Genetic effect on apgar score

Genetic effect on apgar score

Carla Franchi-Pinto¹, Glória Maria Duccini Dal Colletto², Henrique Krieger² and Bernardo Beiguelman²

¹Serviço de Genética Médica do Hospital da Sta. Casa de Misericórdia, São Paulo, SP, Brasil.

²Laboratório de Epidemiologia Genética, Departamento de Parasitologia, ICB, USP, São Paulo, SP, Brasil. Rua Angelina Maffei Vita, 408, Apto. 41, 01455-070 São Paulo, SP, Brasil. Send correspondence to B.B. E-mail: [e-mail]

ABSTRACT

Intraclass correlation coefficients for one- and five-min Apgar scores of 604 twin pairs born at a southeastern Brazilian hospital were calculated, after adjusting these scores for gestational age and sex. The data support a genetic hypothesis only for 1-min Apgar score, probably because it is less affected by the environment than 4 min later, after the newborns have been under the care of a neonatology team. First-born twins exhibited, on average, better clinical conditions than second-born twins. The former showed a significantly lower proportion of Apgar scores under seven than second-born twins, both at 1 min (17.5% vs. 29.8%) and at 5 min (7.2% vs. 11.9%). The proportion of children born with "good" Apgar scores was significantly smaller among twins than among 1,522 singletons born at the same hospital. Among the latter, 1- and 5-min Apgar scores under seven were exhibited by 9.2% and 3.4% newborns, respectively.

INTRODUCTION

In 1953, Apgar proposed a method to evaluate the clinical condition of newborns by giving them a score that represents a sum of several findings. This method was originally utilized one minute after birth (1-min Apgar score). However, presently it is evaluated at least once more, at 5 min after birth (5-min Apgar score).

The Apgar score takes into account five signs that can be easily determined and do not interfere with the care of the newborn. Each of these signs has a rating of zero, one or two. The sum of these ratings varies from zero to 10. Score 10 indicates excellent newborn condition, but scores varying from seven to nine are also accepted as a good prognosis, since they indicate very low risks of respiratory or neuropsychomotor developmental complications. If the 5-min Apgar score is lower than seven, the newborn is re-evaluated at 5-min intervals until clinical normalization (Fanaroff and Martin, 1992), since the correlation between neurological development and Apgar scores gradually becomes higher as the time elapsed between birth and examination increases (Grothe and Rüttgers, 1985; Committee on Fetus and Newborn, 1986). Scores from four to six indicate moderate asphyxia, and the newborn requires some kind of resuscitation, while scores between zero and three are considered severely depressed, since they are associated with high mortality risk (5-10% in the first minute, and 53% if the Apgar score remains in this range for more than 20 min) and cerebral palsy (1% in the first minute and 57% if this range is maintained for more than 20 min) (Drage and Berendes, 1966; Committee on Fetus and Newborn, 1986; Delascio and Guariento, 1987; Fanaroff and Martin, 1992).

Studies on the Apgar scores of twins have been mostly concerned with comparisons of the average scores of first-born and second-born twins, as well as with comparisons of the means observed in twins and singletons or in growth retarded and normal twins (Young et al., 1985; Nakano and Takemura, 1988; Spellacy et al., 1990; Ordorica et al., 1991; Rachdi et al., 1992; Bartnicki et al., 1992). Riese (1990) seems to have been the only author who calculated correlation coefficients for 1- and 5-min Apgar scores for a small group of monozygotic (MZ) (47) and dizygotic (DZ) (39) pairs, apparently not adjusted for gestational age and sex.

The present paper deals with the study of the Apgar score values in a large series of Brazilian twins who were not classified according to zygosity. In spite of the limitation introduced by the non-separation of monozygotic from dizygotic twins, this study could give an indication of a fetus' genetic effect on Apgar score, because if allowance is made for the rare heterokaryotic monozygotic pairs (Turpin et al., 1961), unlike-sex twins are surely dizygotic, while like-sex twins certainly include all monozygotic pairs, besides like-sex dizygotic twins.

SUBJECTS AND METHODS

One-minute and 5-min Apgar scores of 698 twins (250 male (MM), 256 female (FF), and 192 unlike-sex (MF) pairs born from 1984 to 1993 were investigated from the records of the largest maternity hospital in Campinas, SP, Brazil (Maternidade de Campinas). A control group of 1,522 singletons (754 M and 768 F) was sampled at random among the children born during the same period at the same hospital. Newborns weighing 500 g or less were not included in the present analyses, since these weights correspond to aborted fetuses (Belitzki et al., 1978). When gestational age, in weeks, was taken into account together with sex to adjust Apgar scores by multiple regression analysis, the twins were reduced to 604 pairs (218 MM, 223 FF and 163 MF), since the routinely registered records were not always complete. Intraclass correlation coefficients were calculated for Apgar values of the twins adjusted for gestational age and sex. This adjustment is mandatory since the method for calculating these coefficients demands absence of order either between or within pairs (Kempthorne and Osborne, 1961).

