British Journal of Obstetrics and Gynaecology May 1975. Vol. 82. pp. 374-381.
THE EFFECT OF PREGNANCY ON THE RENAL HANDLING OF GLUCOSE BY
J. M. DAVISON AND
F. E. HYTTEN MRC Reproduction and Growth Unit Princess Mary Maternity Hospital Great North Road, Newcastle upon Tyne NE2 3BD Summary Tubular reabsorption of glucose has been measured during glucose infusion in 29 healthy women during and after pregnancy. All the women had normal glucose tolerance to an oral load, and normal glucose excretion when not pregnant, but exhibited a wide range of daily glucose excretion in pregnancy. Throughout pregnancy the renal reabsorption of glucose is less effective than in the nonpregnant state and, in general, the greater the amount of glycosuria which develops in pregnancy, the less effective is the reabsorption during infusion. Post partum, women with minor degrees of glycosuria during the preceding pregnancy return to a normal highly efficient reabsorption performance during infusion, but women who exhibit greater degrees of glycosuria have a reduced capacity to reabsorb even though they are no longer glycosuric after the pregnancy. It is concluded that pregnancy imposes some specific change in the glucose reabsorptive capacity of the proximal tubule and that women with more than usual degrees of glycosuria in pregnancy may, in addition, have an element of tubular damage. This is discussed in relation to other renal function changes in pregnancy in an attempt to explain the characteristic intermittency of clinical glycosuria in pregnancy.
HEALTHY pregnant women excrete more than the usual amounts of glucose in urine. The losses may be ten or more times greater than those of non-pregnant women, but the amounts vary widely from day to day and during the course of a day without any obvious relation to meals (Lind and Hytten, 1972). Normal non-pregnant patterns of excretion are re-established within a week after delivery (Davison and Lovedale, 1974), and the evidence points to an intermittent renal tubular failure to reabsorb glucose, specific to pregnancy. Earlier investigators (Christensen, 1958 ; Welsh and Sims, 1960) concluded that the maximum tubular capacity to reabsorb glucose,
the so-called Tm,, is unaltered during pregnancy, and that the tendency to excrete glucose is due to a raised glomerular filtration rate (GFR), which so increases the filtered load of glucose that tubular capacity is overwhelmed. That view was based on methods and concepts used by the classical workers on renal function (Shannon and Fisher, 1938; Goldring et al., 1940; Smith et al., 1943) and it has become clear that both their results and interpretations were misleading. For example, the estimation of glucose in urine by the traditional methods grossly underestimated the amount of glucose excreted, so that the kidney was credited with a far greater reabsorptive capacity than it had 374
RENAL HANDLING OF GLUCOSE
(Davison and Cheyne, 1972 and 1974), and replotting of the earlier data on which the concept of a ceiling of reabsorption, the Tm,, was based shows no convincing evidence of any such ceiling (Davison and Hytten, 1974). Eggleton and Shuster (1953) and Dempster et al. (1956) were also unable to find evidence of a Tm, in cats and dogs; instead reabsorption continued to rise, though at a diminishing rate, as the filtered load increased. Furthermore, it is now clear that glucose reabsorption is related to the state of sodium and fluid reabsorption by the proximal tubule and therefore to all the physiological circumstances which influence that (Keys and Swanson, 1971; Schultze and Berger, 1972);it also varies with alterations in glomerular filtration rate (GFR), and GFR is itself altered by high levels of glucose in the blood (BrarchnerMortensen, 1973; Kurtzman and Pillay, 1973). Since the development of the only method so far capable of determining glucose accurately in urine, using the enzymes hexokinase and glucose6-phosphate dehydrogenase (Barthelmai and Czok, 1962; Schmidt, 1963; Lind et al., 1971) there seems to have been no detailed study of tubular reabsorption of glucose in “glycosuric” and “non-glycosuric” pregnant women. Renschler and von Baeyer (1966) reported briefly without data, that 15 pregnant women, I1 of whom were “glycosuric”, showed a reduced ability to reabsorb glucose compared to the usual performance of non-pregnant subjects. Since “glycosuria” is an extremely variable