Original Article | Published:

Dexamethasone Therapy and Candida Sepsis in Neonates Less Than 1250 Grams


OBJECTIVE: To determine whether dexamethasone use increases the risk for Candida sepsis (CS) in very low birth weight premature infants (<1250 g).

DESIGN: Retrospective chart review of all infants with a birth weight <1250 g, admitted to the neonatal intensive care unit of the MetroHealth Medical Center, Cleveland, Ohio between January 1, 1996 and December 31, 1999. Infant groups with (n=65) and without (n=229) CS were compared.

RESULTS: Two hundred and ninety four infants with a birth weight <1250 g were identified. CS was diagnosed at a median age of 18 days, and 6 of 65 (10%) infants died directly from Candida-related complications. Candida albicans (n=30, 60%) and Candida parapsilosis (n=14, 25%) were the predominant isolates. Use of dexamethasone in infants at risk for chronic lung disease before 14 days of age (p=0.001), duration of antibiotics (p=0.001), and total duration of parenteral nutrition and intralipid (p=0.0001) were all significantly greater in infants who developed CS. Regression analysis showed that duration of antibiotics before the diagnosis of Candida infection (r2=0.69, p=0.0002) and duration of dexamethasone (r2=0.93, p=0.0002) correlated with Candida infection. Early dexamethasone use was also related to the age at diagnosis of Candida infection (r2=0.51, p=0.01).

CONCLUSIONS: Dexamethasone therapy and prolonged duration of antibiotics are associated with Candida infection in premature infants.


Candida sepsis (CS) has demonstrated an increased incidence, especially in very low birth weight (VLBW) infants. From only 40 reported cases (1964 to 1984) of systemic CS in VLBW infants, the reported incidence increased from 4.5% in the 1980s1,2,3,4 to 10% in extremely low birth weight (ELBW) infants in the 1990s.5,6 This increase in the incidence of CS in ELBW infants may be a consequence of increased survival of extremely premature infants. Impaired immune defense mechanisms,2,7 frequent and prolonged antibiotic use, and multiple invasive procedures such as long-term intravenous central catheters3,4,8 have been implicated in the increased predisposition of premature infants in developing Candida infection.

Improved survival rates among infants with birth weight <750 g is associated with an increased incidence of chronic lung disease. Dexamethasone is widely used in this population of infants in an attempt to minimize the magnitude and duration of ventilatory support, and to decrease pulmonary morbidity. However, dexamethasone use in premature infants is associated with numerous metabolic, cardiovascular, and infectious complications.9,10 Dexamethasone is known to alter phagocytosis,11 reduce the circulating lymphocyte subpopulations,12 inhibit the cytokine responses,13 and impair cell-mediated immunity in premature infants.14 Because both humoral and cell-mediated immune responses are required in the host defense against CS, we hypothesized that dexamethasone use would increase incidence of CS in preterm infants. We further hypothesized that prolonged antibiotic therapy in infants who receive dexamethasone would also lead to an increased predisposition of these babies to Candida infection. The above hypotheses were tested using a retrospective study design in preterm infants with a birth weight <1250 g.


Study Design

A case–control study was performed to determine the relationship between dexamethasone use and CS in preterm infants. All infants with a birth weight <1250 g and admitted to the neonatal intensive care unit (NICU) of the MetroHealth Medical Center (MHMC), Cleveland, Ohio between January 1, 1996 and December 31, 1999 were studied.

Study Population

Inborn infants with a birth weight <1250 g and born at the MHMC during the study period were included. Transported infants to our NICU were excluded. Infants who died during the neonatal period and those with lethal congenital and chromosomal anomalies were also excluded. The investigators (D. K. and P. P.) independently reviewed the medical records, placental pathology, microbiology, and available autopsy data of infants who died during the neonatal period to ensure that the immediate cause of their death was not due to Candida sepsis. Neonatal and laboratory databases were utilized to identify possible cases of Candida sepsis during the study period. CS was defined as isolation of Candida species from a blood culture or'from a catheterized urine sample. Controls included all other infants with a birth weight <1250 g, born during the study period and without microbiologic evidence of Candida infection.

