Comparison of Sodium and Potassium Intake
with Excretion
JOSEPH SCHACHTER, M.D.,
P H . D . , PATRICIA H.
HARPER, R.D.,
A R L E N E W. CAGGIULA, P H . D . , ROBERT H. M C D O N A L D , M.D.,
M A R Y E. R A D I N ,
M.A.,
AND W A R R E N F. DIVEN, P H . D .
Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017
SUMMARY Nine weU-motirated adults, knowledgeable about nutrition, kept food records, sared food portions equal to what had been eaten, and collected 24-hour urine samples for 3 consecutive days. Estimates of
sodium and potassium intake were calculated from food table analyses of written food records and from flame
photometric analyses of food portions. For each subject the mean of the estimates for each of the 3 days was
compared with the mean of urine analyses for sodium and potassium for each of the 3 days. For the group of
nine subjects, the average estimate of sodium intake from analyses of food records was 11% lower than the
average estimate of urinary sodium excretion; the average estimate of sodium intake from analysis of food portions was 2% higher than urinary sodium excretion. For individuals, there were large differences between estimates of intake and measurement of sodium excretion. For the group of nine subjects, the average estimate of
potassium intake from analysis of food records was less than 1% lower than the average estimate of potassium
urinary excretion; the average estimate of potassium intake from analysis of food portions was 13% higher
than potassium urinary excretion. For individuals, as with sodium, there were large differences between estimates of intake and measurement of potassium excretion. (Hypertension 2: 695-699, 1980)
KEY WORDS
* sodium
• potassium
• urinary excretion measurements
T
Two studies have reported comparison between sodium intake and sodium urinary excretion. In one study
of 11 adults, after discussions with the subjects on the
importance of lowering sodium intake, a single 24hour diet recall was found to be a reliable measure of
24 hour-urine excretion of sodium and potassium.2 In
this study, the diet recall analysis consistently underestimated sodium intake by 30%. In another study,
when intake and urinary excretion of sodium were
averaged over a period of 28 days, they were almost
identical in 10 young men with relatively low sodium
intake, and were highly correlated in 10 young men
with moderate sodium intake.3
Other studies have reported wide fluctuations in 24hour urinary sodium excretion in subjects on constant
sodium intakes.4-5 Several factors have been identified
that influence this variability: oscillation of balance
for water and electrolytes,8 level of sodium intake,5
level of potassium intake,1 degree of hydration,7 and
the delay in equilibration following a change in intake.6- 7 -'
In the present study of adults, estimates of sodium
intake from food table analyses of written food
records and from flame photometric analyses of food
portions are compared to 24-hour urine excretion of
sodium. Estimates of potassium intake and excretion
are also compared.
HE PURPOSE of this study was to determine
whether in free living adults measurement
of sodium and potassium from food ingested
yields a valid estimate of sodium and potassium excreted in the urine. This is of interest because 24-hour
urinary excretion, which is believed to provide the best
estimate of sodium intake, cannot be measured in free
living infants and is difficult to measure in free living
adults. In a prior study of 6-month-old infants,
sodium intake was assessed from food records, and a
positive association was found between sodium intake
and systolic blood pressure.1 When the infants were
reexamined at 15 months of age, the association
between sodium intake and systolic pressure was no
longer present. We wondered whether the more complex food intakes of 15-month-old infants rendered
the assessment of sodium intake from food records
invalid.
From the Department of Epidemiology, Graduate School of
Public Health; and the Departments of Pathology and
Biochemistry, University of Pittsburgh; the Department of
Psychiatry, University of Pittsburgh School of Medicine; and the
Presbyterian-University Hospital, Pittsburgh, Pennsylvania.
Supported by Research Grant HL 19864 from the National
Heart, Lung and Blood Institute.
Address for reprints: Joseph Schachter, M.D., Ph.D., 5400
Darlington Road, Pittsburgh, Pennsylvania 15217.
Received June 13, 1979; revision accepted December 12, 1979.
