Europ.J.clin.Invest. 6, 75-83 (1976)

Sodium Balance and Renal Tubular Sensitivity to Aldosterone During Total Fast and Carbohydrate Refeeding in the Obese l) J. Kolanowski, P. Desmecht2), and J. Crabbg Department of Physiology and Internal Medicine, University of Louvain, Belgium Received: February 7 , 1 9 7 5 , and in revised form: August 8, 1975

Abstract. In man the first days of fasting are characterized by enhanced natriuresis despite an increase in aldosterone secretion. Therefore the possibility of a decreased renal tubular sensitivity to this hormone was considered. The response to aldosterone infused before a fast and again on day 4 of fasting was evaluated in 12 starving obese women in terms of urinary sodium, chloride and potassium excretion. The data were compared to those obtained from 20 untreated starving obese women. In addition, salivary flow and sodium output were measured before and after aldosterone infusion in 4 out of the 12 patients treated. The involvement of aldosterone in the mechanism of fasting natriuresis and glucose-induced sodium retention was evaluated by means of spironolactone treatment (Aldactone A; 300 mgfday p . 0 . ) on days 6 to 8 of the fast, with an oral glucose load (100 g) on day 7. Aldosterone infused on day 4 of the fast caused on average only 40 % of the antinatriuretic effect it achieved before the fast. On the other hand even before aldosterone infusion, salivary sodium output was markedly reduced during the fast to levels comparable to those observed after aldosterone treatment in the pre-fast period. Furthermore, spironolactone administration on day 6 of the fast was associated with a prompt and marked increase in natriuresis. These 3 sets of facts indicate a definite biological activity of aldosterone during the initial phase of fasting with factor(s) interfering at the renal level with the normal expression of the hormonal action on sodium balance. There was still a distinct antinatiuretic effect of glucose in the presence of spironolactone, but less pronounced than when glucose was administered alone. The well-documented hyperaldosteronism of a total fast may represent a compensatory mechanism to decrease sodium loss at the end of a week-long fast. In addition a marked sodium-retaining effect of glucose can be demonstrated after aldosterone action is blocked by spironolactone. This provides another indication that glucose stimulates sodium retention through mechanism(s) which do not involve aldosterone. Key words: Sodium balance, aldosterone, total fast, refeeding, obesity.

Introduction Several investigators have confirmed the original reports of Bloom and Mitchell (2, 3) that a total fast is initially characterized by enhanced natriuresis which lasts for the first three to five days and is usually followed by a progressive reduction in sodium excretion ( 4 - 8). Carbohydrate administration causes a prompt and drastic reduction of sodium excretion during this initial phase of negative sodium balance. The underlying mechanisms are not yet fully understood ( 5 , 8 - 1 2 ) . During the first week of a fast investigators have noted, with a few exceptions (6, lo), that a progressive increase in aldosterone secretion occurs, as indicated by measurements of secretion rate ( 5 , 8, 15, 17, 18), plasma level (16) and "Presented in part at the eighth annual meeting of the European Society for Clinical Investigation, Rotterdam, The Netherlands, April 1974 ( l ) . . 2XDepartment of Medical Biochemistry, Cliniques Universitaires St. Pierre, Louvain

urinary excretion (13, 1 4 ) . The negative sodium balance which occurs during the first days of starvation while aldosterone secretion is enhanced, has led to the suggestion that the nephron somehow becomes "refractory" to the sodium-retaining action of mineralocorticoids ( 1 , 18, 19). This would be a transient event and the spontaneous re-equilibration of sodium balance observed at the end of a week-long fast would reflect recovery of renal tubular sensitivity to aldosterone (8, 18). Actually, inhibition of aldosterone action by spironolactone during a fast is associated with enhanced natriuresis (20, 21). Although aldosterone probably plays a role in the regulation of sodium balance during fasting, it is uncertain that this hormone is involved in the glucose-induced antinatrimesis classically observed in these conditions. Data reported by Gersing and Bloom (20) and Hansen e t aZ. (221, suggest that the mechanism by which glucose lowers sodium excretion operates independently of aldosterone since spironolactone did not prevent the antinatriuretic effect of carbohydrates. However according to Boulter et aZ. (21), sodium retention did not set in when the glucose load was administered to fasted obese subjects receiving

76

J. Kolanowski e t ai!.: Sodium Balance and Renal Tubular Sensitivity

spironolactone. However the dosage was unusually high in their study. In the present study the renal tubular responsiveness to aldosterone was estimated by infusing aldosterone into obese subjects before and on day 4 of a total fast, when natriuresis is usually at its peak (8, 1 1 , 1 2 ) . This was to assess the alleged refractoriness to mineralocorticoids during starvation. The extra-renal effects of aldosterone were evaluated by measuring concomitantly salivary sodium output in some of the patients. To determine the possible role of aldosterone at the time of reduced natriuresis which OCCUKS at the end of a weeklong fast, sufficient spironolactone was given from day 6 to 8 of the fast to block aldosterone action in patients suffering from hyperaldosteronism ( 2 3 ) . Glucose was ingested on day 7, i.e. on the second day of this treatment.