Data Collection

Infants with and without CS were identified. Standardized forms were used to collect perinatal characteristics (birth weight, gestation, prenatal steroid and antibiotic use, and percentage of mothers with cerclage), postnatal risk factors (duration of assisted ventilation, antibiotics, dexamethasone, parenteral nutrition, intralipid, central line(s), and feeding), and outcome (length of stay and survival) characteristics. Infant age at initiation of dexamethasone, duration, and number of courses of dexamethasone for each infant was noted. Infants in whom dexamethasone was started before 14 days of age were designated as early dexamethasone group and those in whom it was started after 14'days as late dexamethasone group. Age at diagnosis, type of Candida species, days needed to sterilize blood or urine culture, evidence of systemic seeding, serum creatine concentration, cumulative dose of amphotericin, survival and mortality data directly attributable to CS were also documented.

Statistical Analysis

Infants with and without CS were compared. Group comparisons were made for infants with birth weight <1250 g. Statistical analysis was done using GB STAT 6.5 (Dynamic Microsystems, Silver Spring, MD). For discrete variables, chi-square tests and for continuous variables Student t test or Mann-Whitney U test was performed. Multivariate logistic regression analysis was done to control for the effect of birth weight and gestation age. Pearson correlation was performed to determine the relationship between the age of initiation of dexamethasone, duration of antibiotics before the onset of CS, and'age when CS was diagnosed. Data presented is either mean±SD or median with interquartile range. Statistical significance was defined a priori as a p value <0.05.


From a total of 294 infants who met the inclusion criteria with birth weight <1250 g, 65 (22%) developed CS. The majority of the infants (n=60, 84%) with CS weighed <1000 g. Infants with birth weight <750 g accounted for (n=35) 53% of cases. The diagnosis was established by isolation of Candida species in the blood or urine culture at a median age of 18 days (25th to 75th percentile, 12 to 27'days). Candida albicans was the predominant isolate (n=41, 60%) followed by Candida parapsilosis (n=14, 25%). The blood cultures became sterile at a median age of 7 days (25th to 75th percentile, 5 to 20 days) after initiation of amphotericin therapy. Renal function and serum creatinine concentrations remained normal during the course of amphotericin therapy in all cases.

Candida Sepsis in Infants With a Birth Weight <1250 Grams

Infants who developed CS were smaller at birth (p=0.018) than controls. Use of prenatal steroids, prenatal antibiotics, cerclage, and route of delivery were similar between the groups (Table 1). The survival rate between the group of infants with and without CS was similar. Duration of antibiotic use during first month, dexamethasone use before 14 days of age, indwelling central lines, mechanical ventilation, total parenteral nutrition (TPN) and intralipid infusion, and hospital stay were all significantly longer in infants with CS compared to those without Candida sepsis. CS was identified as the chief cause of death in six infants (10%) with Candida infection (Table 2). Logistic regression analysis was performed using CS as a dependent variable. After controlling for birth weight, duration of TPN, intralipids, and central lines, days on antibiotics during the first month and initiation of dexamethasone therapy before 14 days of age were found to be significantly associated with CS (Table 3).

Table 1 Perinatal Characteristics of Infants With Birth Weight <1250 Grams
Table 2 Postnatal Risk Factors and Outcome Characteristics (<1250 Grams)
Table 3 Logistic Regression Analysis With Candida Sepsis as a Dependent Variable (<1250 Grams)

Dexamethasone and Antibiotic Use and Candida Sepsis

A total of 41 of 65 (63%) infants with CS received dexamethasone. Of'the 41 infants, 22 (54%) were placed on steroids before the onset of CS. Dexamethasone was begun at a median age of 15 days (12 to 30 days, 95% CI) and CS was diagnosed at a median age of 26 days (18 to 42 days, 95% CI). The median duration of dexamethasone therapy was 40 days (30 to 55 days, 95% CI). On the contrary, infants who did not develop Candida sepsis were placed on dexamethasone at a median age of 20.5 days (15 to 26, 95% CI). Duration of dexamethasone before the confirmation of CS correlated significantly with age the candidal infection was diagnosed (r2=0.93, p=0.0002, Figure 1). In addition, early dexamethasone use correlated significantly with the age of onset of CS (r2=0.51, p=0.01). Duration of antibiotics before the diagnosis of CS also correlated with the age infants developed CS (r2=0.69, p=0. 00002, Figure 2).

Figure 1

Correlation between duration of dexamethasone before Candida sepsis and age at diagnosis of Candida sepsis. y=1.36x+10.3; r2=0.93; p=0.0002.

Figure 2

Correlation between duration of antibiotics before Candida sepsis and age at diagnosis of Candida sepsis. y=0.97x+3.0; r2=0.69; p=0.0002.