and as yet unexplained phenomenon during pregnancy, it seemed logical to re-examine tubular reabsorption by modern techniques in women whose patterns of glucose excretion throughout pregnancy were known. SUBJECTS AND METHODS Twenty-nine normal pregnant women agreed to co-operate in the study after detailed explanation ; only two were specifically recruited because routine urine analysis at the antenatal clinic had revealed heavy glycosuria on several occasions. All were in good general health, had normal oral glucose tolerance tests at 30 weeks of pregnancy, and had no family history of diabetes; repeated mid-stream specimens of urine were sterile on culture in all cases. Each patient was delivered of a normal child whose weight at birth was above
375
the 25th centile for gestation as judged by the sex and parity specific standards of Thomson et al. (1968). From the time they entered the study all of these pregnant women provided 24-hour samples of urine every two to three weeks, including the 24 hours before each infusion; the number of 24-hour samples ranged from 6 to 30 per patient. On the occasion of their non-pregnant (postnatal) infusion they again collected urine for the preceding 24 hours. Each 24-hour urine specimen was stored in a polythene bottle containing 5 ml. of 20 per cent aqueous chlorhexidine gluconate, which has been shown to prevent loss of glucose. The glucose content was determined by the hexokinase technique (as above) which is known to determine accurately urinary glucose as low as 5 mg. per 100 nil. (Lind et al., 1971). In addition, every sample of urine passed during the week before the 24-hour urine collection was tested with “Clinistix” (Ames Company) in order to study the pattern of glycosuria. These reagent strips were of recent manufacture and kept in a tightly-closed screw-top bottle. Measurements of glomerular filtration rate and tubular reabsorption of glucose were made at 15 to 18 weeks, 25 to 28 weeks and 35 to 38 weeks of pregnancy. Thirteen of the subjects were studied on three occasions, two were studied on two occasions and fourteen on one occasion during pregnancy. The measurements were repeated on all 29 subjects 8 to 12 weeks after delivery. Details are given in Table 1. A standard intravenous infusion technique was used (Davison and Cheyne, 1972) based on that described by Goldring and Chasis (1944). A priming infusion of 10 per cent dextrose containing 1 g. of inulin per 100 ml. was given by a constant infusion pump at a rate of 16 ml. per minute for 10 minutes. Thereafter a sustaining infusion of 10 per cent dextrose containing 0 5 g. of inulin was given for 20 minutes at 8 ml. per minute aimed to give plateau plasma levels of 150-250 mg. glucose per 100 ml. and 30 mg. inulin per 100 ml. The actual plateaux achieved are shown in Table I. When stable plasma levels had been achieved after 30 minutes, clearances were estimated for three successive 20 minute periods. A blood sample was taken at the midpoint of each such period. The women passed urine spontaneously.
-
376
DAVISON AND HYTTEN
TABL Detaile,
Infant h
0
‘g c.
0
.g I WI
2
Glucose excretion (mg./24h)
?
9
g’g gg
q
E
c
9
3s a*
.5
M
3
3 .0
3
8
23
g$
9 .-
!n
c.
During pregnancy
d
$3
‘3 b
de) m .?
{
Post-
No. days tested
Mean
Range
92 149 117 98 87 146 95 116 99 134 78 118 116 82
61-124 96197 71-165 67-168 46168 92-366 68-159 98-142 69-146 96174 69-101 94-158 81-173 54-130
Nonglycosuric
JB HC ED
cc
KO MS YY IB KB EY DF MH IM PC
30 29 34 26 29 29 32 22 29 27 29 32 27 27
1 0 4 2 3 3 1 0 3 1 4 0 1 2
164 166 168 159 164 162 160 163 171 170 159 155 156 156
51.3 59.0 62.7 66.3 55.0 62.5 53.7 57.7 64.1 63.4 61.5 71.7 51.5 56.5
3156 4330 3386 3480 3138 3628 3560 3572 3840 3954 2830 2880 3050 2900
39 41 41 37 40 39
22 29 26 30 28 24 28 35 26 26 24 31 20
1 2 0 0 1 1 2 1 0 0 1 0 0
153 156 158 152 155 153 164 164 174 161 162 173 163
55.3 61.0 52.9 47.2 48.9 52.4 52.0 72.5 65.4 57.2 58.6 66.8 62.5
29 26
1 1
162 153
56.7 48.7
F M F M M
25 43 49 52
40
M M
39
M
7 12 10 9 10 8 9 6 6 8 6 8 7 30
3636 3386 3476 2460 3130 3500 3487 3704 3500 2480 3450 3960 3500
40 40 40 38 39 39 40 41 40 40 40 39
M F M F F F M M F F M F F
10 7 7 12 6 6 6 6 10 6 7 8 6
165 302 171 436 397 328 226 299 334 522 225 189 255
75-336 72-665 80-357 63-959 95-889 182-596 159-302 108-492 136-795 302-992 61-390 132-280 130411
34 46 60 37 52 62 38 54 65 62 57 71 32
3660 3100
39 38
F M
8 8
10561 9895
6600-19152 2424-23323