The data concerning body weight l o s s and urinary sodium and potassium excretion by the 12 female patients studied were compared to those of the 20 obese women studied similarly except for the administration of aldosterone and spironolactone. Mixed saliva was also collected from four of the patients after paraffin chewing stimulation, before and after aldosterone infusions. On each day saliva was collected for 30 min. periods, at I 1 a.m. and at 7 p.m. These times were i m e diately before and 5 hours after completion of the aldosterone infusions. The salivary volume

ib1ateriaZ and Methods Studies were carried out on 12 obese nondiabetic women whose body weight was 41 C 6 % (SEM) above standard weight. There was no evidence of pathology apart from exogenous obesity, and no diuretics were taken for at least three weeks before the study. After a control period of 5 days on a 1500 kcal sodium-free diet, supplemented daily with 4 2 . 6 mmol sodium chloride and 4 0 . 5 mmol potassium gluconate taken orally, the subjects were submitted to an 8-day total fast with a single oral 100 g glucose load on day 7. Throughout the period of fasting the subjects continued to receive the same supplements of sodium and potassium. Water intake varied from 1 . 5 to 2.5 l/day. Aldosterone (0.5 mg Aldocorten ( R ) ) was infused 1.V. in 300 ml of 0.85 % sodium chloride over 3 hours on the penultimate day of the control period and on day 4 of the fast. The oral NaCl supplement was withdrawn on these days. From day 6 to 8 of the starvation period the patients received spironolactone (Aldactone A (€3 100 mg p.0.1 every 8 hours. Thus the oral glucose load was taken on the second day of the spironolactone treatment. Patients were weighed every morning on rising. Urine was collected daily from the third day of the control period and venous blood was d r a m into heparin on the third day of the control 1500 kcal diet, on days 1 , 3 , 5, 7 of the fast and 48 hours after the oral glucose load, immediately before refeeding (with a 1000 kcal diet) from day 9 on. Samples of whole blood were analysed for glucose and the haematocrit. Plasma was used for measurements of total protein, urea, urate, creatinine, bicarbonate, sodium, chloride, potassium, calcium, magnesium and phosphate levels. Creatinine, urate, sodium, chloride, potassium, phosphate, calcium and magnesium were measured in fresh 24 hour urine samples. Calcium and magnesium were determined by atomic absorption. All other measurements were made using Technicon Auto-Analyser equipment.

y,

6 -

b , ,

,

,

,

,

,

I

,

,

,

I

Urinary K (mEq124h)

U r i n a r y Na

-?!-+

(mEq / 2 L h 1 A

100

j

1

A

GI

A

4

1

4

-

-3-2-1

1500 k c a l

I 1 3 L 5 6 7

Bdays-3-2-1 b

Total

GI

4 . 1

1 2 3

L 5 6

7 Bdays

__*

1500 k c a l fast

A

Total f a s t

Fig. I . Changes in body weight and in urinary sodium (Na) and potassium (K) excretion induced by aldosterone infusion (A; Aldocorten(R) 0.5 mg I . V . in 300 ml of 0.85 % saline over 3 h), spironolactone (Sp; Aldactone A, 100 mg q . 8 h p . 0 . ) and glucose (G1; 100 g p . 0 . ) in 12 obese patients undergoing an 8-day total fast (hatched columns). The data are compared to those obtained from 20 starving obese subjects fasting for 8 days. Both groups received a glucose load on day 7 but the control group did not receive aldosterone or spironolactone (dotted line for body weight loss and background open columns for Na and K excretion). Days indicated in figure as - 3 , - 2, - I correspond t o the last 3 days of the 5-day control, 1500 kcal diet period. A l l subjects were females

J. Kolanowski e t ai!.: Sodium Balance and Renal Tubular Sensitivity was measured and the sodium concentration determined by the Technicon Auto-Analyser method. The differences between mean values were analysed statistically using the Student's ttest for paired values within the aldosteronetreated group. The unpaired t-test was used when the results obtained from aldosterone treated patients were compared to those of the control group.

77

crease in potassium excretion on the day of aldosterone infusion (day 2), while kaliuresis tended to be spontaneously lower at that time (Fig. I ) . After institution of the fast, natriuresis increased progressively in aldosterone-treated subjects so that on day 3 urinary sodium excretion was comparable to that usually observed in starving obese subjects (Fig. I ) . Aldosterone infusion on the following day (day 4) reduced this natriuresis (p < 0.05), but to a lesser degree (p < 0.01) than before the fast. Furthermore sodium balance remained negative at this stage of the fast despite aldosterone administration. When changes in natriuresis resulting from the steroid administration are evaluated against the profile which is obtained in the absence of exogenous aldosterone, the hormonal effect on sodium excretion appears to be markedly reduced during the fast. Thus sodium excretion was 20 mEq/24h while it averaged 50 mEq/24h before the fast. If the 48 hour period after aldosterone infusion is considered, the interfering effect of the fast is even more clearcut. This is shown in Fig. 1. The administration of spironolactone on day 6 caused a prompt enhancement of natriuresis (p < O.OOl), whereas sodium excretion usually dropped from day 5 to 6 in control subjects (Fig. I ) . Despite spironolactone administration glucose ingestion was followed by a greater reduction of

