International Archives of
Int Arch Occup Environ Health (1990) 62: 227-232
Enzymatic assay of formic acid and gas chromatography of methanol
for urinary biological monitoring of exposure to methanol
Masana Ogata and Tomoko Iwamoto
Department of Public Health, Okayama University Medical School, 2-5-1 Shikata-cho, Okayama Japan
Received June 30 / Accepted December 20, 1989
Summary An enzymatic assay method for the deter-
mination of urinary formic acid is described Formic acid
in urine was cleaved to carbon dioxide and water by for-
mic acid dehydrogenase, whereby NAD + was converted
to NADH, which reacted with INT (p-iodonitrotetra-
zolium violet) in the presence of NAD-diaphorase The
color thus produced was determined at 500 nm In addi-
tion, a simple gas chromatographic method of urinary
formic acid is described, in which head space gas of for-
mic acid methylester was applied into the wide bore col-
umn The urinary formic acid concentrations by the en-
zymatic method agreed well with that by the gas chro-
matographic method A simple gas chromatographic
method for urinary methanol assay is also described.
Acetonitrile was added to an equal volume of urine con-
taining methanol After centrifugation, the supernatant
was injected into gas chromatography (GC) The peaks
of urinary methanol and ethanol were separated by GC.
Formic acid and methanol in urine of unexposed healthy
subjects and workers exposed to methanol were ana-
lyzed by the colorimetric and gas chromatographic meth-
ods Geometric mean concentrations of urinary formic
acid and methanol in the healthy subjects were 7 82 mg/g
creatinine and 1 34 mg/l, respectively The concentration
ratio of formic acid to methanol in the urine of the work-
ers exposed to methanol was calculated to be 3 67 ± 2 10,
which agreed with the ratio under a controlled exposure
experiment A slower excretion of formic acid than that
of methanol in the urine of a volunteer was also ob-
served.
Key words: Formic acid Methanol Biological moni-
toring Enzymatic analysis Gas chromatography
Introduction
Methanol has been widely used as an ingredient of thin-
ner and for the extraction of biological materials When
methanol is inhaled, it is converted into formaldehyde
by alcohol dehydrogenase and into formic acid by alde-
Offprint requests to: M Ogata
OCCUpationala
Environmental
Health
© Springer-Verlag 1990
hyde dehydrogenase, and formic acid is excreted into
urine l9 l as a major metabolite of methanol A part of
formic acid is converted into carbon dioxide by formic
acid dehydrogenase and then exhaled as carbon dioxide.
In the urine of workers exposed to methanol, the con-
centration of formic acid is higher than that of methanol
l5 l A gas chromatographic method has been used for
the determination of urinary methanol and formic acid
l8 l In the gas chromatographic assay, formic acid should
be converted into N,N-dimethylformamid by 2-step
reactions and the latter is measured by gas chromatog-
raphy (GC) with TCD l2l or with thermoionic N-detec-
tor l7 l The procedure of these methods is complicated
and time-consuming On the other hand, Grady et al l 4 l
described a method for the determination of blood for-
mic acid using formic acid dehydrogenase, and Ohmori
et al l10l developed a method involving the benzimidazol
derivation from formic acid for high performance liquid
chromatography (HPLC), and the latter group com-
pared the results obtained by the enzymatic method with
those obtained by their HPLC method on biological
specimens.
The present report describes simple method of en-
zymatic assay of formic acid and gas chromatographic
analysis of methanol on urine samples of workers ex-
posed to methanol.
Materials and methods
Urine specimens Urine specimens were collected from eight male
workers, 41 years of age, exposed to about 120 ppm of methanol in
workshops at the end of shift and from unexposed male workers,
36 years of age in their thirties under similar conditions The ex-
posed and unexposed workers ate breakfast and lunch prepared
from the same menus A volunteer (male, 63-years-old) was also
placed in an exposure chamber and exposed to 200 ppm methanol
for 4 h Urine was collected during and after exposure and was
analyzed for formic acid and methanol.
Reagents All the reagents used were analytical grade from Wako
Pure Chemicals Ltd (Osaka) A boron trifluoride-methanol rea-
gent was purchased from Wako Pure Chemicals Ltd Formate
dehydrogenase (FDH; 0 50 units/mg solid, NAD-diaphorase con-
taining 6 g/l of NAD and 500 units/l of diaphorase) and p-iodo-
nitrotetrazolium violet(INT) were purchased from Sigma Co (IL,
228