46 100
40 39 39 40 39 39
M M M M F M
THE EFFECT OF PREGNANCY ON THE RENAL HANDLING OF GLUCOSE BY
J. M. DAVISON AND
F. E. HYTTEN MRC Reproduction and Growth Unit Princess Mary Maternity Hospital Great North Road, Newcastle upon Tyne NE2 3BD Summary Tubular reabsorption of glucose has been measured during glucose infusion in 29 healthy women during and after pregnancy. All the women had normal glucose tolerance to an oral load, and normal glucose excretion when not pregnant, but exhibited a wide range of daily glucose excretion in pregnancy. Throughout pregnancy the renal reabsorption of glucose is less effective than in the nonpregnant state and, in general, the greater the amount of glycosuria which develops in pregnancy, the less effective is the reabsorption during infusion. Post partum, women with minor degrees of glycosuria during the preceding pregnancy return to a normal highly efficient reabsorption performance during infusion, but women who exhibit greater degrees of glycosuria have a reduced capacity to reabsorb even though they are no longer glycosuric after the pregnancy. It is concluded that pregnancy imposes some specific change in the glucose reabsorptive capacity of the proximal tubule and that women with more than usual degrees of glycosuria in pregnancy may, in addition, have an element of tubular damage. This is discussed in relation to other renal function changes in pregnancy in an attempt to explain the characteristic intermittency of clinical glycosuria in pregnancy.
HEALTHY pregnant women excrete more than the usual amounts of glucose in urine. The losses may be ten or more times greater than those of non-pregnant women, but the amounts vary widely from day to day and during the course of a day without any obvious relation to meals (Lind and Hytten, 1972). Normal non-pregnant patterns of excretion are re-established within a week after delivery (Davison and Lovedale, 1974), and the evidence points to an intermittent renal tubular failure to reabsorb glucose, specific to pregnancy. Earlier investigators (Christensen, 1958 ; Welsh and Sims, 1960) concluded that the maximum tubular capacity to reabsorb glucose,
the so-called Tm,, is unaltered during pregnancy, and that the tendency to excrete glucose is due to a raised glomerular filtration rate (GFR), which so increases the filtered load of glucose that tubular capacity is overwhelmed. That view was based on methods and concepts used by the classical workers on renal function (Shannon and Fisher, 1938; Goldring et al., 1940; Smith et al., 1943) and it has become clear that both their results and interpretations were misleading. For example, the estimation of glucose in urine by the traditional methods grossly underestimated the amount of glucose excreted, so that the kidney was credited with a far greater reabsorptive capacity than it had 374
RENAL HANDLING OF GLUCOSE
(Davison and Cheyne, 1972 and 1974), and replotting of the earlier data on which the concept of a ceiling of reabsorption, the Tm,, was based shows no convincing evidence of any such ceiling (Davison and Hytten, 1974). Eggleton and Shuster (1953) and Dempster et al. (1956) were also unable to find evidence of a Tm, in cats and dogs; instead reabsorption continued to rise, though at a diminishing rate, as the filtered load increased. Furthermore, it is now clear that glucose reabsorption is related to the state of sodium and fluid reabsorption by the proximal tubule and therefore to all the physiological circumstances which influence that (Keys and Swanson, 1971; Schultze and Berger, 1972);it also varies with alterations in glomerular filtration rate (GFR), and GFR is itself altered by high levels of glucose in the blood (BrarchnerMortensen, 1973; Kurtzman and Pillay, 1973). Since the development of the only method so far capable of determining glucose accurately in urine, using the enzymes hexokinase and glucose6-phosphate dehydrogenase (Barthelmai and Czok, 1962; Schmidt, 1963; Lind et al., 1971) there seems to have been no detailed study of tubular reabsorption of glucose in “glycosuric” and “non-glycosuric” pregnant women. Renschler and von Baeyer (1966) reported briefly without data, that 15 pregnant women, I1 of whom were “glycosuric”, showed a reduced ability to reabsorb glucose compared to the usual performance of non-pregnant subjects. Since “glycosuria” is an extremely variable and as yet unexplained phenomenon during