Resu i!t s 1 . Changes i n Body Weight and i n Sodium, Chloride and Potassium Balance

The data presented in Fig. 1 indicate that sodium balance was usually negative during the last three of the 5 days on a 1500 kcal diet, and that a profound and sustained reduction in natriuresis (p < 0.OOl) resulted from aldosterone infusion on the penultimate day of this period (day 2). Although no lasting changes were observed in plasma sodium, chloride and potassium concentration (day 1 of fast 0s. day 3 , Table I ) the aldosterone-induced decrease in sodium and chloride excretion persisted for at least 48 h (Table 2). The urinary profile of natriuresis was significantly different (p < 0.01) from that observed in control obese subjects under the same conditions (Fig. I ) . There was a small in-

Table 1. Plasma ionic composition, total proteins and haematocrit in 12 obese females before and during a week-long fast (mean values f SEM) Days of fast 5

Days of experiment

1500 kcal diet (day-3)

Sodium

144.3 f 0.5

142.5

0.7

141.2 f 0.6

141.4 f 0.6

141.5

103.5 2 0.6

102.5 f 0.7

101.7 f 0.6

101.5 f 0.9

101.2 f 0.6

3

1 f

9X

79

f:

0.8

141.6 f 0.8

(mEq/L)

Chloride (mEq/L) Potassium (mEq/L) Calcium (mg/100 ml)

3.83 f 0.07

4.09

0.14

4.44

f

9.56 f 0.11

9.66 f 0.12

9.77

3.33

0.11

3.52

1.67 f 0.03

1.40

Phosphate (mg/100 ml)

3.08

f 0.13

Magnesium (mEq/L) Bicarbonate (mEq/L) Proteins (g/IoO mi)

1.57

f

26.07 -+ 0.82

Haematocrit (46)

41.82 -+ 0.42

6.72

2

0.02

0.18

27.21

f:

f

k

0.53

6.75 f 0.14 40.55

k

0.77

0.19

4.35 t 0.I I

4.05 f 0.08

k 0.10

9.81

2 0.11

9.94

f

0.22

3.09

f

k

0.03

22.62 2 1.23 6.81

f

0.13

41.41 f 1.02

98.9

f

4.17

f 0.12

0.8

0.14

10.24

f

0.19

0.11

2.96 f 0.16

3.33

f

0.19

1.39 f 0.05

1.36 f 0.02

1.27 f 0.04

0.91

17.98 -+ 0.60

19.38 7.21

f

k 0.12

41.64 f 0.89

f

7.35 f 0.14 42.73 f 1.19

23.42

_+

0.88

7.32 f 0.19

44.18

$24 hours after administration of spironolactone (Aldactone A, 300 mg/day P.o., on days 6, 7 and 8. '48 hours after an oral glucose load (100 g)

f

0.90

J . Kolanowski e t a l . : Sodium Balance and Renal Tubular Sensitivity

78

for the same 8-day period was 6 5 . 2

natriuresis (p < 0.OOl) than in the control group. However, in absolute terms, natriuresis on days 7 and 8 remained significantly higher in the presence of the aldosterone antagonist (p < 0.01). Changes in chloride excretion in response to aldosterone and spironolactone treatment during the fast (Table 2 ) paralleled the sodium changes. On the other hand the chloride retention resulting from the glucose load during spironolactone administration was only half of that observed for sodium (p i 0.01). The pattern of potassium excretion during the fast was comparable in the aldosterone-treated and the control groups (Fig. 1 , Table 2 ) , except for a significant drop on day 6 after spironolactone administration (p < 0.001). The glucose-induced antikaliuresis averaged 26 mEq/24 h in the control starving obese subjects and was almost as marked ( 2 1 mEq/24 h) when the biological effects of aldosterone were prevented by spironolactone treatment. Plasma sodium levels dropped slightly during the initial phase of fasting and then stabilized. However the plasma chloride concentration decreased progressively especially after glucose administration (Table I ) . The increase in plasma potassium level noted during the first days of the fast (Table 1 ) was not significant. The sodium loss, calculated from the sodium intake and urinary excretion values for the 8-day period which followed the start of the fast, averaged 149.4 i 30.3 (SEM) mEq. The mean chloride deficit was 1 4 2 . 4 2 3 2 . 2 (SEM) mEq. However it should be noted that the sodium loss exceeded the chloride loss during the first 3 days of the fast while the opposite was observed after spironolactone and glucose treatment (Table 2 ) . The mean cumulative potassium loss

Changes in body weight loss

?: 18.5

(SEM)

mEq

throughout the whole experimental period are shown in Fig. I . Exogenous aldosterone only slowed down weight reduction during the 1500 kcal diet period. The mean body weight loss from day 4 t o day 5 of the fast, i.e. following aldosterone infusion, was not different from that usually observed on these days of starvation. On the other hand glucose administration on day 7 during the spironolactone treatment, significantly reduced body weight l o s s (day 6 to 7 u s . day 5 to 6 ; p c 0.001). This weight loss was smaller than when control starved patients took glucose at the same stage of the fast (Fig. 1 ) .

