@Copyright 1990 by the Humana Press, Inc. All rights of any nature, whatsoever, reserved. 0163-4984/90/2401-0073 $02.00
Interrelationships of Zinc with Glucose and Insulin Metabolism in Humans J. BRAND,~O-NETO, *'1 J. G. H. VIEIP~,2 T. SHUHAMA,3 E. M. K. Russo) R. V. PIESCO, 1 AND P. R. CURP lEndocrine-/Vletabolic Units, Universidade Estadual Paulista, BR- 1861 O, Botucatu, SP, Brasil; 2Escola Paulista de/Vledicina, BR-04023 S~o Paulo, SP, Brasil; and 3Faculdade de Ci6ncias Farmac6uticas de Ribeir~o Preto da Universidade de S~o Paulo, BR-14049 Ribeir~o Preto, SP, Brazil
ABSTRACT Hyperzincemia has been reported to cause alterations in the homeostasis of glycid metabolism. To determine this effect on plasma glucose and insulin levels, we studied 36 normal individuals of both sexes aged 22-26 y after a 12-h fast. The tests were initiated at 7:00 AM when an antecubital vein was punctured and a device for infusion was installed and maintained with physiological saline. Zinc was administered orally at 8:00 AM. Subjects were divided into an experimental group of 22 individuals who received doses of 25, 37.5, and 50 mg of zinc and a control group of 14 individuals. Blood samples were collected over a period of 240 min after the basal samples ( - 30 and 0 rain). We did not detect any change in plasma glucose or insulin levels, a fact that we attribute either to the ineffectiveness of the 50 mg dose of zinc or to the lack of human response to the acute action of this trace element. The individuals who ingested zinc showed a significant fall in plasma cortisol, probably caused by the action of this trace element. Index Entries: Oral zinc tolerance test; acute hyperzincemia; zinc supplement, effects of on glycid metabolism; zinc supplement, effects of on glucose, insulin and cortisol; serum zinc and plasma glucose, insulin and cortisol levels; normal individuals. *Author to whom all correspondence and reprint requests should be addressed.
Biological Trace Element Research
73
VoL 24, 1990
74
Brand&o-Neto et al.
INTRODUCTION After Scott (1) discovered the role of zinc insulin crystallization in 1934, other investigators confirmed the structural and functional relationship between this trace element and insulin. Zinc has been found in high concentrations in the endocrine pancreas of many species (2) as well as in the insulin molecule (2,3) and in secretory granules (4,5), and has been shown to participate in the stabilization of proinsulin and insulin hexamers by forming complexes with them (6,7). Selective zinc deprivation in rats has demonstrated that this trace element also affects insulin synthesis and secretion since it can increase glucose tolerance (8,9), reduce beta cell degranulation (10), decrease histochemically detectable insulin (10), decrease serum insulin (11,12), and increase insulin degradation by the liver (13), as well as reduce tissue sensitivity to the action of insulin (9) and induce Langerhans islet deficiency in the offspring of zinc-deprived pregnat rats (14). In contrast, zinc supplementation causes increased blood (15-17) and urine (15,18) glucose, decreses serum insulin and increases plasma glucagon (19), in addition to enhancing the effect of insulin in certain tissues (20,21). On the basis of these reports, which have established the structural and functional relationship between zinc and insulin, the present investigation was undertaken to study the effect of zinc on glycid metabolism in humans.
MATERIAL AND METHODS The study was conducted on 36 healthy adults, i.e., male and female medical students aged 22-26 y, taking no medication, of normal weight, and with no history of endocrine disease. The study was approved by the Medical Ethics Committee of this institution. The tests were initiated at 7:00 AM after a 12-h fast, and the subjects maintained resting decubitus throughout the experimental period. The study consisted of two stages. Stage 1 involved 3 groups. Group 1 (control): 8 individuals (4 males and 4 females) ingested 20 mL of physiologycal saline. Group 2:10 individuals (5 males and 5 females) ingested 25 mg of zinc. Group 3:3 females ingested 37.5 mg of zinc, and 3 males ingested 50 mg. Blood samples were collected at - 30, 0, 30, 60, 90, 120, 150, 180, 210, and 240 rain. Stage 2 involved only 2 groups. Group 1: 6 individuals (3 males and 3 females) ingested 20 mL of physiological saline. Group 2:6 individuals (3 males and 3 females) ingested 50 mg of zinc, and serial blood collections were performed at -30, 0, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, and 120 min. The objective of this serial
Biological Trace Element Research
Vol. 24, 1990
Effects of Zinc on Giycid Metabolism
75