USA) Phosphate buffer (100 mmol/l, p H 6 0) was prepared by

Colibration curve
mixing 100 mmol/l KH 2PO 4 and 100 mmol/l Na 2HPO 4 (5:1).
NAD-diaphorase was dissolved in the phosphate buffer to give a
o-o Water
concentration of 1 8mg/ml Buffered NAD-diaphorase-INT was
t Urine
prepared by adding 40 ml of INT to 20 ml of buffered NAD-
diaphorase.

Enzymatic assay One hundred Iglof urine samples were mixed by

a Vortex mixer (5 s) with 100 l acetonitrile in 10-ml test tubes and
I,
centrifuged (1600 g) for 10 min The supernatants (1001 d) were -2
added to 10-ml test tubes containing 2 5 ml buffered NAD- II
diaphorase-INT After mixing, 60 gl of FDH (5 units/ml kept at
5 °C) were added, mixed and left at room temperature (20 °C) for
10 min The reaction mixtures were transferred individually into a
glass cuvette of double beam spectrophotometer (Hitachi Model,
100-60) The zero adjustment of the instrument was made against
water and absorbancies of the developed red colour were read at
500 nm The principle of the determination of formic acid is shown
in the following reaction scheme.

FDH
HCOOH ,CO 2 + H2 0 0 100
I 200
Formic Acid Concentration (mg/i)
NAD + H NADH
Fig 1 Calibration curves for formic acid added to water and urine
diaphorase
reduced INT p-iodonitrotetrazolium violet (INT)

Gas chromatographicassay with head space gas A head space gas
technique by GC was used In this method, a head space gas of for-
mic acid methylester was applied into GC and was allowed to omit
the procedure producing dimethylformamide A rubber cap cov-
ered by a Teflon sheet was lined with an aluminum cap for sealing.
After injecting 1 ml urine, followed by 1 ml boron trifluoride-
methanol reagent into the vial, which was incubated at 200 C for
12 h and then warmed at 37 C for 40 min in a water bath The head
space gas was taken by an air-tight injector from the vial and
injected immediately into the gas chromatograph Conditions of
GC were as follows: column, a 25 m x 0 53 mm i d ; and CARBO-
WAX 20 M, df 3 ipm "fused silica" capillary column from Quadrex
Corporation The column oven temperature was 40 °C for the ini-
tial 2 min, then it was increased from 40 °C to 200 °C at a rate of
20 °C/min A final temperature of 200 °C was maintained for 3 min.
Carried gas flow rates were: He, 20 ml/min and feul gas pressure
were: H 2, 0 5 kg/cm 2 ; air, 0 5 kg/cm 2 .
Determination of urinary methanol Five hundred gl of urine sam-
ples were vortexed (5 s) with 500 11of acetonitrile in 10-ml test
tubes and centrifuged (1600 g) for 10 min One l of supernatant
was injected into a Shimadzu GC-14 A GC connected with a flame
ionization detector (FID) GC was carried out using a 25 m x 0 53
mm i d CARBOWAX 20 M "fused silica" capillary column from
Quadrex Corporation (New Haven, CT, USA) The column oven
temperature was 60°C for the initial 2 min, then it was increased
from 60° to 80 0C at 5°C/min and subsequently from 80° to 1800 C at
20 °C/min A final temperature of 180°C was maintained for 5 min.
Other temperature settings were 250 °C for injector and 280 °C for
FID Gas flow rates were: He, 10 ml/min; H2 , 30 ml/min; and air,
400 ml/min The split flow was set at 40 ml/min and suspended for
the first 0 5 min of the run for splitless injection (injection volume,
1 1)
p Data was processed with a Hitachi D-2500 integrator.
were each added to an equal volume of water or urine.
The changes in optical density after enzymatic reaction
were determinated at 500 nm by spectrophotometry A
linear relationship was obtained between the formic acid
concentration and the optical density (Fig 1) The cali-
bration line of formic acid in urine paralleled that in
water The shift of the line for formic acid in urine from
that in water was due to the endogenous formic acid pre-
sent in urine.
Detection limit
The detection limit for formic acid was 3 mg/l when three
times the maximum absorbance of background noise was
set as a detection limit for the spectrophotometric assay
for formic acid according to the suggestions made by Ca-
rins and Rogers l3l
Recovery rates
Samples with formic acid present at concentrations of
10, 20, 40, 50, 80, 100, 200 and 400 mg/l were each added
to an equal volume of normal urine from unexposed sub-
jects and the recovery rates (mean ± SD) were deter-
mined by the enzymatic assay to be 104 4 + 7 68, 96 9
± 9 40, 97 3 + 10 3, 100 0 + 3 27, 93 9 + 5 45, 89 5
+ 6 57, 97 3 ± 1 35 and 96 2 + 1 68, respectively.

