrs of age were included.

IUGR:< 3rd percentile of weight Males (x ± SD) for gestational age. Controls: ³ 3rd percentile Sample sizes were 30 males 34 females with IUGR and 1690 controls.

Males (x ± SD)

IUGR

Controls

Ht (cm)
Wt (kg)
BMI (kg/m²)

169.4 ± 6.0
62.8 ± 11.7
21.8 ± 3.4

175.0 ± 6.6
64.9 ± 10.1
21.1 ± 2.9

<.0001
N.S.
N.S.

Females (x ± SD)

Ht (cm)
Wt (kg)
BMI (kg/m²)

160.3 ± 6.4
54.7 ± 9.0
21.2 ± 2.9

163.7 ± 5.9
56.6 ± 7.8
21.2 ± 2.7

<.001
N.S.
N.S.

Lagerström et al (1994)
Sweden
18 yrs

Records at birth of male children born in the Stockholm area in 1953 were linked to army tests at 18 yrs of age.

IUGR: ³ 38 wks of gestation and < 2500g. Controls: ³ 38 wks and ³ 2500g. Sample sizes were 72 IUGR cases and 5355 controls.

Results (x+ S.E.) at 18 years for males

IUGR

Controls

H (cm)
Wt (cm)
Chest (cm)
Waist (cm)

175.8 ± 6
62.2 ± 9
86.3 ± 6
72.4 ± 6

179.6 ± 7
66.8 ± 9
88.2 ± 6
74.4 ± 6

<.001
<.001
<.0.5
<.0.5

Albertsson-Wikland and Karlberg (1994) Sweden
18 yrs

Pupils in the final grade of school (~ 18 yrs of age) in 1992 in Gothenburg were selected for study. Records from birth obtained from health facilities and schools.

All were full term (37-43 wks). 111 (3.1%) were of low birthweight (below -2 SD) and 141 (3.5%) were of low birth length (below -2 SD); 54 (1.5%) were both light and short at birth. These were compared to normal for gestational age children. "Catch-up" growth was defined as above -2 SD at 2 yrs of age and "non catch-up" as below -2 SD at 2 yrs.

Being light or short at birth influenced body size. Those light at birth and who experienced catch-up growth (n = 82) were half a standard deviation below the reference mean at 8 yrs and at final height. Those without catch-up (n=9) were-2.3 SD at 8 yrs and-1.7 SD at adulthood. Peak height velocity was not delayed in the IUGR groups.

Williams et al (1992) New Zealand 18 yrs

Prospective study of children born at hospital (April 1972-March 1973).

Cases and controls were ³ 37 wks IUGR-API: < 10 percentile of birthweight for gestational age and appropriate ponderal index. IUGR-LPI: < 10 percentile of birthweight for age and low ponderal index. Controls: 11-89 percentiles of birthweight for age. Large: ³ 90 percentile of birthweight for age.

The results (x) were reported with sexes combined

IUGR-API

IUGR-LPI

Controls

Large

Age (7 yrs)

(n=41)

(n=29)

(n=680)

(n=67)

Ht (cm)
Wt (cm)

117.9
21.4

117.9
21.7

121.2
23.4

123.5
25.0

Age (18 yrs)

(n=41)

(n=30)

(n=683)

(n=73)

Ht (cm)
Wt (kg)

165.1
61.4

167.5
61.8

170.5
66.4

174.4
71.9

Westwood et al (1983) Canada 13-19 yrs

Prospective study of 33 full term, non- asphyxiated, IUGR cases born 1960-66 at a teaching hospital and 33 matched controls (26 females, 7 males).

All were ³ 38 wks in gestation. IUGR: > 2 SD below mean weight. Controls: BW between 25th and 75th percentiles.

IUGR

CONTROLS

Ht (cm)
Wt (cm)
HC (cm)
Triceps (mm)
Subs. (mm)
Arm area (cm²)

159.2
50.8
53.8
15.3
14.6
27.0

165.2
58.8
55.4
18.2
14.9
28.1

.001
.001
.001
.020
N.S.
N.S.

There have been several studies from developed countries linking low birthweight to fatness and other indicators of risk of chronic disease (Barker, 1992; Barker, 1994; Barker, 1996; Fall et al, 1995; Law et al, 1992). However, these studies failed to distinguish IUGR from prematurity. Since premature, low birthweight newborns make up about 55% of all low birthweight cases in developed countries (Villar and Belizan, 1982), these findings cannot be said to necessarily apply to IUGR cases. Few of the studies in Table 1 examined whether IUGR is related to fatness at follow-up. Walther (1988) found that the ponderal index was significantly lower at age 7 yrs in cases than controls. Westwood et al (1983) found a relationship at 19 years with subscapular skinfolds (i.e. IUGR cases had thinner folds) but not with triceps skinfolds or with arm fat area. Paz et al (1993) found that BMI was similar in cases and controls at 17 yrs. Finally, Fritzhardinge and Inwood (1989) reported that a high ponderal index at birth was associated with shortness at 18-24 months of age. Thus, there is no support in the IUGR literature for a relationship between low birthweight and greater fatness at follow-up.

Three major gaps in knowledge about the long-term effects of IUGR were identified in the review of the literature. First, there were no reported studies from developing countries that assessed effects at adolescence or adulthood. It would be important to document whether the consequences of IUGR are magnified in settings of greater poverty, malnutrition and disease during postnatal life. Second, apart from the study by Albertsson-Wikland and Karlberg (1994), no study considered the long-term effects of IUGR in relation to postnatal growth retardation. Finally, very little is known about consequences of IUGR for body composition, not only fatness but lean body mass as well, and for physical performance (work capacity, strength).

Contents - Previous - Next