collection at 10-min rather than 30-min intervals was to guarantee that no possible hormonal variation within our time scale would go undetected
(19). Zinc was administered orally in the form of hepta-hydrated zinc sulfate diluted in 20 mL deionized water. Blood samples were obtained by puncturing a forearm vein without using a tourniquet and collected into plastic syringes. All material used for collection, separation, and storage was propylene plastic, carrefully washed according to laboratory standards and rinsed with cetone, 3 N HCI, 0.1% EDTA and distilled and deionized water. Blood samples were frozen and stored at -20~ until the time for measurement. Serum zinc levels were determined in duplicate using an atomic absorption spectrofotometer (Atomspek, Hilger & Watts, model H-1170, UK) as recommended by the manufacturer. The intra-assay error was 2.8% and sensitivity was 0.02 lag/mL. Samples showing hemolysis were discarded. Plasma glucose levels were measured by the ortho-toluidine method (22). Plasma insulin was measured by the double antibody method (23) with an intra- and interassay error of 4.1 and 8.7%, respectively, and a sensitivity of 1.8 IJ,IU/mL. Plasma cortisol was measured using kits from Diagnostic Products Corporation, US. The intra- and interassay errors were 6.6 and 8.0%, respectively, and sensitivity was 1.0 i~g/dL. The data for each variable in the three experimental groups evaluated at 5 moments (0, 60, 120, 180, and 240 rain) were compared by Profile Analysis (24), with interaction tests between groups and moments (to determine the analogy of the mean profiles for the three groups), differences among group within the set of moments (identity of the profiles), and for each time point separately, and differences among time points within the set of groups and within each group separately. The calculated F parameters were considered significant when p < 0.05, where p = level of significance. When 0.05 < p < 0.10, a tendecy was considered to have occurred. The analysis was complemented with the study of linear correlation between pairs of variables within each group
(24).
RESULTS No alteration in glycid metabolism was detected in the present study since no variation was found in plasma glucose or insulin levels. During stage 1 of the investigation, serum zinc levels were constant in Group 1, and more elevated in Group 3 than in Group 2, who received a lower dose. The highest peak occurred between M3 and M4 (Fig. 1, Table 1). The glucose profiles were identical for the three groups (Fig. 2, Table 2). The same occurred for insulin, whose profiles were identical for all
Biological Trace Element Research
Vol. 24, 1990
76
Brand&o-Neto et aL Q GROUP I
4-
9
/.. A
.J
/'
E3-
.~"
/ Pa-
I$
-'K,,,,
...---
/"
"~ -,,, \
:/~ //
E
9 e~uP 3
\.. i,,~- .....
. '/1
N
m era
~..
GROUP 2
NN
"11
//
.'/
MI
M2
M3 MomeMs
M4
M5
Fig. 1. M e a n zinc profiles for each g r o u p at the 5 experimental moments.
Table 1 Zinc Profile Analysis" Hypothesis 1. Parallelism of the treatment m e a n profiles 4. Equal treatm e n t levels within m o ments 5. Equal treatment means within g r o u p
Statistical p a r a m e t e r F = 15.86
Comment The profiles w e r e not similar
P < 0.01
MI: M2: M3: M4: M5:
F F F F F
= 13.76; P < = 18.81; P < = 50.06; P < = 50.68; P < = 46.16; P < GI: F = 0.04 P > 0.05
G2: F = 27.93 P < 0.01 G3: F = 30.85 P < 0.01
0.01 0.01 0.01 0.01 0.01
G I < (G2 = G3) G I < (G2 = G3) G1 < (G2 = G3) G1 ~ G2 < G3 G1 < G2 < G3 N o significant difference between moments was detected within G1 M1 < (M2 = M5) < (M3 = M4) M1 < (M2 = M5) < (M3 = M4)
9Hypothesis tested, statistical parameters calculated and comments.
BiologicalTrace Element Research
Vol. 24, 1990
77
Effects o f Zinc o n Glycid M e t a b o l i s m I00-
9 GROUP I 9 GROUP 2 9
GROUP:5
12o I
o
80
.
.
.
.
.
.
.
.
.
O
E
W
O co
Q
~r 40,
0
I
MI
I
M2
t
M3
I
M4
M5
Moments
Fig. 2. ments,
Mean glucose profiles for each group at the 5 experimental mo-
Table 2 Glucose Profile Analysis" Statistical parameter
Hypothesis 1. Parallelism of the treatment mean profiles
F = 0.52 P ~ 0.50
2. Equal treatment levels
F P F P
3. Equal response means
~
--> = >
0.39 0.50 0.29 0.50
Comment No significant interaction was detected between groups and moments (similar profiles) Equal profiles No significant difference was detected between moments
tested, statistical parameters calculated, and comments.
Biological TraceElement Research
Vol.24, 1990
Brand, o-Nero et al.
78 55-
9
GROUP I
9
GROUP2
9
GROUP:5
-I
E 26.25-
C
~
r ~
C
o
17.5
E o
9
M
!
r
~..~k--......__ . . . . . . .
C 0 @
.........~k.-
A
M
A
8.75
0
I
Mj
I
Mz
M3
I
!
M4
M5
Moments
Fig. 3. Mean insulin profiles for each g r o u p at the 5 experimental m o m e n t s .