2. Changes i n U r i n u r y Excretion and i n Plasma Levels of Phosphate, Calciwn and Magnesium The data recorded in Table 2 indicate that the pattern .of phosphate excretion during the fasting period was similar to that of natriuresis except that neither aldosterone nor spironolactone administration influenced it. There was a significant correlation between phosphate and sodium excretion during the first 3 days of the fast (r = 0 . 5 9 ; p < 0.001) and again on days 7 and 8 of the fast, following glucose ingestion (r = 0 . 5 6 ; p < 0.01). This decrease in phosphaturia from day 5 to 6 of the fast (Table 2) is probably unrelated to aldosterone administration as similar changes occur in control subjects at this stage of a fast ( 1 2 ) . Furthermore, spironolactone failed to influence the amplitude and/or the duration of the reduction in phosphaturia caused by the glucose load on day 7 . The plasma

Table 2 . Urinary ion excretion in 12 obese females before and during a week-long fast (mean values f SEM) 1500 kcal diet

Days of experiment

-3

Sodium

76.7 14.9

-2 1 )

Days of fast

-1

I

3

2

4l)

5

733)

62)

82)

i

33.9 4.3

?

32.1 6.1

i

45.9 4.9

79.3 f 14.2

f 14.5

f.

56.6 6.0

f

54.6 4.8

94.6 f 11.9

f

34.4 5.2

?

39.5 6.8

2 15.6

C

43.2 6.4

?

43.5 6.0

f

51.8 5.4

?

52.1 6.6

67.4 f 12.1

f

46.9 5.4

f

54.4 4.1

90.7 f 11.9

f

60.9 6.1

f

60.2 6.2

c

65.1 5.4

i

73.9 6.8

?

60.6 4.9

f

57.9 4.5

f

53. I 4.6

f

64.8 6.0

f

66.3 5.1

?

59.0 5.1

f

43.5 1.9

f

il.9 1.3

f

22.7 2.5

Phosphate (mg/24 h)

?

526 54

2

635 53

2

532 39

f

558 35

i

7 99 73

f

94 6 79

f

967 84

?

845 55

?

639 50

?

338 58

f

299 59

Calcium (mg/24 h)

2

121.2 18.1

2

Magnesiurn ( m g / 2 4 h)

I

(mEq/24 h)

?

Chloride (mEq/24 h) Potassium (mEq/24 h)

89.1

67.5 7.3

101.2

104.1

11.0

i: 14.2

80.3 2 10.5

t

75.2 5.8

64.7 f 14.1 f

47.0 6.0

116.0 i 23.5

f

76.6 9.2

86.5

151.7 i 27.5 f

91.6 6.8

f.

84.7 5.9

f ~~

"Aldosterone infusion (0.5 mg in 3 h in 3 0 0 ml of 0 . 8 5 2)Spironolactone (Aldactone A , 3 0 0 mg/day p.0.) 3)Oral glucose load (100 g)

saline)

209.5 k 19.3

190.0 30.0

f.

~

72.6 3.6

274.7

237.8

k 20.4

? 24.6

f

73.1 6.7

t

51.4 4.6

205.5 i: 3 8 . 5

f

40.9 4.9

J. Kolanowski e t a l . : Sodium Balance and Renal Tubular Sensitivity

inorganic phosphate (Table I ) concentration did not vary significantly. There was an abrupt drop in urinary calcium and magnesium excretion on the first day of the fast (Table 2) followed by a progressive increase in excretion. There is a highly significant (p < 0.OOl) correlation between calcium and sodium excretion (r = 0 . 7 0 ) as well as between magnesium and sodium excretion (r = 0 . 6 0 ) during the first days of the fast. Although there were no significant changes in calcium and magnesium excretion after aldosterone or spironolactone treatment, the administration of glucose caused the urinary excretion of these ions to decrease on day 7 . This reduction was only statistically significant for magnesium (p < 0.01). The progressive increase in urinary calcium loss during the fast was associated with a slight and insignificant rise in plasma calcium, plasma protein

79

and haematocrit values (Table I ) . Magnesium levels dropped significantly (p < 0.001) as fasting continued . 3. Changes i n Glycaemia and i n Renal Handling of Urea, Urate and Creatinine

During the fast, glycaemia dropped rapidly at first (p 0.001), to stabilize after the third day. 48 h after an isolated glucose load on day 7 glycaemia was still elevated and comparable to prefast values (Table 4 ) . Plasma creatinine concentrations (Table 4 ) rose significantly (p < 0.001) during the week of the fast but decreased slightly 48 h after glucose administration. The urinary creatinine levels (Table 3 ) did not change during the period of study. Both plasma urea concentrations

Table 3 . Urinary volume and creatinine, urea and urate excretion in 12 obese females before and during a week-long fast (mean values f SEM) Days of experiment Urinary volume ( m 1 / 2 4 h)