Results Relative standarddeviations

Calibrationof formic acid Relative standard deviations (RSD) of the determina-

tion of formic acid in the same urine specimens contain-
Standard solutions of formic acid, ranging in concentra- ing 50, 100 and 300 mg/l (n = 5 for each) were 4 22, 3 19
tion from 10 to 400 mg/1, were prepared The solutions and 4 64 %, respectively.

Effect of methanol on enzymatic assay of formic acid
Methanol was added to formic acid to give methanol
concentrations of 10, 20, 40 and 80 mg/i and a formic
acid concentration of 20 mg/l The mean measured ratios
were 0 95 + 0 046, 0 97 + O 012, 0 94 + 0 00 and 0 98
+ 0 035, respectively These results indicate that the
methanol contained in urine of workers exposed to
methanol does not interfere with the enzymatic assay of
formic acid.
Formic acid contents in urine from unexposed workers
Urine samples from 31 male control workers from 30 to
39 years of age (average; 36-years-old) and without oc-
cupational exposure to methanol and formic acid were
studied by the enzymatic assay After correction for uri-
nary creatinine, the formic acid content of unexposed
urine revealed a log-normal distribution The mean
value of the logarithm of the formic acid concentration
(m) corrected for creatinine was 0 893 and the standard
deviation (SD) was 0 289 The geometric mean value
was 7 82 mg/g creatinine The upper 95 % fiducial limit
of urinary formic acid was 23 4 mg/g creatinine (n = 31).
Head space GC of formic acid in urine
Head space gas of formic acid methylester was applied
into GC and was allowed to omit the procedure produc-
ing dimethylformamide with dimethylamine followed by
extraction with chloroform Recoveries of formic acid
added at a concentration of 68 mg/i to urine by this meth-
od was 103 0 ± 4 90 % The relative SD in this assay of
formic acid, ranging 34 to 272 mg/i in serial determina-
tion, was calculated to be 7 61 % (n = 4).
Correlationbetween formic acid concentrationsobtained
by the gas chromatographicand enzymatic methods
Formic acids of various concentrations were added to
normal human urine and the amounts of formic acid in
urine were determined by the gas chromatographic and
enzymatic method The results obtained were compared
(Fig 2) The regression equation with formic acid could
be expressed as y = 1 OOx 0 086, where y = the formic
acid measured by the gas chromatographic method
(Fig 2) The values obtained by both methods gave a
good agreement.
Gas chromatogram of methanol
Authentic samples of methanol showed sharp peaks and
eluated within 3 min of injection (Fig 3A) Chromato-
grams of methanol from urine of subjects not exposed to
methanol and those of workers exposed to methanol are
shown in Fig 3B and 3C respectively Methanol was
well separated from urinary compounds of normal sub-
jects, indicating that the method is suitable for the sep-
aration and quantitative analysis of methanol.
229
o
0
Formic acid by gas chromatographic method
(m mol / I )
Fig 2 Relationship between concentration of formic acid in nor-
mal human urine measured by the gas chromatographic method
and by the enzymatic method
Calibrationof methanol
Standard solutions of methanol, ranging in concentra-
tions from 5 to 40 mg/1, were prepared and added to an
equal volume of water or urine, and the mixed solutions
were directly injected into GC A linear relationship was
obtained between the methanol concentration in water
or urine and the peak height (Fig 4) Calibration lines
are expressed by the following equations:
y = 0 233 x + 0 13 for water,
y = 0 234 x + 1 34 for urine,
where x is the concentration of methanol (mg/l) and y is
the peak height (mv).
The calibration curve of methanol in urine paralleled
that in water, and the displacement of line for urine from
that for water was due to the methanol normally present
in the urine of a person non-occupationally exposed to
methanol.
Detection limit
The detection limit for methanol was 3 mg/l, when three
times the maximum peak height of background noise
was set as a detection limit for the gas chromatographic
analysis of methanol by Cairns and Rogers l3 l.
Recovery rates
Samples with methanol present at concentrations of 5,
10, 20 and 40 mg/1 were each added to an equal volume
of urine from subjects unexposed to methanol, and re-
covery rates were determined by GC to be 90 0, 103 0,
101 5 and 100 2%, respectively.
230