Table 3 Insulin Profile Analysis ~ Hypothesis 1. Parallelism of the treatment m e a n profiles
Statistical p a r a m e t e r F = 0.62 P > 0.50
2. Equal treatment levels
F = 2.57 0.05 < P < 0.10
3. Equal response m e a n s
F = 0.32 P > 0.50
Comment No significant interaction was detected b e t w e e n groups and m o m e n t s (similar profiles) There was a t e n d e n c y toward a difference G 2 / > (G1 = G3) No significant difference was detected b e t w e e n moments
~Hypothesis tested, statistical parameters calculated, and comments. Biological Trace Element Research
Vol. 24, 1990
Effects of Zinc on Glycid Metabolism
79
20-
9 GROUP I I
GROUP 2
9 GROUP 5 15_1
~olO-
E
5: 5-
0
I
M
1
M2
i
M:3
M4
i
M5
Moments
Fig. 4. M e a n c o r t i s o l p r o f i l e s for e a c h at t h e 5 experimental moments.
Table 4 C o r t i s o l Profile A n a l y s i s " Hypothesis 1. P a r a l l e l i s m of t h e treatment mean profiles 4. E q u a l t r e a t ment levels within moments 5. E q u a l t r e a t ment means within group
Statistical p a r a m e t e r
Comment
F = 2.52 P < 0.05
There was a significant interaction between groups and moments. The profiles were not similar. G1 = G 2 = G 3 G1 = G2 = G3 G1 = G2 = G 3 G1 = G 2 = G 3 G1 = G2 = G3 A tendency toward a more e l e v a t e d m e a n at M1 t h a n at M2 (M1 = M2) > (M3 = M 4 = M5) MI> (M2 = M 3 = M 4 = M5)
MI: M2: M3: M4: M5:
F = 1.61; P > 0.10 F = 2.10; P > 0.10 F = 0.52; P > 0.50 F = 2.12; P > 0.10 F = 2.11; P > 0.10 G I : F = 4.01 0.05 < P < 0.10 G2: F = 12.63 P < 0.05 G3: F = 8.39 P < 0.05
"Hypothesis tested, statistical parameters calculated and comments. Biological Trace Element Research
Vol. 24,1990
80
Brand~o-Neto et aL
experimental groups throughout the 240 min of the study (Fig. 3, Table 3). In contrast to glucose and insulin, plasma cortisol levels fell in all experimental groups, and more markedly so in Groups 2 and 3, with different profiles for each group and with significant differences between moments (Fig. 4, Table 4). This fall should not be attributed only to the circadian rhythm, since correlation analysis revealed the existence of a strong negative correlation between cortisol and zinc in Group 2 (r = -0.57, p < 0.001) and Group 3 (r -- -0.83, p < 0.001), showing that the fall in cortisol levels was more marked with increasing zinc doses. No type of correlation was detected among glucose, insulin, and zinc in the three experimental groups. These results were confirmed by the same analyses in the second stage of the investigation. The reduction of the scale to 10 min showed that there was no variation in plasma glucose or insulin.
DISCUSSION In the present investigation, we used equal proportions of males and female subjects. In Group 3, characterized as the group receiving the highest zinc dose, we studied the 6 individuals as a whole from a statistical viewpoint. Thus, the present results demonstrate that acute hyperzincemia did not alter peripheral glucose or insulin levels. These data agree with those reported by B6gin-Heick et al. (25), who found no peripheral changes in glucose in mice submitted to zinc supplementation. However, opposite results of hyperglycemia and glycosuria have been reported in laboratory animals (15-18,26). Decreased insulin secretion during the GTT has also been reported by Ghafghazi et al. (27), and this inhibitory effect of zinc was later explained by interference of this trace element with intracellular calcium in studies using the isolated or perfused pancreas technique and isolated Langerhans islets (28). Etzel and Cousins (29) reported an increase in glucose, a fall in insulin and an elevation of glucagon after ip zinc administration to rats, especially during the first 30 min. Adrenalectomy practically abolished this hyperglycemic response. According to these investigators, and also to Horak and Suderman (30), the hyperglycemic response to zinc depended on the effects of glucagon and cortisol on hepatic glycogenolysis. In contrast, in the present study, we did not observe an increase in plasma cortisol, but rather a fall, which was actually more marked and statistically significant in the group that ingested zinc. Even though we used increasing zinc doses, it is possible that the maximum dose of 50 mg was not sufficient to induce alterations in glucose or insulin. In this respect, there is evidence that zinc is taken up slowly by beta cells (31,32), so that the pancreas needs a longer time of exposure in relation to extracellular zinc, as well as supraphysiological doses (15,16,29,33). Biological Trace Element Research
Vol. 24, 1990
Effects o f Zinc on Glycid /Vletabolism
81
Thus, we conclude that acute zinc administration did not modify plasma glucose or insulin levels either because the 50 mg dose was insufficient or because the h u m a n species, in contrast to others, is refractory to the acute action of this trace element. As to cortisol, zinc may have an inhibitory effect on its synthesis and secretion. Further studies are n e e d e d to elucidate this p h e n o m e n o n .
