Dexamethasone-induced leucocytosis in pregnancy


  • Fiona C. Denison,

    Senior House Officer
    1. Simpson Memorial Maternity Pavilion, Royal Infirmary of Edinburgh NHS Trust, Edinburgh
    Search for more papers by this author
  • Catherine L. Elliott,

    Research Fellow
    1. Department of Obstetrics and Gynaecology, University of Edinburgh, Edinburgh
    Search for more papers by this author
  • Euan M. Wallace

    Lecturer , Corresponding author
    1. Department of Obstetrics and Gynaecology, University of Edinburgh, Edinburgh
      Correspondence: Dr E. M. Wallace, Department of Obstetrics and Gynaecology, Monash University, 246 Clayton Road, Clayton, Victoria 3168, Australia.
    Search for more papers by this author

Correspondence: Dr E. M. Wallace, Department of Obstetrics and Gynaecology, Monash University, 246 Clayton Road, Clayton, Victoria 3168, Australia.


The effect of intramuscular dexamethasone administration in late pregnancy on the maternal peripheral white cell count was examined in 20 women. The mean total white cell count increased from a baseline of 11–3 × 109/L (SD 2–3) to 16–2 × 109/L (SD 4–6) on day 1, normalising thereafter. This 43% increase represented composite changes in the neutrophil and lymphocyte counts which, on average, increased by 62% and decreased by 22%, respectively. It is concluded that prenatal dexamethasone induces a significant neutrophilia on the first day following administration. This information may be helpful when monitoring for infection.

Antenatal glucocorticoid administration as prophylaxis for neonatal respiratory distress syndrome is a proven and established component of obstetric care for women at risk of delivery before 34 weeks of gestation1. It is also well recognised that spontaneous preterm labour and preterm prelabour spontaneous rupture of the membranes (PPROM) is significantly associated with infection2. While there is no evidence that administration of glucocorticoids significantly increases the risk of perinatal infection in women with PPROM, such therapy may complicate the management of these women in a more subtle manner. In the absence of infection a conservative management approach allowing prolongation of pregnancy and unproved fetal maturity significantly improves neonatal outcome3. However, when infection is identified it may be clinically appropriate to effect delivery in the hope of improving outcome. Thus, monitoring for infection, particularly in the acute phase, is an important element of the management of women with PPROM. Many and varied strategies have been employed to improve the early detection of infection and so guide the timing of delivery, but with very limited success4. Indeed, in the absence of specific and sensitive tests for infection, serial assessment of the maternal peripheral white cell count remains a useful and popular component of strategies4.

It has long been appreciated that, in nonpregnant individuals, the acute administration of exogenous corticosteroids induces a transient leucocytosis5,6. Were similar effects to occur in pregnancy the management of women with PPROM may be complicated. If a corticosteroid-induced leucocytosis is interpreted as being secondary to infection then inappropriate intervention may ensue. Conversely, if an infection-related leucocytosis is wrongly attributed to corticosteroid administration then a necessary delivery may be delayed.

However, to our knowledge there are no published studies quantifying either the extent or the duration of a glucocorticoid-induced leucocytosis in pregnancy, information which clearly would be useful in the management of women receiving glucocorticoid prophylaxis. Pregnancy itself is associated with significant haematological changes, including a moderate leucocytosis. The neutrophil count is almost double that observed in nonpregnant women7. It is conceivable that this response to pregnancy may alter the leucocytic effects characteristic of glucocorticoids in nonpregnancy and we therefore wanted to assess whether changes in the circulating white cell number occur following corticosteroid administration in pregnancy. Further, if glucocorticoids did induce a leucocytosis then we wanted to detail the extent and duration of the effects, thereby assisting our management of women in whom infection was a significant risk, such as those with PPROM. This study aimed to investigate the effects of dexamethasone on the leucocyte count in pregnant women.


The study reported here formed part of a larger, ongoing study of dexamethasone administration which received ethical approval from the Lothian Paediatric and Reproductive Medicine Research Ethics Subcommittee. Twenty women (median age 30 years, range 15–40) were recruited and gave informed consent to the study. Each woman had been assessed by their attending obstetric team and glucocorticoid treatment was prescribed, in anticipation of impending preterm delivery. Women with objective evidence of infection were not recruited to the study. Each woman received two intramuscular doses of 12 mg dexamethasone 12 hours apart in line with our hospital policy. Blood was collected before the administration of the first dose, and then daily in the morning for the following four days. The blood was analysed for haemoglobin, leucocyte count, including differential count, and platelet count using an automated NE 800 Sysmex counter.