-3

f

1578 255

1.23 Creatinine f 0.08 ( g / 2 4 h) Urea 16.5 f 1.1 ( g / 2 4 h) Urate 646.8 f 46.4 (mg/24,h)

1500 kcal diet -2 1 ) -1

2

I574 168

I .32 f 0.11 fi

17.5 1.7

693.2 f 38.1

Days of fast

1377

1382 237

-+ 319

1.28 f 0.09

f 0.09

f

f

14.9 1.3

571 .O 2 29.5

1.25

f

0.8

433 .O f 49.1

1596 283

1.39 f 0.13

11.8

f

3

2

1

f

16.3 1.6

277.3 f 43.0

f

4l)

1671 260

1.34 f 0.09

f

I .34 f 0.11

15.6 f

1.2

1518 214

f

14.4 1.0

241.6

223.3

fi 23.8

f 25.3

5

62>

1392 f

180

1.32 f 0.07 f

11.4 0.8

275.2 f 31.3

7213)

82)

fi

1372 259 1.22

1.18

f 0.08

c

0.08

2 0.08

10.9 0.6

fi

8 .O 0.5

?

551.4 54.2

f

1513 253 1.35

_+

266.5 fi 23.0

2

f

I170 271

6.5 0.7

305.4 f 23.7

"Aldosterone infusion ( 0 5 mg in 3 h in 300 ml of 0.85 2 saline) *)Spironolactone ( 3 0 0 mg/day p . 0 . ) 3)Oral glucose load (100 g)

Table 4 . Changes in glycaemia and in plasma urea, urate and creatinine levels in 12 obese females during a week-long fast (mean values 2 SEMI Days of experiment

1500 kcal diet (day-3)

Glycaemia

81.7 f 2.4

79.3 f 2.8

55.4 f 4.0

5 3 . 3 f 1.7

54.5 f 2.2

74.4

31.4 f 2.0

27.9 f 1.8

30.1 f 1.4

27.8 f 1.8

23.2 f 1.8

21.6 f 2 . 8

5.6 f 0.4

5 . 8 f 0.4

8 . 4 f 0.6

11.2 f 0.6

12.8 f 0.7

11.0 f 0 . 8

0.96 f 0.04

0.97 f 0.03

1.26 f 0.08

1.54 f 0.07

1.52 f 0.05

1.34 f 0.05

1

3

Days of fast 5

9X

79

2

5.3

(mg/iOO ml)

Urea (mg/ 100 ml)

Urate (mg/100 mij Creatinine (mg/100 ml)

$24 hours after administration of spironolactone (Aldactone A, 300 mg/day P . o . , on days 6 , 7 and 8 . x48 hours after an oral glucose load (100 g).

J. Kolanowski e t a z . : Sodium Balance and Renal Tubular Sensitivity

80

Table 5. The influence of aldosterone infusion on salivary flow and sodium output before and on day 4 of a total fast (mean values t SEM; n = 4 ) 1500 kcal diet

Period

Total fast

7 p.m.

7 p.m.

7 0.m.

1 1 a.m.§

Salivary flow (m1/30 min.)

10.8

10.8

9.8

c 2.3

c 4.1

e 2.5

c 1.1

f. 0 . 5

c 1.1

54.0

34.0

25.7

21.6

28.0

k 3.9

t 2.4

Sodium output ( p E q / 3 0 min.)

c18.5

ir12.3

~

1 1 a.m.

1 1 a.m.§

Hour

8.5

9.30

5

5.3

7.1

5

7 p.m. 9.8

1 1 a.m. 6.0

c 4.9

c 1.0

20.4

19.7

25.7

3.5

c 4.1

ell

.o

'Saliva collected immediately before aldosterone infusion (0.5 mg in 300 ml 0.85 Z saline I . V . over 3 hours) (Table 4 ) and urea excretion values (Table 3) dropped to a comparable extent, in keeping with the progressive reduction in urea production during the fast. This decrease was accentuated by glucose administration. There was a rapid drop in urinary urate excretion on starvation (Table 3) with a concomitant rise in plasma urate levels (Table 4 ) . This indicates an important reduction in renal clearance of urates. On the day glucose was administered there was a significant increase in urinary urate (p < 0.01; Table 3). This increase in excretion was not noticed on the following day. The drop in plasma bicarbonate level during the fast (Table 1; p < 0.001) was also corrected by glucose administration. The behaviour of these organic compounds in the 1 2 obese subjects who underwent this study was comparable to the changes usually observed during a fast and following glucose administration ( 1 2 ) . Thus aldosterone and spironolactone obviously had no effect here. 4 . Influence o f Aldosterone on Salivary Sodium Excretion

Aldosterone infusion during the control 1500 kcal period (on day 2) caused a significant reduction in sodium output in saliva, (p < 0.01) determined in 4 subjects 5 h (day 2 , 7 p.m.) and 24 h (day 1 , 1 1 a.m.) later, and compared to control values obtained before aldosterone treatment. There were no appreciable changes in salivary flow (Table 5). After three days of total fast salivary flow and sodium output were significantly reduced (p < 0.01) when compared to prefast control values and aldosterone infusion'only caused a small, insignificant further drop in sodium output.

