Laboratory Diagnosis in Neurology, 1 Ed.

Lactate and Glucose

H. Reiber


General Remarks


The drop in CSF glucose observed in bacterial and tuberculous meningitis was described as early as 1893 by Lichtheim. In 1924, Nishimura demonstrated that the drop in pH observed in purulent meningitis is caused by an increase in lactate (lactic acidosis). The inverse relationship between glucose and lactate was recognized in 1925 by Kilian, and it soon became clear (Garcia et al., 1928) that the lactate concentration in CSF is a better marker of the progress of the infection than is the CSF glucose concentration. Nevertheless, glucose determination still dominates worldwide, and the switch to lactate analysis is only slowly taking place.

Forms of Lactate

L-Lactate. L-Lactate is the end product of anaerobic glucose metabolism and, depending on the availability of oxygen, is further utilized by oxidation in the citric acid cycle or by gluconeogenesis in the Cori cycle. Lactate is the link between aerobic and anaerobic metabolism. Differential diagnostic evaluation of hyperlactatemia with and without transition to lactic acidosis is an important aspect of internal medicine and emergency medicine (Thomas, 2005).

In CSF diagnostic analysis, L-lactate determination plays an important role in distinguishing between bacterial and viral meningitis. For the prognosis of various brain injuries after craniocerebral trauma or subarachnoid hemorrhage, serial analysis of L-lactate may be used as a prognostic marker if CSF lactate levels are moderately increased.

D-Lactate. D-Lactate is also generally formed in mammals, but as a proportion of the total lactate concentration it is less than 1%. By contrast, D-lactate production is dominant in bacteria, although depending on the species it may be more or less pronounced in comparison to L-lactate production. There is no established diagnostic role for D-lactate analysis.

Physiology and Pathophysiology

Physiological metabolism. In a healthy brain, L-lactate is produced primarily by astrocytes, where it is a natural metabolite (Walz and Mukerji, 1988). Lactate is metabolized aerobically by neurons.

Pathophysiological metabolism. Damaging processes of various origins—in particular, traumatic, ischemic, and inflammatory insult—generally result in increased activity of the brain, not only in the neurons, but also in the astrocytes, and this is followed by the release of lactate. In patients with craniocerebral trauma, lactate production seems to be a physiological astrocytic response mediated by the release of glutamate. The capacity of the neurons for aerobic utilization of the lactate released by the astrocytes is exceeded when brain activity increases under pathological conditions, and even the neurons turn to glycolysis for energy.

In severe bacterial meningitis, the increased lactate levels in the CSF are mainly due to the mechanisms described above, and only to a minor degree to bacterial production of D- and L-lactates and granulocytic production of L-lactate in the CSF (Wellmer et al., 2001; Prange, 2004). The diseases in Table 5.11 showed the highest mean granulocyte counts when the mean lactate values were also high (Felgenhauer and Beuche, 1999); nevertheless, it would be wrong to conclude from this that there is a direct connection. Cell counts may be high (12 700/μL) (Pseudomonas spp.) with a low lactate concentration (10 mmol/L) or may be low (1194/μL) (pneumococci) with a high lactate concentration (29 mmol/L). This is supported by test results from a further case of pneumococcal meningitis, which showed a cell count of 17/μL together with 3.4 mmol/L lactate (Prange, 2004).

The raised lactate levels in the CSF are primarily due to posttraumatic and postinfectious neuronal excitation mediated by glutamate (Walz and Mukerji, 1988; Prange, 2004).


L-lactate determination is an established means of differentiating between bacterial and nonbacterial meningitis. A moderate increase in lactate is observed in many inflammatory, vascular, metabolic, and neoplastic diseases of the brain and meninges, but it is less relevant for differential diagnosis (Thomas, 2005). When analyzed serially, it may help to establish a clinical prognosis.


Preanalytical Requirements

Whereas the determination of L-lactate in serum requires tubes containing sodium fluoride, L-lactate in CSF does not. L-lactate in CSF samples is stable for 24 h at 4°C without the addition of sodium fluoride, even when leukocyte or erythrocyte counts are high, and for much longer when cell counts are low.


Methods for L-lactate. Amperometry is rarely used (Thomas, 2005). Enzymatic tests are usual, in particular the following two methods (Noll, 1984):

• Lactate is converted to pyruvate by lactate dehydrogenase (LDH, EC in the presence of NAD, and NADH2 is measured photometrically. The equilibrium of the reaction is shifted strongly toward lactate. For quantitative conversion of lactate, pyruvate must be removed from equilibrium by transamination with alanine aminotransferase (ALT, EC


• Lactate is converted by lactate oxidase to pyruvate and hydrogen peroxide (H2O2). Hydrogen peroxide reacts with peroxidase, p-aminoantipyrine, and phenol to form a dye that is measured photometrically.

