Correspondence

Relationship between Cerebrospinal Fluid Glucose and Serum Glucose

N Engl J Med 2012; 366:576-578February 9, 2012

Article

To the Editor:

Levels of glucose in the cerebrospinal fluid (CSF) are used to discriminate bacterial meningitis from viral meningitis.1,2 Children with bacterial meningitis typically have low levels of CSF glucose because of glycolysis by both white cells and the pathogen and impaired CSF glucose transport. The level of CSF glucose is typically interpreted in relation to that of serum glucose, since glucose passes across the blood–brain barrier.3 Standard reference texts cite the normal ratio of CSF to serum glucose as 0.6,4 although this ratio has not been rigorously investigated.5

Our retrospective cross-sectional study included children 18 years of age or younger who underwent lumbar puncture for clinical indications in the emergency department of a single tertiary-care hospital between January 1, 1995, and September 15, 2011. The institutional review board at Children's Hospital Boston approved the study, with waiver of the requirement for informed consent.

We included children undergoing lumbar puncture whose levels of both CSF and serum glucose had been obtained and received by the hospital laboratory within 60 minutes of each other. Children with conditions that could alter the normal ratio of CSF to serum glucose (e.g., bacterial meningitis, aseptic meningitis, ventricular shunt, and diabetes) were excluded.

Locally weighted scatterplot smoothing (bandwidth, 0.8) was used to explore the relationship between serum and CSF glucose levels. Linear regression (with R² as a measure of model fit) was then used to further explore this relationship in all patients and in subgroup analyses.

Overall, CSF glucose levels were obtained from 19,374 children, and serum glucose levels were obtained concomitantly from 3963 (20%). Of these patients, 158 were excluded, including 18 who had bacterial meningitis, 44 who had aseptic meningitis, 87 who had a ventricular shunt, and 9 who had diabetes. Among the 3805 study patients, the median age was 2.2 months (interquartile range, 1.1 to 13.3). Levels of CSF glucose rose with increasing levels of serum glucose (Figure 1Figure 1The Relationship between Levels of Glucose in Serum and Cerebrospinal Fluid in Children 18 Years of Age or Younger.). Although the ratio decreased minimally for children with a serum glucose level of more than 240 mg per deciliter, this relationship remained highly linear across the range of serum glucose values. For every increase of 1.0 mg per deciliter in serum glucose levels, CSF glucose levels increased by 0.56 mg per deciliter (95% confidence interval, 0.56 to 0.57; R²=0.94). The relationship between CSF glucose and serum glucose was similar in all age groups and in the following subgroups: patients for whom data on serum glucose levels were received by our laboratory before data on CSF glucose levels were received, patients without a traumatic lumbar puncture, and children without CSF pleocytosis (see Table 1 in the Supplementary Appendix, available with the full text of this letter at NEJM.org).

We observed a linear relationship between the levels of serum and CSF glucose. This relationship remained constant in each of the patient subgroups. Specifically, the normal level of CSF glucose should be approximately 60% of the level of serum glucose. Knowledge of this relationship may help clinicians in interpreting CSF glucose concentration for the evaluation of bacterial meningitis in children.

Lise E. Nigrovic, M.D., M.P.H.
Amir A. Kimia, M.D.
Children's Hospital Boston, Boston, MA
lise.nigrovic@childrens.harvard.edu

Samir S. Shah, M.D., M.S.C.E.
Cincinnati Children's Hospital Medical Center, Cincinnati, OH

Mark I. Neuman, M.D., M.P.H.
Children's Hospital Boston, Boston, MA

Disclosure forms provided by the authors are available with the full text of this letter at NEJM.org.

5 References

1. 1

   Spanos A, Harrell FE Jr, Durack DT. Differential diagnosis of acute meningitis: an analysis of the predictive value of initial observations. JAMA 1989;262:2700-2707
   CrossRef | Web of Science | Medline

2. 2

   Bonsu BK, Harper MB. Differentiating acute bacterial meningitis from acute viral meningitis among children with cerebrospinal fluid pleocytosis: a multivariable regression model. Pediatr Infect Dis J 2004;23:511-517
   CrossRef | Web of Science | Medline

3. 3

   Menkes JH. The causes for low spinal fluid sugar in bacterial meningitis: another look. Pediatrics 1969;44:1-3
   Web of Science | Medline

4. 4

   McMillan JA, DeAngellis CD, Feigin RD, Warshaw JB, eds. Oski's pediatrics: principles and practice. 3rd ed. Philadelphia: Lippincott Williams & Wilkins, 2006.
   

5. 5

   Marks V. True glucose content of lumbar and ventricular cerebrospinal fluid. J Clin Pathol 1960;13:82-84
   CrossRef | Web of Science | Medline