Intended for healthcare professionals


Children born after intracytoplasmic sperm injection: population control study

BMJ 1999; 318 doi: (Published 13 March 1999) Cite this as: BMJ 1999;318:704
  1. Alastair G Sutcliffe (icsi{at}, lecturera,
  2. Brent Taylor, professora,
  3. Jun Li, epidemiologista,
  4. Simon Thornton, medical directorb,
  5. J Gedis Grudzinskas, professorc,
  6. Brian A Lieberman, medical directord
  1. a Royal Free and University College Medical School, Department of Child Health, Royal Free Campus, London NW3 2PF
  2. b Centre for Assisted Reproduction (CARE), Park Hospital, Nottingham NG5 8RX
  3. c Department of Obstetrics and Gynaecology, St Bartholomew's and Royal London School of Medicine and Dentistry, Royal London Hospital, London E1 1BB
  4. d Department of Reproductive Medicine, St Mary's Hospital, Manchester M13 0JH
  1. Correspondence to: Dr Sutcliffe
  • Accepted 9 November 1998

Intracytoplasmic sperm injection is often successful for treatment of male infertility; over 20 000 children have been born as a result.1 This bypassing of natural barriers to sperm selection has raised concerns about the children conceived.2 We report a population control study of children born in the United Kingdom as a result of this treatment.

Subjects, methods, and results

Children between 12 and 24 months old who had been singleton births were identified from a list of couples who had received the treatment and their parents were invited to participate; 123 of 137 families (90%) agreed. Control children, conceived naturally, were recruited from associated nurseries (105/123) or were social peers of cases (18/123). Altogether, 123 children born after intracytoplasmic sperm injection (study children) and 123 control children were seen. Children were matched for social class, maternal educational level, region, sex, and race but not maternal age. Multiple births were excluded to avoid confounding factors. Primary outcome measures were developmental scoring on the Griffiths scales of mental development3 and rates of congenital abnormalities. The Griffiths scales are an objective method of assessing development which uses five subscales. All subscales have a normal mean score of 100 (75-125, SD 1).

Clinical data obtained included date and type of delivery, birth weight, gestation, resuscitation required, duration and reason for admission to neonatal unit (if admitted), and details of ventilatory support. Congenital abnormalities were classed according to the ICD-10 (international classification of diseases, 10th revision). Sociodemographic data obtained included date of birth, sex, and age in months. Information obtained about parents included date of birth, social class, occupation, smoking status, alcohol intake, marital or support status (that is, the number of full time carers in the household), type of housing, education, race, and mother's gravidity and parity.

One observer (AGS) assessed all children. The mean age at assessment was 17.5 months. There was similarity across sociodemographic factors, although mothers of study children were more likely to be 35 years or older (P<0.001). Study children were more likely to have been born earlier (38.83 weeks v 39.59, P<0.01), to be of lower mean birth weight (3167 g v 3365 g, P<0.01), and to have been born by caesarean section (44 v 26, P<0.05), but neonatal admission rates were similar (P>0.7).

The mean mental age (17.3 months for study infants v 17.6 for controls) and the mean Griffiths quotient (101 for study infants v 102 for controls) were comparable (table). Difference in the eye-hand coordination subquotient persisted despite adjustment for gestation (P<0.05). However, scores on all subscales were normal for control and study children.

The number of study infants with minor congenital anomalies (14) was higher in comparison with controls (9), but not significantly. The number of children with a major congenital malformation was comparable (6 study v 5 controls). Malformations found in the study infants were scrotal fusion, undescended testis, exomphalos, congenital cataract, and congenital hip dislocation. In the control group malformations were buphthalmos, horseshoe kidney, cleft lip, cleft palate, and ventricular septal defect.

Chronological age and mean scores on Griffiths scales of mental development for infants conceived by intracytoplasmic sperm injection (study) and infants conceived naturally (control)

View this table:


In our study a narrow age range was used to match cases and controls for assessment with the revised Griffiths scales (standardised on the UK population), which have a 91.3% power to detect a five point difference between groups. Children conceived naturally were chosen as controls in preference to another in vitro fertilisation group as being a more appropriate standard by which to test normal development. Control children were not matched by parity, history of infertility, or mode of delivery, which may be predictive of adverse outcome in later childhood. Our 90% follow up compares favourably with 25% in a Belgian study.4 Spontaneous abortions during the study period were not documented, possibly distorting the rates of congenital abnormality.

The difference in eye-hand coordination subscales is unlikely to be of functional significance; scores in both groups were normal. Congenital abnormality rates were consistent with national data in the United Kingdom (overall rate 5%)5 but a larger study might identify a true increase in minor anomalies.

We are recruiting further children, and follow up at 5 years of age is planned.


We would like to thank the following participating centres: Assisted Conception Unit, Ninewells Hospital, Dundee; BUPA Roding Hospital, Essex; Holly House Hospital, Essex; Churchill Clinic London; Lister Hospital, London; Midland Fertility Services, Birmingham; Manchester Fertility Services, St Mary's Hospital, Manchester; NURTURE, Department of Obstetrics and Gynaecology, Nottingham University; CARE, Park Hospital, Nottingham; General Infirmary at Leeds; Assisted Conception Unit, St James's Hospital, Leeds; and London Women's Clinic, Assisted Conception Unit, University College Hospital. Children from the following nurseries acted as controls: Tiger Tots (Aberdeen); University of Nottingham Crèche; Queen's Medical Centre Nottingham Crèche; the Surgery, 2 Ritchie Street, London; and the Mouse Hole Nursery (Middlesex Hospital Nursery). We are grateful for the advice of Dr Kerryn Saunders, Monash University, Melbourne, Australia.


  • Contributors All authors participated in the design of this study, recruitment of patients, analysis of data, and writing of the paper. AGS assessed each child and is guarantor for the study. JL undertook the statistical analysis. BT advised on epidemiological aspects of the design of the study. ST, JGG, and BAL arranged the recruitment of patients and assisted with ethical approval and raising funding.

  • Funding The study was supported by a grant from the Sir Halley Stewart Trust Fund. Additional grants were received from the Manchester Fertility Services Trust, the Burgess Bequest, Cook UK, Serono UK, Organon Laboratories, Ferring UK, IBSA Biochemique, and Smith Industries.

  • Competing interests None declared.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
View Abstract