Table 5.11 CSF lactate in patients with inflammatory disease or tumor (diagnostic lumbar puncture) (Felgenhauer and Beuche, 1999)


Mean value, mmol/L

Range, 5th–95th percentiles

Pneumococcal meningitis



Meningococcal meningitis



Listeria meningitis



Tuberculous meningitis






Lymphoma, leukemia






Guillain-Barré syndrome









Herpes simplex encephalitis



Viral meningitis



Herpes zoster ganglionitis



Multiple sclerosis



Table 5.12 Age-dependent reference ranges for L-lactate in CSF. Conversion factor: mg/dL × 0.11 = mmol/L

0–15 years

1.1–1.8 mmol/L (9.9–16.2 mg/dL)

16–50 years

1.5–2.1 mmol/L (13.5–18.9 mg/dL)

> 51 years

1.7–2.6 mmol/L (15.3–23.4 mg/dL)

The enzymatic tests can be used as monotests (coefficient of variation: 6%; detection limit: ≤ 0.2 mmol/L). Internal quality control should be performed with the reference values of commercially available control sera, since there is no reference test for L-lactate.

Methods for D-lactate. D-lactate in CSF is analyzed photometrically and fully enzymatically using D-lactate dehydrogenase (EC, or else in the same way as L-lactate.



Effect of serum levels. L-lactate in normal CSF is derived from cerebral tissue and from serum. Lactate concentrations in the blood may change under normal conditions without affecting the CSF concentration because the lactate transporter between blood and brain is saturated. Hence, it is not necessary to know the serum concentration when evaluating the lactate concentration in the CSF (Posner and Plum, 1967).

Reference range. Reference ranges for L-lactate in CSF are age-dependent (Table 5.12, Kleine et al., 1979; Kleine, 2003).

The normal upper limit for ventricular CSF is given as 2.2–2.4 mmol/L (Kalff and Kluge, 2003).

Clinical importance. For lumbar CSF, with a cut-off set at 3.5 mmol/L, the following apply:

• In acute (untreated) bacterial meningitis, including tuberculous meningitis, concentrations of 3.5 mmol/L or higher are found in over 90% of cases (Table 5.11) (Kleine et al., 1979; Luft and Götz, 1983).

• Even in bacterial meningitis that has been treated with antibiotics, 50–80% of CSF samples show elevated levels. Lactate concentration normalizes within 10 days after effective treatment with antibiotics.

• In tuberculous meningitis and fungal meningitis, the measurement of lactate concentration yields levels in the lower range of the group with bacterial meningitis. Despite effective treatment of tuberculous meningitis, normalization of lactate levels may take several weeks (Felgenhauer and Beuche, 1999).

• In viral forms of meningitis, lactate concentrations exceed the cut-off level in less than 0.1% of cases.

• In intracerebral hemorrhage (erythrocyte count ≥ 1000/μL), lactate levels may be above the cut-off level.

• Primary and secondary tumors, circulatory problems, and acute cerebral seizures (status epilepticus or acute intoxication of the CNS) are largely found in the range below 3.5 mmol/L (Table 5.11).

In bacterial meningitis, with a cut-off set at 4.5 mmol/L, the diagnostic sensitivity is 100% with a diagnostic specificity of 93% and a positive predictive value of 48%. In combination with the CSF cell count (> 800/μL), the specificity increases to 99% and the positive predictive value to 88%, but the diagnostic sensitivity decreases to 71% (Kleine et al., 1979).

Compared to glucose analysis, which requires determination of both CSF and serum concentrations (to calculate the CSF/serum ratio), lactate analysis is preferred because:

image Only one determination is required, i. e., in CSF.

image Pathological lactate levels are detectable earlier than pathological glucose levels.

image Pathological levels remain detectable for longer, even in patients undergoing treatment.

The increase and fluctuation in the L-lactate level of ventricular CSF are both essential markers of the extent and prognosis of either generalized or localized cerebral ischemia and hypoxia. Lactate values obtained from ventricular drainages are suitable for monitoring during the first 6–8 days in order to identify prognostic trends:

• Initially elevated base levels (> 2.5 mmol/L) after trauma may normalize within a few days to values lower than 2 mmol/L, a sign of good prognosis, whereas constant values or even further elevation indicate a poor prognosis.

• In cases of subarachnoid hemorrhage, patients with ventricular CSF lactate levels of 2.4 ± 0.8 mmol/L have a very good prognosis, whereas those with levels of 4.0 ± 2.1 mmol/L have a poor prognosis (Glasgow Outcome Scale: 3–5).


D-lactate levels of 0.2 mmol/L or less in both blood and CSF indicate bacterial infection even if the L-lactate concentration in the CSF is below 3.5 mmol/L.

D-lactate production dominates over L-lactate production in some bacterial strains in culture (e. g., Escherichia-coliNeisseria meningitidis). However, this has not been replicated under physiological in vivo conditions in animal experimental models of the corresponding meningitis: here, too, L-lactate production dominates (Wellmer et al., 2001; Prange, 2004). In humans, no connection between Dlactate production and severity of meningitis has been found:

• D-lactate levels were continuously low, or below the detection levels of the methods.

• The maximum value of D-lactate concentration (0.6 mmol/L) nevertheless constituted only 11% of total lactate in the CSF of this patient (Enterobacter sp. infection; Prange, 2004).

• The D-lactate fraction in E. coli meningitis was, on average, 1/80 of the L-lactate concentration.