695
696
HYPERTENSION
Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017
Method
Adult subjects, rather than infants, were selected so
that 24-hour urine collections could be obtained. To
ensure a high degree of compliance with protocol, two
groups of adults were recruited: six nutrition students
at the University of Pittsburgh Graduate School of
Public Health who were studying assessment of food
intake, and seven members of our research team. Data
were collected during December and January.
Subjects received both oral and written instructions
for urine collection, and then kept a written record of
the time and amount of each voiding. To allow for a
lag time between eating and urine excretion,6'7i * a 24hour period was defined as being terminated by collection of the first morning specimen on the day following the day in which food portions were collected.
Therefore, each first morning specimen was kept
separate and analyzed separately from the rest of the
urine collected in that same calendar day. Because of
the variation in daily sodium excretion,4'6 urines were
collected for 3 consecutive days, with a first morning
specimen only being collected on the fourth day.
Subjects were instructed orally and in writing to
record times of ingestion and to describe each food
item and the amount ingested. In addition, they saved
portions equal in amount to all of the foods and
liquids ingested. A salt shaker containing approximately 15 g of salt was provided for each of the 3
days. The subjects were instructed not to use salt in
cooking or food preparation, but to add salt to their
food from the salt shaker provided. Each salt shaker
was weighed before and after each day's use to determine the amount of salt ingested. Written food
records were analyzed for sodium and potassium by
means of a computer file of nutrient values of foods
based on U. S. Department of Agriculture Handbook
456, plus lists of supplementary food items. Food portions were blenderized but not diluted or digested
prior to analysis; the solids were spun out, and the
supernatants analyzed for sodium and potassium.
Sodium and potassium concentrations of urine and
food supernatants were measured by flame
photometry using Instrumentation Laboratory Model
443 flame photometer with an internal lithium standard. Blood samples were collected from each of the
seven research team subjects for measurement of
plasma creatinine on one of the days of food and urine
collection, to permit calculation of creatinine
clearance.
Results
Mean 24-hour urine creatinine of each subject was
compared with published values for age- and sexspecific mean 24-hour creatinine in a New Zealand
population." The latter data were utilized because they
provided age- and sex-specific means and standard
deviations for creatinine excretion. Adults in New
Zealand did not differ in height from United States
adults, and differences in weight for age and sex
groups varied from not significant to a maximum of
VOL 2, No 5, SEPTEMBER-OCTOBER
1980
less than 4%.10 Since differences between the two populations in body mass and presumably in muscle mass
seemed small, it was assumed that differences in
creatinine excretion, similarly, would be small. Data
from three of the nutrition students were eliminated
because their mean creatinine excretions were more
than 1.2 standard deviations below the New Zealand
mean. These data were rejected because of the
probability that urine collection was incomplete, and
therefore urine sodium and potassium excretion rates
not valid. Data from one of the research team
members were eliminated because urine samples were
lost. For the remaining nine subjects (seven women,
two men) mean creatinine and mean urine volume
were compared to age- and sex-specific means in table
1. Daily electrolyte excretions for the nine subjects are
included.
Creatinine clearance calculated for further assessment of the adequacy of urine collection ranged from
83 to 105 ml/min in five subjects, and in one, 151
ml/min. The usual range of creatinine clearance is 60
to 110 ml/min. The creatinine excretion and
creatinine clearance data all indicated that adequate
24-hour urine samples were obtained in these subjects.
The average amount of sodium found in 24-hour
urine samples for the nine subjects ranged from 76 to
153 mEq, which was similar to the average of 123
mEq for 11 subjects trained to lower sodium intake.2
It was, however, lower than average levels observed in
the New Zealand population for age- and sex-specific
groups, 139 to 149 mEq for females and 177 to 184
mEq for men.' In another study of 142 males with
high normal or borderline elevation of blood pressure
who were undergoing a trial of primary prevention of
hypertension, the average of four 24-hour urine
measurements of sodium excretion ranged from 156 to
196 mEq.11
In the present study, for the group of nine subjects,
the estimate of sodium intake from food portions
agreed closely with the measurement of urinary excretion of sodium. The average estimate of sodium intake
from food portions for the nine subjects was 2 mEq
higher than the average measurement of urinary excretion (table 2). The average estimate of sodium intake from food records was 13 mEq lower than the
average urinary excretion.