Discussion Despite intensive efforts the mechanism of sodium excretion and antinatriuretic properties

of carbohydrates during fasting are still poorly understood. Whether the increased secretion of aldosterone during fasting (5, 8, 13 - 18) is transiently prevented from acting and/or is involved in the striking sodium-retaining effect of glucose is also unsettled. I . Renal Tubule and Salivary Gland Response t o Aldosterone h A n g Fasting

Although the early phase of a total fast is usually associated with negative sodium balance, there are indications that aldosterone is biologically active at this stage. The I . V . injection of this hormone on the 6th day of the fast caused a clearcut reduction of urinary sodium (and chloride) excretion (8). Injection on the 4th day (as in the present study) when natriuresis is usually close to or at its peak, was also followed by retention of sodium (and chloride). Furthermore, administration of spironolactone on day 6, to prevent endogenous aldosterone reaching its receptor site(s), resulted in a de.finite increase in sodium (and chloride) excretion. It is therefore difficult to conclude that the nephron is even transiently refractory to aldosterone during a fast (18, 19, 2 1 ) . On the other hand, it is clear that the sodiumretaining effect of aldosterone was smaller on the 4th day of the fast than a week previously (Fig. 1 ) . Indeed, while sodium intake was kept constant throughout, with comparable natriuresis on the days preceding the intravenous aldosterone infusions, sodium excretion only dropped significantly below the intake level before the fast as a response to the hormone. Using the natriuresis profile observed in 20 female obese subjects not given aldosterone as a reference, it can be seen that the sodium-retaining effect of aldosterone during the fast was less than half its effect before the fast. Since the secretion and plasma level of aldosterone increase rapidly during the first

J. Kolanowski e t a1.t Sodium Balance and Renal Tubular Sensitivity

days of a fast (5, 8, 15 - 18), the weaker sodium-retaining effect of infusion on day 4 of the fast probably reflects a decrease in available mineralocorticoid receptors for exogenous aldosterone. Determination of salivary sodium output before and during a total fast agrees with this interpretation. In fact, the salivary sodium output on day 3 of the fast, i.e. before the second aldosterone infusion, was comparable to that observed after aldosterone administration during the control period. Only a small further reduction in salivary sodium occurred following aldosterone infusion on day 4 of the fast. The mechanism of the natriuresis seen early in the fast might be due to increased sodium delivery to the distal tubule, resulting from decreased sodium reabsorption in the proximal tubule and/or in the diluting segment of the nephron ( 4 , 7 , 2 4 ) . The increase in urinary phosphate, calcium and magnesium excretion taken together with a profoundly reduced renal urate clearance, suggest that during early fasting transport mechanisms operating proximally to aldosterone-mediated sodium reabsorption in th8 distal tubule are less efficient. These changes in urinary excretion cannot be accounted for by the increased protein breakdown associated with starvation. For instance, the increase in phosphate excretion during the first days of the fast corresponded to more than twice the amount which could be ascribed to protein catabolism estimated from urea excretion ( 2 5 ) . Renal phosphate clearance calculated from values given in Table 1 and 2 increased 70 Z after a few days of the fast. Although urinary phosphate excretion is not an ideal parameter of proximal tubular reabsorption, most of the filtered phosphate is effectively reabsorbed in the proximal tubule and a close relationship exists between reabsorption of sodium, phosphate and calcium in the proximal tubule ( 2 6 ) . It is interesting that there was a significant correlation between phosphate and sodium excretion during the fast in our patient. The decreased proximal sodium reabsorption early in fast could arise from an enhanced organic anion excretion at that stage with "obligatory" sodium coverage as long as renal ammonia excretion fails to match it ( 2 7 , 2 8 ) . Thus, if renal ammoniogenesis is stimulated by chronic acidification before a fast, the fast is no longer characterized by a phase of negative sodium balance ( 2 9 ) . 2. Atdosterone and GZucose-Induced Sodium

Re tention

Glucose-stimulated sodium reabsorption is said to occur mainly in the proximal segment of the nephron ( 4 , 7 , 2 0 , 2 4 , 3 0 ) . It has been suggested, however, that both the proximal and distal tubules are involved in the enhancement of sddium retention (7) when glucose is given to fasting subjects. It seems unlikely that the