This is the first study to show an association between dexamethasone therapy and Candida infection in VLBW infants. Our data shows that 25% of the study infants developed Candida infection, and the majority of these (83%) infants were below the birth weight of 1000 g. In addition to the already well-described risk factors for Candida infection, we noted that dexamethasone therapy and use of dexamethasone during the first 2 weeks after birth is associated with the development of Candida infection in high-risk premature infants.

Given the inherent, albeit transient, deficiencies in the humoral and cellular immunity in preterm infants, we predicted that dexamethasone therapy by virtue of its effects on neutrophils and lymphocytes enhances the risk of Candida sepsis in preterm infants. Although, dexamethasone therapy in premature infants has been shown to increase bacterial sepsis and meningitis,10 there are no reports of increased incidence of Candida sepsis in these infants. Only a single report exists in the literature examining the association between disseminated Candida infections and intravenous hydrocortisone in preterm infants. These authors report that a significantly higher number of infants (10 of 17, 59%) receiving hydrocortisone developed disseminated Candida infection.15 In our study, 54% of'preterm infants were started on dexamethasone before they developed Candida infection. The interval between dexamethasone and diagnosis of Candida sepsis was approximately 2 weeks. The linear positive correlation between the duration of dexamethasone and confirmation of Candida infection supports the argument that dexamethasone use in preterm infants is associated with Candida sepsis. Furthermore, use of steroids in the first 2 weeks after birth is significantly associated with development of Candida sepsis. Alterations in blood cell counts, peripheral blood lymphocyte subpopulations, impairment of the phagocytic ability of neutrophils and inhibition of the cytokine production by the mononuclear cells16,17,18,19,20 have been shown to be associated with the use of dexamethasone. These data suggest that dexamethasone therapy impact the immune competence of premature infants and, thus, predispose them to fungal as well as bacterial infections. It is also likely that dexamethasone treatment in preterm infants placed on empirical and prolonged course of antibiotics further accentuates their predisposition to acquire infection by Candida.

Prolonged use of antibiotics has been implicated in the acquisition of CS in premature infants.21,22 We noted that a higher proportion of infants in our study received multiple antibiotic courses during the first 30 days after birth for presumed sepsis. In addition, the duration of antibiotics before the diagnosis of Candida infection was significantly greater in infants who subsequently developed CS (Figure 2). Although we did not identify association with any particular group of antibiotics, Benjamin et al.23 have shown that candidemic patients had more exposure to third generation cephalosporins. Multiple antibiotics and for a prolonged period have been proposed to suppress normal flora leading to proliferation of Candida.24 Antibiotics have also been thought to remove antifungal substances released by other organisms, compete for nutrients, and directly stimulate Candida. In preterm infants colonized with Candida, empirical antibiotic therapy when used in conjunction with dexamethasone enhances tissue susceptibility to invasion and dissemination of fungi.25,26

In summary, our data show that dexamethasone use in patients at risk for chronic lung disease and prolonged duration of antibiotics during the neonatal period is associated with CS in VLBW infants. Prophylaxis with antifungal agents, judicious use of antibiotics, and an expedited dexamethasone regimen for infants at risk for chronic lung disease together may reduce the incidence of CS in VLBW premature infants.


  1. 1

    Johnson DE, Thompson TR, Green TP, Ferrari P . Systemic Candidiasis in very low birth weight infants (&lt;1500 grams) Pediatrics 1984 73: 138–43

  2. 2

    Baley JE, Kliegman RM, Boxerbaum B, Fanaroff AA . Fungal colonization in the very low birth weight infant Pediatrics 1986 78: 225–32

  3. 3

    Baley JE, Kliegman RM, Fanaroff AA . Disseminated fungal infections in very low birth weight infants: clinical manifestations and epidemiology Pediatrics 1984 73: 144–52

  4. 4

    Faix RG, Kovarik SM, Shaw TR, Johnson RV . Mucocutaneous and invasive Candidiasis among very low birth weight (&lt;1500 grams) infants in intensive care nurseries: a prospective study Pediatrics 1989 83: 101–7

  5. 5

    Huang YC, Li CC, Lin TY, Lien et al. Association of fungal colonization and invasive disease in very low birth weight infants Pediatr Infect Dis J 1998 17: 819–22

  6. 6

    Kossoff EH, Buescher ES, Karlowicz MG . Candidiasis in a neonatal intensive care unit: trends during fifteen years and clinical features of 111 cases Pediatr Infect Dis J 1998 17: 504–8