'I
A B C 1
4,

oh

I
Fig 3 Gas chromatograms of methanol.
t K A: authentic sample (14 Omg/1) of
_<JLJ
_~ methanol; B: normal urine specimen;
C: urine specimens of a worker exposed
to methanol
0 5 O 5 O 5
time (min) time (min) time (min)

Calibration Curve

o Waler

* Urine

C-

WC'

40
Me OH concentration by GC injected urine
methanol concentration (mg/i)
Fig 5 Linear relationship between methanol concentrations deter-
Fig 4 Calibration curves for methanol added to water and urine
mined by two different methods, one using urine directly and one
with acetonitrile extract of urine Capillary column-GC was used

Relative standard deviation
RSD of determined values on 10 4 mg/l of methanol in
the same urine specimen of workers exposed to methanol
was 5 16 % (n = 6).
Methanol contents of urine from unexposed workers
Urine samples from 30 male controls between 22 and 46
years of age and without occupational exposure to
methanol were studied by GC with acetonitrile extract.
The geometric mean (m) of urinary methanol measured
by GC on acetonitrile extract was 1 34 mg/ or 1 12 mg/g
creatinine, the geometric SD was 1 66 or 2 14, and the
upper 95% fiducial limit of methanol was 3 86 mg/i or
4.31 mg/g creatinine (n = 30).
Comparison of urinary methanol concentrations
of workers measured by the direct method
with those by the acetonitrileextraction method
with a capillary column
Methanol concentrations in the urine of workers ex-
posed to methanol were determined by GC on urine and
with capillary column on acetonitrile extract of methanol
from urine A linear relationship was obtained between
the methanol concentrations determined directly on
 231

Table 1A, B Concentrations of formic acid and methanol in the urine of workers (A) and a volunteer (B) exposed to methanol
A
Case no Formic acid Methanol Concentration ratio Molar
on the w/v base concentration ratio
mg/l mmol/l mg/ mmo/1
1 185 7 4 04 24 9 0 78 7 46 5 18
2 480 1 10 4 106 5 3 33 4 51 3 21
3 23 5 0 51 68 0 21 3 46 2 43
4 18 2 0 40 10 4 0 33 1 75 1 21
5 21 O 0 46 66 0 21 3 18 2 19
6 84 5 1 84 83 8 2 62 1 01 0 70
7 49 8 1 08 11 6 0 36 4 29 3 00
x + SD 123 3 + 168 1 2 68 + 3 64 35 8 + 41 5 1 12 + 1 30 3 67 + 2 10 2 56 + 1 47