pregnancy, it seemed logical to re-examine tubular reabsorption by modern techniques in women whose patterns of glucose excretion throughout pregnancy were known. SUBJECTS AND METHODS Twenty-nine normal pregnant women agreed to co-operate in the study after detailed explanation ; only two were specifically recruited because routine urine analysis at the antenatal clinic had revealed heavy glycosuria on several occasions. All were in good general health, had normal oral glucose tolerance tests at 30 weeks of pregnancy, and had no family history of diabetes; repeated mid-stream specimens of urine were sterile on culture in all cases. Each patient was delivered of a normal child whose weight at birth was above
375
the 25th centile for gestation as judged by the sex and parity specific standards of Thomson et al. (1968). From the time they entered the study all of these pregnant women provided 24-hour samples of urine every two to three weeks, including the 24 hours before each infusion; the number of 24-hour samples ranged from 6 to 30 per patient. On the occasion of their non-pregnant (postnatal) infusion they again collected urine for the preceding 24 hours. Each 24-hour urine specimen was stored in a polythene bottle containing 5 ml. of 20 per cent aqueous chlorhexidine gluconate, which has been shown to prevent loss of glucose. The glucose content was determined by the hexokinase technique (as above) which is known to determine accurately urinary glucose as low as 5 mg. per 100 nil. (Lind et al., 1971). In addition, every sample of urine passed during the week before the 24-hour urine collection was tested with “Clinistix” (Ames Company) in order to study the pattern of glycosuria. These reagent strips were of recent manufacture and kept in a tightly-closed screw-top bottle. Measurements of glomerular filtration rate and tubular reabsorption of glucose were made at 15 to 18 weeks, 25 to 28 weeks and 35 to 38 weeks of pregnancy. Thirteen of the subjects were studied on three occasions, two were studied on two occasions and fourteen on one occasion during pregnancy. The measurements were repeated on all 29 subjects 8 to 12 weeks after delivery. Details are given in Table 1. A standard intravenous infusion technique was used (Davison and Cheyne, 1972) based on that described by Goldring and Chasis (1944). A priming infusion of 10 per cent dextrose containing 1 g. of inulin per 100 ml. was given by a constant infusion pump at a rate of 16 ml. per minute for 10 minutes. Thereafter a sustaining infusion of 10 per cent dextrose containing 0 5 g. of inulin was given for 20 minutes at 8 ml. per minute aimed to give plateau plasma levels of 150-250 mg. glucose per 100 ml. and 30 mg. inulin per 100 ml. The actual plateaux achieved are shown in Table I. When stable plasma levels had been achieved after 30 minutes, clearances were estimated for three successive 20 minute periods. A blood sample was taken at the midpoint of each such period. The women passed urine spontaneously.
-
376
DAVISON AND HYTTEN
TABL Detaile,
Infant h
0
‘g c.
0
.g I WI
2
Glucose excretion (mg./24h)
?
9
g’g gg
q
E
c
9
3s a*
.5
M
3
3 .0
3
8
23
g$
9 .-
!n
c.
During pregnancy
d
$3
‘3 b
de) m .?
{
Post-
No. days tested
Mean
Range
92 149 117 98 87 146 95 116 99 134 78 118 116 82
61-124 96197 71-165 67-168 46168 92-366 68-159 98-142 69-146 96174 69-101 94-158 81-173 54-130
Nonglycosuric
JB HC ED
cc
KO MS YY IB KB EY DF MH IM PC
30 29 34 26 29 29 32 22 29 27 29 32 27 27
1 0 4 2 3 3 1 0 3 1 4 0 1 2
164 166 168 159 164 162 160 163 171 170 159 155 156 156
51.3 59.0 62.7 66.3 55.0 62.5 53.7 57.7 64.1 63.4 61.5 71.7 51.5 56.5
3156 4330 3386 3480 3138 3628 3560 3572 3840 3954 2830 2880 3050 2900
39 41 41 37 40 39
22 29 26 30 28 24 28 35 26 26 24 31 20
1 2 0 0 1 1 2 1 0 0 1 0 0
153 156 158 152 155 153 164 164 174 161 162 173 163
55.3 61.0 52.9 47.2 48.9 52.4 52.0 72.5 65.4 57.2 58.6 66.8 62.5
29 26
1 1
162 153
56.7 48.7
F M F M M
25 43 49 52
40
M M
39
M
7 12 10 9 10 8 9 6 6 8 6 8 7 30
3636 3386 3476 2460 3130 3500 3487 3704 3500 2480 3450 3960 3500
40 40 40 38 39 39 40 41 40 40 40 39
M F M F F F M M F F M F F
10 7 7 12 6 6 6 6 10 6 7 8 6
165 302 171 436 397 328 226 299 334 522 225 189 255
75-336 72-665 80-357 63-959 95-889 182-596 159-302 108-492 136-795 302-992 61-390 132-280 130411
34 46 60 37 52 62 38 54 65 62 57 71 32
3660 3100
39 38
F M
8 8
10561 9895
6600-19152 2424-23323
46 100
40 39 39 40 39 39
M M M M F M