To determine the acute effect of hyperzincemia on the homeostasis of glycid metabolism in humans, we applied the oral zinc tolerance test to 22 normal individuals. No peripheral change was observed, even though a significant fall in plasma cortisol occurred. These results may be explained either by the ineffectiveness of the 50 mg zinc dose used or by a lack of response by the h u m a n species to the acute action of zinc. The fall in cortisol may have been owing to the inhibitory effect of this trace element.
ACKNOWLEDGMENTS This work was supported by FAPEfiP (86/1889-2; 86/2511-3), CNPq (407220/84), and FUNDUNESP (4789/85). We thank Maria Nice de S. Tavares for technical assistance.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.
D. A. Scott, Biochem. J. 28, 1602 (1934). N. Havu, G. Lundgren, and S. Falkmer, Acta Endocrinol. 86, 570 (1977). D. C. Hodgkin, Diabetes 21, 1131 (1972). E. Pihl, J. Microscopie 7, 509 (1968). S. O. Emdin, G. G. Dodson, J. M. Cutfield, and S. M. Cutfield, Diabetologia 19, 174 (1980). T. L. Blundell, J. F. Cutfield, S. M. Cutfield, E. J. Dodson, G. G. Dodson, D. C. Hodgkin, and D. A. Mercola, Diabetes Suppl. 2, 492 (1972). B. J. Frank, A. H. Pekar, and A. J. Veros, Diabetes Suppl. 2, 486 (1972). E. Hove, C. A. Elvehjem, and E. B. Hart, Am. J. Physiol. 119, 768 (1937). J. Quarterrnan, C. F. Mills, and W. R. Humphries, Biochem. Biophys. Res. Comm. 25, 354 (1966). L. Boquist and A. Lernmark, Acta Path. Microbiol. Scand. 76, 215 (1969). A. M. Huber and S. N. Gershoff, J. Nutr. 103, 1739 (1973). H. P. Roth and M. Kirchgessner, Internal. Z. Vit. Ern. Forsch. 45, 201 (1975). D. G. Hendricks and A. W. Mahoney, J. Nutr. 102, 1079 (1972). L. K. Robinson and L. S. Hurley, ]. Nutr. 111, 869 (1981). W. Salant and L. E. Wise, J. Biol. Chem. 34, 447 (1918). F. Y. Berenshtein and M. I. Shkol-Nik, Fiziol. Zhur. S.S.S.R. 37, 120 (1951). N. Hartmann, H. Fiedler, M. Ahrens, and M. Hansch, Wiss. Z. Karl-MarxUniv. Leipzig, Math-Naturwiss. R.21, 531 (1972).
Biological Trace Element Research
Vol. 24, 1990
82
B r a n d 2 o - b l e t o et aL
18. G. Weitzel, F. J. Strecker, and U. Roester, Ztschr. Physiol. Chem. 292, 286 (1953). 19. K. R. Etzel and R. J. Cousins, J. Nutr. 113, 1657 (1983). 20. E. R. Arquilla, S. Packer, W. Tarmas, and S. Myamoto, Endocrinoloy,y 103, 1440 (1978). 21. L. Coulston and P. Dandona, Diabetes 29, 665 (1980). 22. K. M. Dubowski, Clin. Chem. 8, 215 (1962). 23. J. G. H. Vieira, E. M. K. Russo, R. M. B. Maciel, O. A. Germek, and A. R. Chacra, Rev. Bras. Patol. Clin. 16, 108 (1980). 24. D. F. Morrison, Multivariate Statistical Methods, 1st ed., McGraw-Hill, New York, 1967, pp. 338. 25. N. B6gin-Heick, M. Dalp6-Scott, J. Rowe, and H. M. C. Heick, Diabetes 34, 179 (1985). 26. W. R. Sutton and V. E. Nelson, Proc. Iowa Acad. Sci. 45, 115 (1938). 27. T. Ghafghazi, C. W. Ludvigsen, M. L. McDaniel, and P. E. Lacy, IRCS Med. Sci. 7, 122 (1979). 28. T. Ghafghazi, M. L. McDaniel, and P. E. Lacy, Diabetes 30, 341 (1981). 29. K. R. Etzel and R. J. Cousins, J. Nutr. 113, 1657 (1983). 30. E. Horak and F. W. Suderman Jr., Toxicol. Appl. Pharmacol. 32, 316 (1975). 31. D. P. Figlewicz, B. Formby, A. T. Hodgson, F. G. Schmid, and G. M. Grodsky, Diabetes 29, 767 (1980). 32. C. Ludvigsen, M. McDaniel, and P. E. Lacy, Diabetes 28, 570 (1981). 33. D. P. Figlewicz, A. Heldt, S. E. Forhan, and G. M. Grodsky, Endocrinology 108, 730 (1981).