B
Timing (h) Formic acid Methanol Concentration ratio
mg/i mg/l: mg/g creatinine mg/l mg/l: mg/g creatinine on the w/v base
1.020a 1 020a
I (3-4) 23 2 17 2 14 3 6 50 4 81 4 01 36
II (0-4) 24 0 + 1 80 18 3 + 1 84 16 3 + 2 08 5 27 + 0 76 4 02 + 0 55 3 57 + 0 43 46
III (4-9) 18 1 + 2 20 15 9 + 1 78 14 4 + 2 86 1 86 + 0 50 1 57 + O40 1 42 + 0 46 10 1
a Concentration was corrected for the specific gravity of 1 020
I: last hour of exposuse
II: last 4 hours of exposure
III: 5 h after end of exposure

urine and those determined on acetonitrile extract
(Fig 5) The regression equation was: y = 3 76 + 1 05 x
(correlation coefficient, 0 99), where y represents meth-
anol concentration determined directly on urine and x
represents methanol concentration determined on ace-
tonitrile extract The concentration of the methanol ex-
tracted with acetonitrile shifted to a higher value than
that determined directly on urine The data indicated
that methanol bound to urinary protein was released by
the acetonitrile treatment.
Concentrationsof methanol and formic acid in the urine
from workers and a volunteer exposed to methanol
Results of the determination of urinary methanol and
formic acid in exposed workers and a volunteer are
shown in Table 1 The mean weight/volume concentra-
tion ratio and molar concentration ratio of formic acid to
methanol in urine of workers were 3 67 2 10 and
2.56 ± 1 47, respectively, indicating that the concentra-
tion of formic acid excreted in urine was higher than that
of methanol (Table 1A).
The concentration of methanol in the volunteer's
urine of the last 1h of exposure to 200 ppm methanol
was 6 50 mg/l (Table 1B) The data agree with that of
Sedivec and Mraz l 11l; namely 5 mg methanol/l urine for
143 ppm methanol exposure and 6 7 mg/l for 210 ppm
methanol in controlled exposure l11l The concentration
ratio of formic acid to methanol in the urine of the last
1 h of exposure was 3 6, which agreed with the ratio
found in workers' urine in this paper The concentration
ratio of the urine from the end of exposure to 5 h after
exposure was 10 1, which was higher than the value of
4.6 for the last 4 h of exposure, indicating that formic
acid excretion is slower than the methanol excretion.
Discussion
Determination of the concentration of urinary methanol
and formic acid has been used for biological monitoring
of exposure to methanol Recently, the German Re-
search Society documented that the BAT (biological
tolerance value for working materials) for urinary
methanol of workers exposed to methanol should be
30 mg/l Determination of the levels of urinary methanol
and formic acid for biological exposure indices for
methanol exposure are in progress by ACGIH.
Gas chromatographic and enzymatic assay methods
have been reported for the determination of formic acid
in biological specimens In the gas chromatographic
method, boron trifluoride-methanol used as a reagent
for the methylation of formic acid as the first step
strongly irritates the eyes In the second step, methyl
formate was extracted with chloroform, reacted with di-
methylformamide and injected into the gas chromato-
graphic column Moreover, it is disadvantageous, be-
cause the separation of methylformic acid and methanol
by the gas chromatographic analysis is not complete l7 l.
Head space GC of formic acid with wide bore capillary
column was also useful for the simple determination of
urinary formic acid as the second step can be omitted in
this method The concentration of urinary formic acid by
GC agreed with that by enzymatic assay method The
232
enzymatic method is simple for formic acid, and the
method has a higher analytical sensitivity and requires a
relatively short assay period of time as compared with
GC The results of recovery experiment of enzymatic
methods show sufficient accuracy for this method The
enzymatic method allows us to analyze a great number
of urine specimens to monitor workers exposed to
methanol.
The mean concentration of formic acid determined
by our method on urine of workers unexposed to metha-