  7. 7

    Bektas S . Decreased adherence, chemotaxis and phagocytic activities of neutrophils from preterm neonates Acta Pediatr Scand 1990 79: 1031

  8. 8

    Blaschke-Hellmessen R . Vertical transmission of Candida and its consequences Mycoses 1998 41: 2 31–6

  9. 9

    Stark AR, Carlo WA, Tyson JE et al. Adverse effects of early dexamethasone in extremely-low-birth-weight infant. National Institute of Child Health and Human Development Neonatal Research Network N Engl J Med Jan 11, 2001 344: 2 95–101

  10. 10

    Stoll BJ, Temprosa M, Tyson JE et al. Dexamethasone therapy increases infection in very low birth weight infant Pediatrics Nov 1999 104: 5 e63

  11. 11

    Lortie C, King GM, Adamson IY . Effects of dexamethasone on macrophages in fetal and neonatal rat lung Pediatr Pulmonol 1990 8: 3 138–44

  12. 12

    Parimi PS, Birnkrant DJ, Rao LV, Diaz G, Moore JJ . Effect of dexamethasone lymphocyte subpopulations in premature infants with bronchopulmonary dysplasia J Perinatol 1999 19: 347–51

  13. 13

    Bessler H, Straussberg R, Gurary N, Aloni D, Sirota L . Effect of dexamethasone on IL-2 and IL-3 production by mononuclear cell in neonates and adults Arch Dis Child Fetal Neonatal Ed 1996 Nov 75: 3 F197–201

  14. 14

    Scwarze J, Bartmann P . Influence of dexamethasone on lymphocyte proliferation in whole blood cultures of neonates Biol Neonate 1994 65: 5 295–301

  15. 15

    Botas CM, Kurlat I, Young SM, Sola A . Disseminated Candidal infections and intravenous hydrocortisone in preterm infants Pediatrics June 1995 95: 6 883–7

  16. 16

    Xanthou M, Valassi-Adam E, Kintsonidou E, Matsaniotis N . Phagocytosis and killing ability of Candida albicans by blood leucocytes of healthy term and preterm babies Arch Dis Child 1975 50: 1 72–5

  17. 17

    Katikaneni LD, Mathur S . Humoral immunity to Candida albicans (anti-candida antibody titers) in premature infants Am J Reprod Immunol 1990 May 23: 1 1–3

  18. 18

    Shareef MJ, Myers TF, Mathews HL, Witek-Janusek L . Reduced capacity of neonatal lymphocytes to inhibit growth of Candida albicans Biol Neonate 1999 75: 1 31–9

  19. 19

    Bektas S, Goetze B, Speer CP . Decreased adherence, chemotaxis and phagocytic activities of neutrophils of preterm neonates Acta Pediatr Scand 1990 Nov 79: 11 1031–8

  20. 20

    Peng CT, Lin HC, Lin YJ, Tsai CH, Yeh TF . Early dexamethasone therapy and blood cell count in preterm infants Pediatrics Sep 1999 104: 3 Pt 1 476–81

  21. 21

    Singh K, Chakrabarti A, Narang A, Gopalan S . Yeast colonization and fungemia in preterm neonates in a tertiary care unit Indian J Med Res Nov 1999 110: 169–73

  22. 22

    Saiman L, Ludington E, Pfaller M . Risk factors for Candidemia in neonatal intensive care unit patients Pediatr Infect Dis J 2000 19: 4 319–24

  23. 23

    Benjamin DK, Ross K, Mickenney R, Auten R, Fischer RG . When to suspect fungal infection in neonates; a clinical comparison of Candida albicans and Candida parapsilosis fungemia with coagulase negative staphylococcal bacteremia Pediatrics 2000 106: 712–8

  24. 24

    Seelig MS . The role of antibiotics in the pathogenesis of Candidal infections Am J Med 1966 40: 887–917

  25. 25

    Mullett MD, Cook EF, Gallagher R . Nosocomial sepsis in the neonatal intensive care unit J Perinatol Mar–Apr 1998 18: 2 112–5

  26. 26

    Leibovitz E, Iuster-Reicher A, Amitai M, Mogilner B . Systemic Candidal infection associated with use of peripheral vascular catheter in neonates — a 9-year experience Clin Infect Dis Feb 1992 14: 2 485–91

Download references

Author information

Correspondence to Prabhu Parimi MD.

Rights and permissions

Reprints and Permissions

About this article

Further reading