Changes in the serial haematological parameters were subjected to analysis of variance (ANOVA) for repeated measures using Statview 4.1 (Abacus, Berkeley, California, USA). Changes were considered significant when P < 0.05.


The median gestation at treatment was 31 weeks (range 23–34). The indications for treatment in the 20 women were: seven for pregnancy-induced hypertension, five for threatened preterm labour (three of which were twin pregnancies), four for antepartum haemorrhage, two for PPROM, and one each for intrauterine growth retardation and placenta praevia. Table 1 shows the haematological parameters before and after dexamethasone administration. The mean total white cell count rose significantly, by 43%, on day one and returned to baseline on day two. This rise represented composite changes in the neutrophil and lymphocyte counts. The mean neutrophil count rose by 62% (range −26% to +147%) on day one, normalising on day two, and the mean lymphocyte count fell significantly (by 22%) on day one with a significant rebound elevation on day two, an increase of 24% compared with baseline, normalising on day three.

Table 1.  Changes in haematology parameters following the administration of dexamethasone in 20 women. Values are shown as mean (SD). NS = not significant
Haematology parameterBefore dexamethsoneDay 1Day 2Day 3Day 4P
Haemoglobin (g/L)112 (10.1)108 (11.5)104 (10.9)107 (10.4)111 (12.-3)<0.01
Platelet count (×109/L)226 (63.9)228 (69.7)220 (68.9)232 (62.3)229 (63.7)NS
Total white cell count (× 109/L)11.3 (2.3)16.2 (4.6)11.8 (2.8)11.3 (3.0)11-4 (3.5)< 0.0001
Neutrophil count (× 109/L)8.7 (1.9)14.0 (4.4)8.8 (2.4)8.7 (2.6)8-9 (2.7)< 0.0001
Lymphocyte count (× 109/L)2.0 (0.6)1.6 (0.6)2.5 (0.6)2.1 (0.9)1-9 (1.0)< 0.0001

The mean haemoglobin level fell significantly over days one to three with the nadir on day two. The maximum percentage fall was 7%, recovering to normal on day four. There were no significant changes in the platelet count nor in other white cell parameters.


Our data confirm that, as in nonpregnancy, dexamethasone induces a transient neutrophilia when given antenatally, with a mean increase of 62% in the count, resulting in a mean increase of the total white cell count of 43%. This increase was only apparent on day one following dexamethasone administration. Levels normalised by the following day. Various explanations have have been proposed for this corticosteroid effect, including a drug-induced increase in the rate at which the cells enter the blood or a decrease in the rate that the cells leave the blood5,6. While the exact mechanism(s) remain unknown, monitoring of the maternal peripheral white cell count continues to be widely used as an indicator of infection4 thereby guiding management. Our data show that the raised white cell count attributable to dexamethasone, while variable in extent between individuals, is apparent only on the day after administration. In all but one woman the white cell count on day two was lower than that on the first day after steroid treatement. The exception was a woman who showed a fall in her white cell count on day one (by 26%) which returned to baseline on day two. We would therefore suggest that if an elevated white cell (or neutrophil) count is found beyond day one following corticosteroid administration then it is suggestive of infection. Our study did not examine whether steroid administration alters the expected host response to infection, and we are not able to consider this potential complication of corticosteroid treatment.

We recognise that many centres use betamethasone or other corticosteroids in preference to dexamethasone and, while it is likely that the effects will be similar, we did not test this specifically. Furthermore, while pharmacokinetic data would support the administration of dexamethasone 12 hourly8, a commonly used regimen is two corticosteroid doses 24 hours apart. It is possible that the more prolonged corticosteroid exposure afforded by the latter regimen may induce a more prolonged elevation of the white cell count than we report here. It would be worthwhile therefore for centres using the white cell count as a marker of infection to consider assessing the specific leucocyte effects of their local corticosteroid regimen.

The number of women involved were too small for us to properly assess effects of parity, gestation, indication for treatment, age, smoking habits and drug history—factors which may influence the leucocyte count9,10. There were no obvious confounding effects but they cannot be excluded.

The changes in haemoglobin are interesting and, to our knowledge, have not been reported before. It is likely that they were secondary to changes in intravascular volume, but against this explanation would be the lack of a concomitant changes in the platelet count. Nonetheless, the extent of the change (maximum fall of 7%) and the return to baseline by day 4 suggest that this finding is of limited clinical relevance.