The rationale for estimating the heritability (h²) of a quantitative trait from data on like-sex and unlike-sex twins has taken into account the following premises:

1) If the number of the surely DZ unlike-sex twin births (192 pairs, in our case) is subtracted from the number of like-sex twins (250 MM + 256 FF = 506 pairs) then, according to the classical Weinberg's rule, the remainder like-sex pairs (314) will represent the estimated number of MZ twin births. Therefore, in our data 314/506 = 0.62 would estimate the proportion of like-sex twins who are MZ.

2) Since, theoretically, the intraclass correlation (r_i) for a quantitative trait among MZ pairs estimates the heritability, i.e. r_iMZ = h², while among DZ pairs r_{iD Z} = h²/2, then the intraclass correlation of like-sex pairs (r_il) would estimate r_il = ch² + (1-c)h²/2, where c is the proportion of like-sex twins, and r_ul = h²/2 would estimate the intraclass correlation of unlike-sex pairs. Therefore, r_il + r_ul= ch² + (1-c)h²/2 + h²/2 = h²(1 + c/2), which enables us to solve for h² = (r_il + r_ul)/(1 + c/2).

Ratings of zero, one or two attributed to each of the five signs used for the evaluation of the Apgar scores of all newborns were as follows:

1) Heart rate: zero = no heart beat can be seen, felt or heard; one = a heart rate under 100; two = a heart rate over 100.

2) Respiratory effort: zero = apnea after birth; one = all types of respiratory effort, such as irregular, shallow ventilation; two = breath and lusty cry.

3) Reflex irritability after suctioning the oropharynx and nares with a soft rubber catheter: zero = absence of response; one = facial movements; two = facial grimaces, sneezing or coughing.

4) Muscle tone: zero = complete flaccidity; one = some movements; two = good tone, arms and legs spontaneously flexed and resistant to extension.

5) Skin color: zero = cyanosis; one = cyanosis only of the extremities; two = the entire child is pink.

RESULTS AND DISCUSSION

The 250 MM, 256 FF, and 192 MF twins were distributed according to birth order of the co-twins, with Apgar scores divided into two classes: under 7 (bad prognosis) and 7-10 (good prognosis) (Table I). Both 1- and 5-min Apgar scores of first and second born twins were closely associated in like-sex, but were weakly associated in unlike-sex twins, as pointed out by the chi-square values. The weaker association exhibited by the known dizygotic twins suggested that 1- and 5-min Apgar scores of newborns might depend on genetic factors.

Table I
- Independence tests to investigate the association between first and second born twins for 1- and 5-min Apgar scores of male (MM), female (FF) and unlike-sex (MF) twin pairs. Percentage values are within parentheses.

Second twin	First twin
	MM			FF			MF
	< 7	³ 7	Total	< 7	³ 7	Total	< 7	³ 7	Total
< 7	33	48	81	30	42	72	15	40	55
			(32.4)			(28.1)			(28.6)
³ 7	16	153	169	12	172	184	16	121	137
			(67.6)			(71.9)			(71.4)
Total	49	201	250	42	214	256	31	161	192
	(19.6)	(80.4)		(16.4)	(83.6)		(16.1)	(83.9)
c 2; 1 d.f.	33.98; P << 0.001			46.61; P << 0.001			7.05; P < 0.01

One-minute Apgar score

Five-minute Apgar score

Second
twin

First twin

MM

FF

MF

< 7

³

7

Total

< 7

³

7

Total

< 7

³

7

Total

< 7

14

16

30

7

22

29

5

19

24

(12.0)

(11.3)

(12.5)

³

7

9

211

220

7

220

227

8

160

168

(88.0)

(88.7)

(87.5)

Total

23

227

250

14

242

256

13

179

192

(9.2)

(90.8)

(5.5)

(94.5)

(6.8)

(93.2)

c

2; 1 d.f.

57.29; P << 0.001

22.05; P << 0.001

8.59; P < 0.01

The possibility of a genetic influence prompted us to calculate intraclass correlation coefficients for 1- and 5-min Apgar scores of twins after adjusting for gestational age and sex (Tables II and ^III). The total variances of the adjusted scores did not differ significantly (^{Table III}), while the covariances of like-sex pairs were significantly higher than those of unlike-sex twins both for 1-min (F_(440;162) = 2.39) and 5-min Apgar scores (F_(440;162) = 1.64). However, the within pair variance was significantly higher among the unlike-sex twins (F_(162;440) = 1.27), but only for the 1-min Apgar score. Coherently, the intraclass correlation coefficient of like-sex twins was significantly higher than that of unlike-sex pairs only for the 1-min Apgar score (t_(¥) = 2.513; P < 0.02).