For individuals, differences between the estimate of
sodium intake from food portion and urine measurement ranged from -36% to +44% (SD 36.4);
differences between the estimate of sodium intake
from food record and urine measurement ranged from
-51% to +61% (SD 39.6). The differences between
food record and urine measurement of sodium exceeded ± 25% in four of the nine subjects.
Differences between sodium intake as estimated by
food portion analysis and urine excretion did not seem
to be a function of the magnitude of fluctuations in
daily urine volume either as reflected by variability in
daily creatinine excretion or by variability in daily
urine collected. The absolute differences among each
of the three daily creatinine excretions and the mean
creatinine excretion for each subject were calculated
SODIUM INTAKE AND SODIUM EXCRETION/Schachter et al.
697
TABLE 1. Age- and Sex-Specific Mean SJf-How Creatinine Excretion and Urine Volume in Six New Zealand Subjects Compared to
Nine Subjects in the Present Study
n
138
138
138
86
138
138
138
147
82
New Zealand Study
Urine volume
Creatinine
(mg)
(ml)
mean ± SD
mean ± SD
1097 ± 455
1346 ± 370
1097 ± 455
1346 ± 370
1346 ± 370
1097 ± 455
1437 ± 547
1753 ± 543
1097 ± 455
1346 ± 370
1097 ± 455
1346 =* 370
1097 ± 455
1346 •* 370
1397 ± 542
1967 ± 711
1245 ± 542
1323 * 360
Creatinine
2
(mg)
mean
1064
1013
3
940
4
6
1555
1077
1343
7
844
8
1428
9
972
n
1
5
Present Study
Sodium excretion
Urine volume
(mEq)
Day
Day
Day
(ml)
3
2
1
mean
1454
165
149
143
1272
86
55
117
62
121
1315
160
1405
.168
115
—
1325
54
77
97
1162
111
138
155
114
122
1116
101
82
1147
180
163
112
1513
67
87
<
Potassium excretion
(mEq)
Day
1
Day
2
Day
3
61
48
48
56
42
58
60
51
62
79
61
—
64
60
61
40
54
60
64
53
63
73
56
42
63
81
54
Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017
(SD 6.6); differences between food record and urine
measurement ranged from -22% to +26% (SD 10.7).
and then averaged. The squares of these absolute
differences for each subject were calculated also and
then averaged. The four subjects with the largest
differences in sodium between food portion analysis
and urine excretion were compared with the four subjects with the smallest differences. The absolute
differences and the squares of the differences in daily
creatinine excretion were similar for the two groups of
four subjects. The same analyses with daily urine
volumes indicated that the absolute differences and the
squares of the differences in daily urine volume were
similar for the two groups of four subjects.
For the group of nine subjects, the average estimate
of potassium intake from food portions averaged 7
mEq higher than the average measurement of urinary
potassium. The average estimate of potassium intake
from food records averaged less than 1 mEq lower
than measurement from urine samples (table 3).