Biological TraceElement Research
Vof.24, 1990
Interrelationships of Zinc with Glucose and Insulin Metabolism in Humans J. BRAND,~O-NETO, *'1 J. G. H. VIEIP~,2 T. SHUHAMA,3 E. M. K. Russo) R. V. PIESCO, 1 AND P. R. CURP lEndocrine-/Vletabolic Units, Universidade Estadual Paulista, BR- 1861 O, Botucatu, SP, Brasil; 2Escola Paulista de/Vledicina, BR-04023 S~o Paulo, SP, Brasil; and 3Faculdade de Ci6ncias Farmac6uticas de Ribeir~o Preto da Universidade de S~o Paulo, BR-14049 Ribeir~o Preto, SP, Brazil
ABSTRACT Hyperzincemia has been reported to cause alterations in the homeostasis of glycid metabolism. To determine this effect on plasma glucose and insulin levels, we studied 36 normal individuals of both sexes aged 22-26 y after a 12-h fast. The tests were initiated at 7:00 AM when an antecubital vein was punctured and a device for infusion was installed and maintained with physiological saline. Zinc was administered orally at 8:00 AM. Subjects were divided into an experimental group of 22 individuals who received doses of 25, 37.5, and 50 mg of zinc and a control group of 14 individuals. Blood samples were collected over a period of 240 min after the basal samples ( - 30 and 0 rain). We did not detect any change in plasma glucose or insulin levels, a fact that we attribute either to the ineffectiveness of the 50 mg dose of zinc or to the lack of human response to the acute action of this trace element. The individuals who ingested zinc showed a significant fall in plasma cortisol, probably caused by the action of this trace element. Index Entries: Oral zinc tolerance test; acute hyperzincemia; zinc supplement, effects of on glycid metabolism; zinc supplement, effects of on glucose, insulin and cortisol; serum zinc and plasma glucose, insulin and cortisol levels; normal individuals. *Author to whom all correspondence and reprint requests should be addressed.
Biological Trace Element Research
73
VoL 24, 1990
74
Brand&o-Neto et al.
INTRODUCTION After Scott (1) discovered the role of zinc insulin crystallization in 1934, other investigators confirmed the structural and functional relationship between this trace element and insulin. Zinc has been found in high concentrations in the endocrine pancreas of many species (2) as well as in the insulin molecule (2,3) and in secretory granules (4,5), and has been shown to participate in the stabilization of proinsulin and insulin hexamers by forming complexes with them (6,7). Selective zinc deprivation in rats has demonstrated that this trace element also affects insulin synthesis and secretion since it can increase glucose tolerance (8,9), reduce beta cell degranulation (10), decrease histochemically detectable insulin (10), decrease serum insulin (11,12), and increase insulin degradation by the liver (13), as well as reduce tissue sensitivity to the action of insulin (9) and induce Langerhans islet deficiency in the offspring of zinc-deprived pregnat rats (14). In contrast, zinc supplementation causes increased blood (15-17) and urine (15,18) glucose, decreses serum insulin and increases plasma glucagon (19), in addition to enhancing the effect of insulin in certain tissues (20,21). On the basis of these reports, which have established the structural and functional relationship between zinc and insulin, the present investigation was undertaken to study the effect of zinc on glycid metabolism in humans.
MATERIAL AND METHODS The study was conducted on 36 healthy adults, i.e., male and female medical students aged 22-26 y, taking no medication, of normal weight, and with no history of endocrine disease. The study was approved by the Medical Ethics Committee of this institution. The tests were initiated at 7:00 AM after a 12-h fast, and the subjects maintained resting decubitus throughout the experimental period. The study consisted of two stages. Stage 1 involved 3 groups. Group 1 (control): 8 individuals (4 males and 4 females) ingested 20 mL of physiologycal saline. Group 2:10 individuals (5 males and 5 females) ingested 25 mg of zinc. Group 3:3 females ingested 37.5 mg of zinc, and 3 males ingested 50 mg. Blood samples were collected at - 30, 0, 30, 60, 90, 120, 150, 180, 210, and 240 rain. Stage 2 involved only 2 groups. Group 1: 6 individuals (3 males and 3 females) ingested 20 mL of physiological saline. Group 2:6 individuals (3 males and 3 females) ingested 50 mg of zinc, and serial blood collections were performed at -30, 0, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, and 120 min. The objective of this serial
Biological Trace Element Research
Vol. 24, 1990
Effects of Zinc on Giycid Metabolism
75