Table II
- Regression coefficients obtained by multiple stepwise regression analysis of 1- and 5-min Apgar scores (Y) on gestational age and sex.

± SD	One minute	Five minutes
	7.35 ± 2.17	8.90 ± 1.75
Effects a gestational age sex	b ± SE	b ± SE
	-2.242 ± 0.682 ***	0.251 ± 0.538
	0.264 ± 0.018 ***	0.239 ± 0.015***
	-0.254 ± 0.116 *	-0.269 ± 0.091**
SQ regression	836.705	687.434
d.f.	2	2
SQ residual	4855.360	3021.446
d.f.	1205	1205
F	103.826	137.080
r2	0.147	0.185

Table III - Intraclass correlation coefficients (r_i) for 1- and 5-min Apgar scores of 604 twins adjusted for gestational age and sex.

Estimates

One-minute Apgar score

Five-minutes Apgar score

Like-sex

Unlike-sex

Like-sex

Unlike-sex

s²(B)

5.652

4.524

3.408

3.025

s²(W)

2.519

3.214

1.622

1.940

Cov.

1.567

0.655

0.893

0.543

s²(T)

4.086

3.869

2.515

2.483

ri

0.383

0.169

0.355

0.219

N

441

163

441

163

s

²

(B) = Between pair variance; s

²

(W) = within pair variance; Cov. = covariance; s

²

(T) = total variance; N = number of twins.

The data (^{Table III}) support a genetic hypothesis only for 1-min Apgar score, since one may infer a genetic basis for a quantitative trait if: 1) the total variances in the twin groups do not differ; 2) the within pair variance of unlike-sex twins is larger than that observed in like-sex twins; 3) the covariances and intraclass correlation coefficients of like-sex twins are larger than the corresponding estimates of unlike-sex pairs. Therefore, the heritability (h²) of 1-min Apgar score was estimated as h² = 0.421 based on the proportion of MZ pairs among the like-sex twin births (c = 0.62) and on the intraclass correlation coefficients for like-sex (r_il = 0.383) and unlike-sex twin births (r_ul = 0.169). Although there is no evidence of a genetic mechanism affecting 5-min Apgar score, a genetic hypothesis for this trait cannot be rejected by the data (^{Table III}), mostly because like-sex twins include both monozygotic and dizygotic pairs. Nevertheless, the intraclass correlation coefficients calculated by Riese (1990) for Apgar scores of 47 monozygotic and 39 dizygotic twins also support a genetic hypothesis only for 1-min Apgar score. Therefore, it seems reasonable to accept that it is easier to detect a genetic effect for 1-min Apgar score, since it is less affected by the environment than 4 min later, after the newborns have been under the care of a team of neonatologists.

In spite of the Apgar scores of twins being more closely associated among same-sex pairs than among MF twins (Table I), the proportions of individuals in the two main classes of Apgar scores (< 7 and ³ 7) did not differ significantly among the MM, FF and MF groups of twins at 1 and at 5 min after birth. These comparisons presented by the first-born twins yielded c²₍₂₎= 1.22, 0.50 < P < 0.70, and c²₍₂₎= 2.70, 0.20 < P < 0.30, for 1-min and 5-min Apgar scores, respectively. The same comparisons for second-born twins resulted in c²₍₂₎= 1.27, 0.50 < P < 0.70, and c²₍₂₎= 0.14, 0.90 < P < 0.95.

This similarity provides a joint estimate that, among the first-born twins, Apgar scores under 7 were exhibited by 17.5% at 1 min and 7.2% at 5 min. Among the second-born twins, Apgar scores under 7 were shown by 29.8% at 1 min and by 11.9% at 5 min. Since the first-born twins showed a significantly lower proportion of Apgar scores under 7 than second-born twins both at one (c²₍₁₎= 29.35; P << 0.001) and 5 min (c²₍₁₎= 9.05; P < 0.001) our data agree with previous claims that first-born twins exhibit, on average, better clinical conditions than second-born twins (Young et al., 1985; Nakano and Takemura, 1988; Spellacy et al., 1990; Ordorica et al., 1991; Rachdi et al., 1992).

One- and 5-min Apgar scores under 7 were exhibited, respectively, by 9.2% and 3.4% of 1,522 singletons born at the same hospital. Since these proportions are significantly smaller than the corresponding frequencies observed among twins, our data also show that the proportion of newborns with Apgar scores having good prognostic values is significantly larger among singletons than among twins. These observations agree with those reported by Spellacy et al. (1990) for a large twin sample.