For individuals, differences between food portion
and urine measurement ranged from —7% to +31%
Discussion
In a group of 13 well-motivated adult subjects, data
from four had to be rejected because urine collection
appeared inadequate. There were large differences in
the estimates of sodium intake for each individual
based on the mean of the values for each of 3 days for
either food records or food portions compared to
measurement of sodium excretion in three 24-hour
urine samples. Similar differences between intake and
urinary excretion of sodium were reported in earlier
studies of subjects on constant sodium intakes.2''
From these studies data were selected for the last 3
days of measurement for each subject (table 4);
average differences for individuals between sodium intake and excretion ranged from -18% to +52% (SD
23.4). The large differences found in these prior
studies support the conclusion of the present study
TABLE 2. Average Sodium Intake and Excretion Per Day (mEq)
Food
record
Food
portion
Food
portion—urine
Difference
%
Food
record—urine
Difference
Urine
1
94
126
153
-27
-18
-58
-38
2
83
80
86
-6
-7
-3
-4
3
116
158
114
44
38
2
2
4
92
91
141
-51
-36
-50
-35
5
77
104
76
27
36
1
1
-68
-51
Ss
%
6
66
92
135
-42
-31
7
133
163
113
50
44
20
18
8
127
160
142
18
13
-15
-11
9
143
96
89
7
8
54
61
103.4
118.9
116.6
2.2
5.2
-13.0
-6.3
Average
Standard deviation
36.4
39.6
698
HYPERTENSION
VOL 2, No
5, SEPTEMBER-OCTOBER
1980
TABLE 3. Average Potassium Intake and Excretion Per Day (mEq)
Food
record
Food
portion
Food
portion—urine
Difference
%
Food
record—urine
Difference
Urine
1
51
67
52
15
29
-2
-4
2
62
68
52
16
31
3
53
61
58
3
5
-5
-8
4
55
76
70
6
9
-15
-22
5
48
57
62
-5
-7
-13
-22
6
45
59
51
9
17
-6
-12
7
67
68
60
8
14
6
11
8
62
60
57
3
6
5
8
9
83
78
66
12
18
17
26
58.4
66.2
57.8
-0.3
-0.4
Ss
Average
Standard deviation
10.7
6.6
Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017
that, based upon 3-day collection in well-motivated
adults who were knowledgeable regarding nutrition,
neither food record estimates nor analyses of food
portions provided valid measures for individuals of 3day urine excretions of sodium.
When, in a previous study, sodium intake for individuals was averaged over a 28-day period, there
was close agreement with urinary sodium averaged
over this same period.3 It is likely, therefore, that for
some number of days greater than 3 and less than 28,
measurement of sodium intake from analysis of food
portions would provide valid estimates of urinary excretion of sodium in individuals.
The present study was Undertaken to assess the
validity of food intake measures of electrolytes as indices of urinary excretion of electrolytes. In addition
to the issue of validity, there is also the question of
feasibility and difficulty in measuring either food in-
20
10
13.6
7.4
%
take or urinary excretion. Among our well-motivated
subjects, there was the subjective impression that the
collection and analysis of food portions was approximately as difficult a task, and involved as much
inconvenience, as obtaining 24-hour urine samples.
Food record estimates of sodium intake were 11%
lower than urinary excretion in the present study
(table 2) and 30% lower than urinary excretion in a
prior study.2 Perhaps the subjects, aware of the
desirability of minimizing sodium intake, had underestimated the quantities of sodium-containing foods in
their food records. Food records, in the present study,
did not underestimate the potassium-containing foods
ingested. An alternate hypothesis is that the food table
values underestimate the sodium content of numerous
foods. There seems no plausible reason for this.
Further, the food table values do provide valid
measures of the potassium content of foods.
TABLE 4. Average Sodium Excretion Per Day (mEq) of Subjects on Constant Sodium Intake in Other Studies
24-Hour urine sodium excretion
Day 3
Day 2
Average
Ref
no.
Ss
Sodium
intake
Day 1
3
FW
149
243
122
112
3
BM
156
120
99
130
3
SV
99
69
86
3
JP
108
121
Difference
Intake—Urine
%
159
-10
-6
116
40
34
91
82
17
21
99
87
102
6
6
3
EH
109
148
96
132
125
-16
-13
3
FS
129
142
125
135
134
-5
-4
3
KR
DM
NC
FMc
DB
AT
140
130
120
143
131
9
7
141
183
193
141
172
-31
-18
146
140
160
127
142
4
3
123
147
121
124
131
-8
-6
150
182
112
96
130
20
15
150
129
97
68
98
52
53
3
3
3
2
2
Average
Standard deviation
133.3
126.8
6.5
23.4
7.7
SODIUM INTAKE AND SODIUM EXCRETION/Sc/iac/i/er et al.
Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017
Results from the present study of nine subjects
suggest that, despite the large differences for individuals between intake and excretion of sodium, the
average of the measurements of sodium in food (food
portion analysis or ingestion of prescribed amounts of
sodium) will not deviate from average urinary excretion by more than 2% to 5% (tables 2 and 4). For
epidemiological study, therefore, measurement of
sodium intake by analysis of food portions over a 3day period should provide a fairly accurate measure of
average sodium excretion for the group. It should be
possible, for example, to obtain reasonable estimates
of urinary excretion of sodium from measurement of
sodium in food ingested in groups with different blood
pressure levels, including groups containing only nine
or more subjects. Measurement of sodium intake by
analysis of food records is not expected to provide the
degree of accuracy as analysis of food portions.
Several of the differences for individuals between
3-day estimates of potassium intake and urinary excretion exceeded 25%, and a considerable number
exceeded 10%. Consequently, measures of potassium
intake are not accurate estimates of urinary excretion
over a 3-day period. Differences for individuals
between potassium intake and urinary excretion
showed a smaller variance than for sodium (tables 2
and 3). For this group of nine subjects, the average estimate of potassium intake deviated from the average
urinary excretion by less than 8%. For groups of this
size or larger, therefore, 3-day estimates of intake may
provide reasonable estimates of urinary excretion of
potassium.
699
References
1. Schachter J, Kuller LH, Perkins JM, Radin ME: Infant blood
pressure and heart rate: Relation to ethnic group (black or
white), nutrition, and electrolyte intake. Am J Epidemiol 110:
205, 1979
2. Gillum RF, Elmer P, Thorpe W, Kehede J: Is the dietary recall
worth its salt? Hypertension. In press
3. Mickelson O, Makdani D, Gill JL, Frank RL:, Sodium and
potassium intakes and excretions of normal men consuming
sodium chloride or a 1:1 mixture of sodium and potassium
chlorides. Am J Clin Nutr 30: 2033, 1977
4. Black DAK, Thomson AE: Day-to-day changes in sodium and
water output with and without posterior pituitary extract. Clin
Sci 10: 511, 1951
5. Baldwin D, Alexander RW, Warner EG: Chronic sodium
chloride challenge studies in man. J Lab Clin Sci 55: 362, 1960
6. Gamble JL: Lane Medical Lectures: Companionship of Water
and Electrolytes in the Organization of Body Fluids. Stanford,
Stanford University Press, 1949, pp 33-38
7. Ladd M: Effect of prehydration upon renal excretion of sodium
in man. J Appl Physiol 3: 603, 1951
8. Liu K, Cooper R, McKeever J, McKeever P, Byington R,
Soltero I, Stamler R, Gosch F, Stevens E, Stamler J: Assessment of the association between habitual salt intake and high
blood pressure: Methodological problems. Am J Epidemiol
110: 219, 1979
9. Simpson FO, Nye ER, Bolli P, Waal-Manning HJ, Goulding
AW, Phelan EL, de Hamel FA, Spears GFS, Leek M, Stewart
AC: The Milton survey: Part 1, General methods, height,
weight, and 24-hour excretion of sodium, potassium, calcium,
magnesium and creatinine. NZ Med J 87: 379, 1978
10. National Center for Health Statistics: Weight and Height of
Adults 18-74 Years of Age, Series 11, No. 211. Hyattsville,
Maryland, Department of Health, Education, and Welfare,
1979
11. Leaf A, Couter WT: Evidence that renal sodium excretion by
normal human subjects is regulated by adrenal cortical activity.
J Clin Invest 28: 1067, 1949
Comparison of sodium and potassium intake with excretion.
J Schachter, P H Harper, M E Radin, A W Caggiula, R H McDonald and W F Diven
Hypertension. 1980;2:695-699
doi: 10.1161/01.HYP.2.5.695
Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017
Hypertension is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1980 American Heart Association, Inc. All rights reserved.
Print ISSN: 0194-911X. Online ISSN: 1524-4563
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://hyper.ahajournals.org/content/2/5/695
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Hypertension can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial
Office. Once the online version of the published article for which permission is being requested is located, click
Request Permissions in the middle column of the Web page under Services. Further information about this
process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Hypertension is online at:
http://hyper.ahajournals.org//subscriptions/