81

glucose-induced sodium retention simply reflects increased tubular sodium transport associated with, or resulting from enhanced glucose reabsorption secondary to an increased glucose filtered load, since sodium retention usually occurs after the period of hyperglycaemia ( 7 , 3 1 ) . The data from the literature concerning a role for aldosterone in the sodium retention induced by glucose after a few days of fasting are controversial. According to some investigators ( 2 0 , 2 2 ) spironolactone fails to interfere with glucose-induced antinatriuresis. However, Boulter e t a t . ( 2 1 ) recently reported that during a fast spironolactone prevents the antinatriuretic effect of glucose, but very large amounts of sprionolactone (1200 mg per day) were used, and it is not inconceivable that under such conditions the effects of spironolactone are no longer limited to the displacement of aldosterone from its receptor sites ( 3 2 ) . The present investigation indicates that after administration of a widely accepted dosage of spironolactone which blocks most aldosteronedependent sites of renal sodium reabsorption the administration of glucose to fasted subjects still results in a clearcut decrease in sodium excretion. Thus glucose appears to effectively stimulate, directly or indirectly, renal sodium retention by a mechanism bypassing aldosterone. In addition no change in aldosterone secretion rate could be demonstrated after glucose ingestion during a total fast (8). Because this glucose-induced sodium retention is associated with a reduction in phosphate and magnesium excretion and with a rise in renal ur'ate clearance, it appears that glucose administration during fasting may influence transport mechanisms at sites proximal to aldosterone-mediated sodium reabsorption. It has been suggested that glucagon rather than aldosterone causes the fluctuations in sodium balance described here. Indeed, the glucagonaemia profile during a fast parallels that of natriuresis ( 1 2 , 1 9 , 3 3 ) . Thus fasting hyperglucagonaemia is rapidly reduced when glucose is given to starved subjects while glucagon administration enhances the natriuresis associated with fasting ( 1 9 , 3 4 ) and prevents the antinatriuretic properties of carbohydrates ( 3 1 , 3 3 , 3 4 ) . However, this renal effect of glucagon appears complicated, as is shown by the fact that this hormone does not interfere with the stimulating effects of aldosterone, insulin and glucose on sodium transport of amphibian epithelia i n vitro ( 3 5 ) . More significant yet, glucagon fails to abolish the glucose-induced antinatriuresis when the hormone is infused into starved subjects the day after a glucose load ( 3 4 ) . Thus we can conclude that during the transient phase of negative sodium balance following the beginning of,a fast, aldosterone remains biologically active. On the other hand the marked sodium-retaining effect exerted by glucose at this stage can be demonstrated when aldosterone action is blocked. The hyperaldoster-