ACKNOWLEDGMENTS

The authors are deeply indebted to Professor Dr. Newton E. Morton for the helpful comments, which improved this article. This work was supported by CNPq. Publication supported by FAPESP.

RESUMO

Os coeficientes de correlação intraclasse foram calculados para os índices de Apgar 1 e 5 minutos após o nascimento de 604 pares de gêmeos em uma maternidade do sudeste brasileiro, depois que esses índices foram ajustados para idade gestacional e sexo. Os dados obtidos apoiaram a hipótese genética apenas em relação ao primeiro índice de Apgar, provavelmente porque ele é menos influenciado pelo ambiente do que 4 minutos depois, quando os recém-nascidos já estiveram sob os cuidados de uma equipe de neonatologistas. Os gêmeos nascidos em primeiro lugar apresentaram, em média, melhor estado clínico que os nascidos em segundo lugar, visto que os primeiros mostraram uma proporção de índices de Apgar inferiores a 7 significativamente menor do que os nascidos em segundo lugar, tanto um minuto (17,5% contra 29,8%) quanto cinco minutos após o nascimento (7,2% contra 11,9%). A proporção de recém-nascidos com índices de Apgar que indicam bom prognóstico foi significativamente menor nos gêmeos do que em 1.522 conceptos únicos nascidos na mesma maternidade. Índices de Apgar menores do que 7, obtidos um e cinco minutos após o nascimento, foram encontrados respectivamente em 9,2% e em 3,4% dos recém-nascidos de parto único.

(Received February 4, 1998)

• Apgar, V. (1953). A proposal for a new method of evaluation of the newborn infant. Anesth. Analg. 32: 260-267.
• Bartnicki, J., Meyenburg, M. and Saling, E. (1992). Small for gestational age twins: a retrospective analysis of clinical and acid-base status immediately after delivery. Int. J. Gynecol. Obstet. 37: 97-98.
• Belitzki, R., Fescina, R.E. and Ucieda, F. (1978). Definiciones y terminologias aplicables al periodo perinatal. Recomendaciones de la Organización Mundial de la Salud y modificaciones de la F.I.G.O. Publicación Científica del CLAP, No. 757: 136-147.
• Committee on Fetus and Newborn (1986). American Academy of Pediatrics. Pediatrics 78: 1148-1149.
• Delascio, D. and Guariento, A. (Ed.) (1987). Recém-nascido. Assistęncia. In: Obstetrícia Normal - Briquet 3rd edn. Sarvier, Săo Paulo, SP, Brasil.
• Drage, J.S. and Berendes, H. (1966). Apgar scores and outcome of the newborn. Pediatr.Clin. North Am. 13: 637-643.
• Fanaroff, A.A. and Martin, R.J. (1992). Neonatal-Perinatal Medicine: Diseases of the Fetus and Infant. 5th edn. Vol. 2. Mosby-Yearbook, New York, NY, USA.
• Grothe, W. and Rüttgers, H. (1985). Twin pregnancies: an 11-year review. Acta Genet.Med. Gemellol. 34: 49-58.
• Kempthorne, O. and Osborne, R.H. (1961). The interpretation of twin data. Am. J. Hum.Genet. 13: 320-339.
• Nakano, R. and Takemura, H. (1988). Birth order in delivery of twins. Gynecol. Obstet Invest. 25: 217-222.
• Ordorica, S.A., Frieden, F.J., Hoskins, I.A. and Young, B.K. (1991). Discordant twins: acid-base status. Acta Genet. Med. Gemellol. 40: 373-381.
• Rachdi, R., Fekih, M.A., Mouelhi, CH. and Brahim, H. (1992). Problčmes posés par l'accouchement de la grossesse gémellaire. Rev. Fr. Gynécol. Obstét. 87: 295-298.
• Riese, M.L. (1990). Genetic influences on neonatal temperament. Acta Genet. Med. Gemellol. 39: 207-213.
• Spellacy, W.N., Handler, A. and Ferre, C.D. (1990). A case-control study of 1253 twin pregnancies from a 1982-1987 perinatal data base. Obstet. Gynecol. 75: 168-171.
• Turpin, R., Lejeune, J., Lafourcade, J., Chigot, P.L. and Salmon, C. (1961). Présomption de monozygotisme en dépit d'un dimorphisme sexuel: sujet masculin XY et sujet neutre haplo X. C. R. Acad. Sci. Paris 252: 2945-2946.
• Young, B.K., Suidasn, J., Antonie, C., Silverman, F., Lustig, I. and Wasserman, J. (1985). Differences in twins: the importance of birth order. Am. J. Obstet. Gynec 151: 915-921.

Publication Dates

• Publication in this collection
  02 June 1999
• Date of issue
  Mar 1999

History

• Received
  04 Feb 1998