collection at 10-min rather than 30-min intervals was to guarantee that no possible hormonal variation within our time scale would go undetected
(19). Zinc was administered orally in the form of hepta-hydrated zinc sulfate diluted in 20 mL deionized water. Blood samples were obtained by puncturing a forearm vein without using a tourniquet and collected into plastic syringes. All material used for collection, separation, and storage was propylene plastic, carrefully washed according to laboratory standards and rinsed with cetone, 3 N HCI, 0.1% EDTA and distilled and deionized water. Blood samples were frozen and stored at -20~ until the time for measurement. Serum zinc levels were determined in duplicate using an atomic absorption spectrofotometer (Atomspek, Hilger & Watts, model H-1170, UK) as recommended by the manufacturer. The intra-assay error was 2.8% and sensitivity was 0.02 lag/mL. Samples showing hemolysis were discarded. Plasma glucose levels were measured by the ortho-toluidine method (22). Plasma insulin was measured by the double antibody method (23) with an intra- and interassay error of 4.1 and 8.7%, respectively, and a sensitivity of 1.8 IJ,IU/mL. Plasma cortisol was measured using kits from Diagnostic Products Corporation, US. The intra- and interassay errors were 6.6 and 8.0%, respectively, and sensitivity was 1.0 i~g/dL. The data for each variable in the three experimental groups evaluated at 5 moments (0, 60, 120, 180, and 240 rain) were compared by Profile Analysis (24), with interaction tests between groups and moments (to determine the analogy of the mean profiles for the three groups), differences among group within the set of moments (identity of the profiles), and for each time point separately, and differences among time points within the set of groups and within each group separately. The calculated F parameters were considered significant when p < 0.05, where p = level of significance. When 0.05 < p < 0.10, a tendecy was considered to have occurred. The analysis was complemented with the study of linear correlation between pairs of variables within each group
(24).
RESULTS No alteration in glycid metabolism was detected in the present study since no variation was found in plasma glucose or insulin levels. During stage 1 of the investigation, serum zinc levels were constant in Group 1, and more elevated in Group 3 than in Group 2, who received a lower dose. The highest peak occurred between M3 and M4 (Fig. 1, Table 1). The glucose profiles were identical for the three groups (Fig. 2, Table 2). The same occurred for insulin, whose profiles were identical for all
Biological Trace Element Research
Vol. 24, 1990
76
Brand&o-Neto et aL Q GROUP I
4-
9
/.. A
.J
/'
E3-
.~"
/ Pa-
I$
-'K,,,,
...---
/"
"~ -,,, \
:/~ //
E
9 e~uP 3
\.. i,,~- .....
. '/1
N
m era
~..
GROUP 2
NN
"11
//
.'/
MI
M2
M3 MomeMs
M4
M5
Fig. 1. M e a n zinc profiles for each g r o u p at the 5 experimental moments.
Table 1 Zinc Profile Analysis" Hypothesis 1. Parallelism of the treatment m e a n profiles 4. Equal treatm e n t levels within m o ments 5. Equal treatment means within g r o u p
Statistical p a r a m e t e r F = 15.86
Comment The profiles w e r e not similar
P < 0.01
MI: M2: M3: M4: M5:
F F F F F
= 13.76; P < = 18.81; P < = 50.06; P < = 50.68; P < = 46.16; P < GI: F = 0.04 P > 0.05
G2: F = 27.93 P < 0.01 G3: F = 30.85 P < 0.01
0.01 0.01 0.01 0.01 0.01
G I < (G2 = G3) G I < (G2 = G3) G1 < (G2 = G3) G1 ~ G2 < G3 G1 < G2 < G3 N o significant difference between moments was detected within G1 M1 < (M2 = M5) < (M3 = M4) M1 < (M2 = M5) < (M3 = M4)
9Hypothesis tested, statistical parameters calculated and comments.
BiologicalTrace Element Research
Vol. 24, 1990
77
Effects o f Zinc o n Glycid M e t a b o l i s m I00-
9 GROUP I 9 GROUP 2 9
GROUP:5
12o I
o
80
.
.
.
.
.
.
.
.
.
O
E
W
O co
Q
~r 40,
0
I
MI
I
M2
t
M3
I
M4
M5
Moments
Fig. 2. ments,
Mean glucose profiles for each group at the 5 experimental mo-
Table 2 Glucose Profile Analysis" Statistical parameter
Hypothesis 1. Parallelism of the treatment mean profiles
F = 0.52 P ~ 0.50
2. Equal treatment levels
F P F P
3. Equal response means
~
--> = >
0.39 0.50 0.29 0.50
Comment No significant interaction was detected between groups and moments (similar profiles) Equal profiles No significant difference was detected between moments
tested, statistical parameters calculated, and comments.
Biological TraceElement Research
Vol.24, 1990
Brand, o-Nero et al.
78 55-
9
GROUP I
9
GROUP2
9
GROUP:5
-I
E 26.25-
C
~
r ~
C
o
17.5
E o
9
M
!
r
~..~k--......__ . . . . . . .
C 0 @
.........~k.-
A
M
A
8.75
0
I
Mj
I
Mz
M3
I
!
M4
M5
Moments
Fig. 3. Mean insulin profiles for each g r o u p at the 5 experimental m o m e n t s .