a2

J. Kolanowski e t a l . : Sodium Balance and Renal Tubular Sensitivity

aldosterone. Canad. med. Ass. J. 98, 1034 ( 1968) 15. Herman, L.S., Hansen, N.C., Grdnbaek, P., Inversen, M.: Hyperaldosteronisme after affedning met absolut sult. Ugeskr. Laeg. Acknow~edgments. We are indebted to Dr. Tyberg131, 192 (1969) hein from CIBA-GEIGY S.A. Laboratories (Be1 ium) 16. Chinn, R.H., Brown, J.J., Fraser, R., Heron, for kindly providing us with the Aldocorten7R) S.M., Lever, A.F., Murchison, L., Robertson, used for this study. J.I.S.: The natriuresis of fasting: relationship to changes in plasma renin and plasma aldosterone concentration. Clin. Sci. 39, 437 ( 1 970) References 17. Garnett, E.S., Cohen, H., Nahmias, C., Viol, G.: The roles of carbohydrate, renin and 1 . Kolanowski, J., Desmecht, P., Crabb’e, J.: aldosterone in sodium retention during and Changes in renal tubular response to aldosterone after total starvation. Metabolism 22, 867 during total fast in the obese. Europ. J. clin. ( I 973) Invest. 4, 375 (1974) 18. Boulter, P.R., Spark, R.F. , Arky, R . A . : DisBloom, W.L., Mitchell, W.R.: Salt excretion in sociation of the renin-aldosterone system fasting patients. Arch. intern. Med. 106, 321 and refractoriness to the sodium-retaining (I 960) action of mineralocorticoid during starvation Bloom, W.L., Mitchell, W.R.: Inhibition of in man. J. clin. Endocr. 38, 248 (1974) salt excretion by carbohydrate. Arch. intern. Med. 109, 26 (1962) 19. O’Brian, J.T., Saudek, C.D., Spark, R.F., Wright, H . K . , Gann, D.S., Albertsen, K.: Arky, R.A.: Glucagon-induced refractoriness Effect of glucose on sodium excretion and to exogenous mineralocorticoid. J. clin. renal concentration ability after starvation Endocr. 38, 1147 (1974) in man. Metabolism 12, 804 (1963) 20. Gersing, A., Bloom, W.L.: Glucose stimulation of salt retention in patients with aldosterone Rapoport, A., From., G.L., Husdon, H.: Metabinhibition. Metabolism 1 1 , 329 (1962) olic studies in prolonged fasting. I. Inorganic metabolism and kidney function. Metab21. Boulter, P.R., Spark, R.F., Arky, R.A.: Effect of aldosterone blockade during fasting and olism 14, 31 (1965) 6. Katz, A.I., Hollingsworth, D.R., Epstein, F.H.: refeeding. Amer. J. clin. Nutr. 26, 397 (1973) Influence of carbohydrate and protein on sodium 22. Hansen, E.L., Hdrlyck, E., Grdnbaek, P., excretion during fasting and refeeding. J. Lab. Iversen, M.: Hyperaldosteronism following clin. Med. 72, 93 (1968) total fasting in obese subjects. Acta med. 7. Schloeder, F.X., Stinebaugh, B.J.: Renal tubuscand. 182, 65 (1967) lar sites of natriuresis of fasting and glu23. Liddle, G.W.: Specific and non-specific inhicose-induced sodium conservation. Metabolism bition of mineralocorticoid activity. Metab19, 1119 (1970) olism 10, 1021 (1961) 8. Kolanowski, J., Pizarro, M.A., De Gasparo, 24. Hoffman, R.S., Martino, J.A., Wahl, G., Arky, M . , Desmecht, P., Harvengt, C., Crabbi?, J.: R.A.: Effects of fasting and refeeding. 11. Influence of fasting on adrenocortical and Tubular sites of sodium reabsorption and pancreatic islet response to glucose loads effects of oral carbohydrates on potassium, in the obese. Europ. J. clin. Invest. 1, 25 calcium and phosphate excretion. J. Lab. ( I 970) clin. Med. 74, 915 (1969) 9. Stinebaugh, B.J., Schloeder, F.X.: Studies 25. Reifenstein, E.C.,Jr., Albright, F., Wells, on the natriuresis of fasting. I. Effect of S . L . : The accumulation, interpretation and prefast intake. Metabolism 15, 828 (1966) presentation of data pertaining to metabolic 10. Smith, R., Ross, E.J., Marshall-Jones, P.: balances, notably those of calcium, phosAldosterone and sodium excretion in obese phorus and nitrogen. J. clin. Endocr. 5, 367 subjects on water diet. Metabolism 18, 700 ( I 945) ( I 969) 26. Agus, Z.S., Gardner, L.B., Beck. L.H., GoldI I . Boulter, P.R., Hoffman, R.S., Arky, R.A.: berg, M.: Effects of parathyroid hormone on Pattern of sodium excretion accompanying renal tubular reabsorption of calcium, sodium starvation. Metabolism 22, 675 (1973) and phosphate. h e r . J. Physiol. 224, 1143 I ? . Kolanowski, J., Desmecht, P., Crabb’e, J.: ( I 973) R&percussions mEtaboliques et hornonales du 27. North, K.A.K., Lascelles, D., Coates, P.: The j e h e total de courte durhe chez l’obdse. mechanisms by which sodium excretion is inSchweiz. med. Wschr. 104, 1022 (1974) creased during a fast but reduced on subse13. Haag, B.L., Reidenberg, M.M., Shuman, C.R., quent carbohydrate feeding. Clin. Sci. Molec. Channick, B.J.: Aldosterone, 17-hydroxy, Med. 46, 423 (1974) corticoid and fluid and electrolyte responses 28. Sigler, M.H.: The mechanism of the natriuresis to starvation and selective reefeding. h e r . of fasting. J. clin. Invest. 55, 377 (1975) J. med. Sci. 254, 652 (1967) 29. Schloeder, F.X., Stinebaugh, B.J.: Studies 14. Verdy, M., Champlain, J. de: Fasting in obese on the natriuresis of fasting. 11. Relationfemales. 11. Plasma renin activity and urinary ship to acidosis. Metabolism 15, 838 (1966) onism associated with a total fast probably represents a compensatory mechanism whereby sodium losses become reduced after a few days of starvation.

J. Kolanowski e t al.: Sodium Balance and Renal Tubular Sensitivity 3 0 . Lenon, E.J., Lemann, J.,Jr., Piering, W.F.,

Larson, L.S.: The effect of glucose on urinary cation excretion during chronic extracellular volume expansion in normal man. J. clin. Invest. 5 3 , 1424 ( 1 9 7 4 ) 31. Kolanowski, J., De Gasparo, M., Desmecht, P., Crabb’e, J.: Further evaluation of the role of insulin in sodium retention associated with carbohydrate administration after a fast in the obese. Europ. J. clin. Invest. 2 , 439 (1972) 3 2 . Crabbb, J.: Inhibition by spironolactone of

the effect of aldosterone on transepithelial sodium transport. In: Extrarenal activity of aldosterone and its antagonist. Brendel et al. (Eds.), Excerpta med., p . 7 . Amsterd,am 1972 3 3 . Saudek, C.D., Boulter, P.R., Arky, R.A.: The natriuretic effect of glucagon and its role in starvation. J. clin. Endocr. 3 6 , 7 6 1 ( 1 9 7 3 )

83

3 4 . Kolanowski, J., Salvador, G., Henquin, J.C.,

Desmecht, P., Gerich, J., Karam, J.H., Crabb’e, J.: Influence of glucagon on sodium (Na) balance during fasting and carbohydrate refeeding in the obese. Europ. J. clin. Invest. 3 , 244 (1973) 3 5 . Crabbb, J.: Insulin, glucagon and active

sodium transport: from man to amphibia and back. In: Transport mechanisms in epithelia. H.H. Ussing and N.A. Thorn (Eds.), p. 1 7 3 . Copenhagen: Munksgaard 1973

Prof. Dr. J. Crabbb Department of Physiology Tour Harvey UCL 5 5 3 0 Avenue Hippocrate, 5 5 B-1200 Bruxelles, Belgium