nol or formic acid was 7 82 mg/g creatinine, which was
similar to or slightly higher than that determined by GC.
The loss of urinary methanol by evaporation during stor-
age and/or transportation should be considered in bio-
logical monitoring using urine of workers exposed to
methanol In contrast to methanol, non-volatile formic
acid in urine is stable during the storage of urine and a
positive correlation has been found between the concen-
tration of methanol in the air and the concentration of
formic acid in urine l8 l The determination of urinary
formic acid is advantageous, because the concentration
of urinary formic acid was about five times higher than
that of methanol.
In conclusion, the enzymatic assay method of urinary
formic acid, head space GC of formic acid methylester
and the direct capillary gas chromatographic method of
urinary methanol assay developed in this paper are re-
commended to determine the methanol concentration in
the urine of workers exposed to methanol.
The concentrations of methanol and formic acid in
the urine of workers and a volunteer were determined.
The ratio of methanol concentration in the urine of ex-
posed workers to that in the air to which the workers
were exposed was higher than that of a volunteer in a
controlled exposure experiment This discrepancy is due
to the fact that the pulmonary ventilation of workers is
larger than that of a volunteer at rest In the present
experiment, the numbers of workers were limited A
number of workers' urine were needed to estimate the
relationship between the concentration of methanol in
air and the concentration of urinary formic acid, mea-
sured by enzymatic method The average concentration
of urinary methanol in unexposed subjects was reported
to be 0 73 mg/l l 11l, 1 1mg/l l5 l and 2 05 mg/l l6l The
normal level of methanol in the present study was 1 40
mg/l, a value which agreed with their reported values.
Urinary concentration of formic acid in unexposed sub-
jects was on average 15 1mg/g creatinine l 8l, 12 7 mg/l
l5 l, and 11 9mg/l l1 l, which agree with the value of
7.82 mg/g creatinine obtained in the present experiment.
Acknowledgments We express our thanks to Prof Shinji Ohmori,
Dr Toyohiro Taguchi and Dr Toshio Kawai for their guidance in
this work, Dr Tokushi Horike for technical assistance, and Prof.
Kazuhisa Taketa for reviewing the manuscripts.
References
1 Baumann K, Angerer J (1979) Occupational chronic exposure
to organic solvents Int Arch Occup Environ Health 42:241-
249
2 Bricknell KS, Finegold SM (1978) Improved method for assay
of formic acid by gas-liquid chromatography J Chromatogr
151: 374-378
3 Cairn T, Rogers WM (1983) Acceptable analytical data for
trace analysis Anal Chem 55: 54A-57 A
4 Grady S, Osterloh J (1986) Improved enzymic assay for serum
formate with colorimetric endpoint J Anal Toxicol 10:1-5
5 Heinrich R, Angerer J (1982) Occupational chronic exposure
to organic solvents (X, Biological monitoring parameters for
methanol exposure) Int Arch Occup Environ Health 50:341-
349
6 Kawai T, Hirashima J, Horiguchi S (1986) Gaschromato-
graphic determination of urinary methanol by direct injection
method and the normal level of Japanese subjects Seikatsu
Eisei 30:260-262
7 Kimmerle G, Eben A (1986) Metabolism studies of N,N-di-
methylformamice Int Arch Arbeitsmed 34:127-136
8 Liesivuori J, Savolainen H (1987) Urinary formic acid as an
indicator of occupational exposure to formic acid and
methanol Am Ind Assoc J 48 (1) :32-34
9 Liesivuori J (1988) Biological monitoring and toxicological
aspects of formic acid and methanol In: Liesivuori J (ed)
Kuopion Yliopiston Julkaisuja Publications of the University
of Kuopio Kuopio, Helsinki Finland, pp 1-53
10 Ohmori S, Sumii I, Toyonaga Y, Nakata K, Kawase M (1988)
High-performance liquid chromatographic determination of
formate as benzimidazol in biological samples J Chromatogr
426:15-24
11 Sedivec V, Mraz M, Flek J (1981) Biological monitoring of
persons exposed to methanol vapours Int Arch Occup Envi-
ron Health 48:257-271