Table 3 Insulin Profile Analysis ~ Hypothesis 1. Parallelism of the treatment m e a n profiles
Statistical p a r a m e t e r F = 0.62 P > 0.50
2. Equal treatment levels
F = 2.57 0.05 < P < 0.10
3. Equal response m e a n s
F = 0.32 P > 0.50
Comment No significant interaction was detected b e t w e e n groups and m o m e n t s (similar profiles) There was a t e n d e n c y toward a difference G 2 / > (G1 = G3) No significant difference was detected b e t w e e n moments
~Hypothesis tested, statistical parameters calculated, and comments. Biological Trace Element Research
Vol. 24, 1990
Effects of Zinc on Glycid Metabolism
79
20-
9 GROUP I I
GROUP 2
9 GROUP 5 15_1
~olO-
E
5: 5-
0
I
M
1
M2
i
M:3
M4
i
M5
Moments
Fig. 4. M e a n c o r t i s o l p r o f i l e s for e a c h at t h e 5 experimental moments.
Table 4 C o r t i s o l Profile A n a l y s i s " Hypothesis 1. P a r a l l e l i s m of t h e treatment mean profiles 4. E q u a l t r e a t ment levels within moments 5. E q u a l t r e a t ment means within group
Statistical p a r a m e t e r
Comment
F = 2.52 P < 0.05
There was a significant interaction between groups and moments. The profiles were not similar. G1 = G 2 = G 3 G1 = G2 = G3 G1 = G2 = G 3 G1 = G 2 = G 3 G1 = G2 = G3 A tendency toward a more e l e v a t e d m e a n at M1 t h a n at M2 (M1 = M2) > (M3 = M 4 = M5) MI> (M2 = M 3 = M 4 = M5)
MI: M2: M3: M4: M5:
F = 1.61; P > 0.10 F = 2.10; P > 0.10 F = 0.52; P > 0.50 F = 2.12; P > 0.10 F = 2.11; P > 0.10 G I : F = 4.01 0.05 < P < 0.10 G2: F = 12.63 P < 0.05 G3: F = 8.39 P < 0.05
"Hypothesis tested, statistical parameters calculated and comments. Biological Trace Element Research
Vol. 24,1990
80
Brand~o-Neto et aL
experimental groups throughout the 240 min of the study (Fig. 3, Table 3). In contrast to glucose and insulin, plasma cortisol levels fell in all experimental groups, and more markedly so in Groups 2 and 3, with different profiles for each group and with significant differences between moments (Fig. 4, Table 4). This fall should not be attributed only to the circadian rhythm, since correlation analysis revealed the existence of a strong negative correlation between cortisol and zinc in Group 2 (r = -0.57, p < 0.001) and Group 3 (r -- -0.83, p < 0.001), showing that the fall in cortisol levels was more marked with increasing zinc doses. No type of correlation was detected among glucose, insulin, and zinc in the three experimental groups. These results were confirmed by the same analyses in the second stage of the investigation. The reduction of the scale to 10 min showed that there was no variation in plasma glucose or insulin.
DISCUSSION In the present investigation, we used equal proportions of males and female subjects. In Group 3, characterized as the group receiving the highest zinc dose, we studied the 6 individuals as a whole from a statistical viewpoint. Thus, the present results demonstrate that acute hyperzincemia did not alter peripheral glucose or insulin levels. These data agree with those reported by B6gin-Heick et al. (25), who found no peripheral changes in glucose in mice submitted to zinc supplementation. However, opposite results of hyperglycemia and glycosuria have been reported in laboratory animals (15-18,26). Decreased insulin secretion during the GTT has also been reported by Ghafghazi et al. (27), and this inhibitory effect of zinc was later explained by interference of this trace element with intracellular calcium in studies using the isolated or perfused pancreas technique and isolated Langerhans islets (28). Etzel and Cousins (29) reported an increase in glucose, a fall in insulin and an elevation of glucagon after ip zinc administration to rats, especially during the first 30 min. Adrenalectomy practically abolished this hyperglycemic response. According to these investigators, and also to Horak and Suderman (30), the hyperglycemic response to zinc depended on the effects of glucagon and cortisol on hepatic glycogenolysis. In contrast, in the present study, we did not observe an increase in plasma cortisol, but rather a fall, which was actually more marked and statistically significant in the group that ingested zinc. Even though we used increasing zinc doses, it is possible that the maximum dose of 50 mg was not sufficient to induce alterations in glucose or insulin. In this respect, there is evidence that zinc is taken up slowly by beta cells (31,32), so that the pancreas needs a longer time of exposure in relation to extracellular zinc, as well as supraphysiological doses (15,16,29,33). Biological Trace Element Research
Vol. 24, 1990
Effects o f Zinc on Glycid /Vletabolism
81
Thus, we conclude that acute zinc administration did not modify plasma glucose or insulin levels either because the 50 mg dose was insufficient or because the h u m a n species, in contrast to others, is refractory to the acute action of this trace element. As to cortisol, zinc may have an inhibitory effect on its synthesis and secretion. Further studies are n e e d e d to elucidate this p h e n o m e n o n .