General Remarks


The drop in glucose concentration in the CSF observed in bacterial and tuberculous meningitis was described by Lichtheim in 1893 and still constitutes the most commonly used method worldwide for differentiating between bacterial and viral meningitis.

Physiology and Pathophysiology

Kinetic studies have demonstrated the existence of saturable, stereospecific carrier mechanisms for the transport of glucose from blood into brain. The metabolic conditions in cerebral tissue are described above with reference to lactate. They also play a part in the drop in CSF glucose observed under pathological conditions. Impaired glucose transport through the blood–brain barrier is also under consideration as a pathological factor (Kornelisse et al., 1995).


Preanalytical Requirements

CSF is collected in tubes with or without sodium fluoride; at the same time, blood is taken in sodium fluoride tubes. Glucose in CSF samples with high leukocyte or erythrocyte counts is stable without sodium fluoride for at least 5 hours at room temperature, and for 24 hours at 4°C—and for much longer in CSF samples with low cell counts.

Table 5.13 Reference ranges of glucose concentrations in CSF and blood. The reference ranges for capillary, venous, and whole blood differ, and they vary with age (increase by 0.1 mmol/L/decade of life). Conversion factor: mmol/L × 0.0555 = mg/dL (Thomas, 2005)

CSF, mmol/L*

Serum, mmol/L*

CSF/serum ratio*

3.4 (2.7–4.2)

4.5 (3.3–5.5) 0.74


* Median with 5th and 95th percentiles.


Glucose is determined quantitatively by photometry using the hexokinase/glucose-6-phosphate dehydrogenase method (reference method) in the lysate of CSF and blood (cited by Vormbrock, 1984).


Effect of serum levels. Unlike lactate, the level of glucose in the blood has a direct effect on the level of glucose in the CSF. Hence, it is necessary to determine glucose in both CSF and blood.

Apart from very rare cases of impaired glucose metabolism (CSF analysis in children), lactate determination is superior to glucose determination (see above, “Lactate”).

Reference values. As a rule, the CSF concentration of glucose should be about 70% of the blood concentration (for reference ranges, see Table 5.13).

Clinical importance. CSF/serum quotients that are lower than 0.5 indicate bacterial, tuberculous, or fungal diseases of the CNS.


Felgenhauer K, Beuche W, eds. Labordiagnostik neurologischer Erkrankungen. Liquoranalytik und-zytologie, Diagnose und Prozessmarker. Stuttgart: Thieme; 1999

Garcia T, Killian JA, De Sanctis A. The lactic acid and the sugar content of the spinal fluid in meningitis. Arch Pathol 1928;6:530

Kalff R, Kluge H. Besonderheiten des Ventrikelliquors. In: Zettl UK, Lehmitz R, Mix E, eds. Klinische Liquordiagnostik. Berlin: Walter de Gruyter; 2005:366–377

Killian JA. Sugar and lactic acid of spinal fluid in meningitis. Proc Soc Exp Biol Med 1925;23:255–257

Kleine TO, Baehrlocher K, Niederer V, et al. Die diagnostische Bedeutung der Lactatbestimmung im Liquor bei Meningitis. Dtsch Med Wochenschr 1979;104:553–557

Kleine TO. Lactat. In: Zettl UK, Lehmitz R, Mix E, eds. Klinische Liquordiagnostik. Berlin: Walter de Gruyter; 2003:130–133

Kornelisse RF, de Groot R, Neijens HJ. Bacterial meningitis: mechanisms of disease and therapy. Eur J Psychiatry 1995:154:85–96

Lichtheim L. Gehirnkrankheiten durch die Punktion der Subarachnoidalräume. Dtsch Med Wochenschr 1893;19:1234–1236

Luft D, Götz R. Lactat im Liquor cerebrospinalis. Bedeutung für Differentialdignose, Therapiekontrolle und Prognose zerebraler und meningealer Erkrankungen. Lab Med 1983;7:55–59

Nishimura K. The lactate content of blood and spinal fluid. Proc Soc Exp Biol Med 1924;22:322–324

Noll F. L-lactate. In: Bergmeyer HU, ed. Methods of enzymatic analysis. 3rd ed, vol 6. Weinheim: Verlag Chemie; 1984:588–592

Panteghini M, Pagani F. Biological variation of lactate and pyruvate in blood. Clin Chem 1993;39:908

Posner JB, Plum F. Independence of blood and CSF lactate. Arch Neurol 1967;16:494–496

Prange J. D- und L-Laktat bei bakterieller Meningitis. Dissertation, University of Göttingen; 2004

Thomas L. Lactat. In: Thomas L, ed. Labor und Diagnose. 6th ed. Frankfurt: TH Books; 2005

Vormbrock R. UV-method with glucose dehydrogenase. In: Bergmeyer HU, ed. Methods of enzymatic analysis. Vol 4. Weinheim: Verlag Chemie; 1984:172–178

Walz W, Mukerji S. Lactate release from cultured astrocytes and neurons: a comparison. Glia 1988;1:366–370

Wellmer A, Prange J, Gerber J, et al. D- and L-lactate in rabbit and human bacterial meningitis. S. Scand J Infect Dis 2001;33:909–913