To determine the acute effect of hyperzincemia on the homeostasis of glycid metabolism in humans, we applied the oral zinc tolerance test to 22 normal individuals. No peripheral change was observed, even though a significant fall in plasma cortisol occurred. These results may be explained either by the ineffectiveness of the 50 mg zinc dose used or by a lack of response by the h u m a n species to the acute action of zinc. The fall in cortisol may have been owing to the inhibitory effect of this trace element.
ACKNOWLEDGMENTS This work was supported by FAPEfiP (86/1889-2; 86/2511-3), CNPq (407220/84), and FUNDUNESP (4789/85). We thank Maria Nice de S. Tavares for technical assistance.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.
D. A. Scott, Biochem. J. 28, 1602 (1934). N. Havu, G. Lundgren, and S. Falkmer, Acta Endocrinol. 86, 570 (1977). D. C. Hodgkin, Diabetes 21, 1131 (1972). E. Pihl, J. Microscopie 7, 509 (1968). S. O. Emdin, G. G. Dodson, J. M. Cutfield, and S. M. Cutfield, Diabetologia 19, 174 (1980). T. L. Blundell, J. F. Cutfield, S. M. Cutfield, E. J. Dodson, G. G. Dodson, D. C. Hodgkin, and D. A. Mercola, Diabetes Suppl. 2, 492 (1972). B. J. Frank, A. H. Pekar, and A. J. Veros, Diabetes Suppl. 2, 486 (1972). E. Hove, C. A. Elvehjem, and E. B. Hart, Am. J. Physiol. 119, 768 (1937). J. Quarterrnan, C. F. Mills, and W. R. Humphries, Biochem. Biophys. Res. Comm. 25, 354 (1966). L. Boquist and A. Lernmark, Acta Path. Microbiol. Scand. 76, 215 (1969). A. M. Huber and S. N. Gershoff, J. Nutr. 103, 1739 (1973). H. P. Roth and M. Kirchgessner, Internal. Z. Vit. Ern. Forsch. 45, 201 (1975). D. G. Hendricks and A. W. Mahoney, J. Nutr. 102, 1079 (1972). L. K. Robinson and L. S. Hurley, ]. Nutr. 111, 869 (1981). W. Salant and L. E. Wise, J. Biol. Chem. 34, 447 (1918). F. Y. Berenshtein and M. I. Shkol-Nik, Fiziol. Zhur. S.S.S.R. 37, 120 (1951). N. Hartmann, H. Fiedler, M. Ahrens, and M. Hansch, Wiss. Z. Karl-MarxUniv. Leipzig, Math-Naturwiss. R.21, 531 (1972).
Biological Trace Element Research
Vol. 24, 1990
82
B r a n d 2 o - b l e t o et aL
18. G. Weitzel, F. J. Strecker, and U. Roester, Ztschr. Physiol. Chem. 292, 286 (1953). 19. K. R. Etzel and R. J. Cousins, J. Nutr. 113, 1657 (1983). 20. E. R. Arquilla, S. Packer, W. Tarmas, and S. Myamoto, Endocrinoloy,y 103, 1440 (1978). 21. L. Coulston and P. Dandona, Diabetes 29, 665 (1980). 22. K. M. Dubowski, Clin. Chem. 8, 215 (1962). 23. J. G. H. Vieira, E. M. K. Russo, R. M. B. Maciel, O. A. Germek, and A. R. Chacra, Rev. Bras. Patol. Clin. 16, 108 (1980). 24. D. F. Morrison, Multivariate Statistical Methods, 1st ed., McGraw-Hill, New York, 1967, pp. 338. 25. N. B6gin-Heick, M. Dalp6-Scott, J. Rowe, and H. M. C. Heick, Diabetes 34, 179 (1985). 26. W. R. Sutton and V. E. Nelson, Proc. Iowa Acad. Sci. 45, 115 (1938). 27. T. Ghafghazi, C. W. Ludvigsen, M. L. McDaniel, and P. E. Lacy, IRCS Med. Sci. 7, 122 (1979). 28. T. Ghafghazi, M. L. McDaniel, and P. E. Lacy, Diabetes 30, 341 (1981). 29. K. R. Etzel and R. J. Cousins, J. Nutr. 113, 1657 (1983). 30. E. Horak and F. W. Suderman Jr., Toxicol. Appl. Pharmacol. 32, 316 (1975). 31. D. P. Figlewicz, B. Formby, A. T. Hodgson, F. G. Schmid, and G. M. Grodsky, Diabetes 29, 767 (1980). 32. C. Ludvigsen, M. McDaniel, and P. E. Lacy, Diabetes 28, 570 (1981). 33. D. P. Figlewicz, A. Heldt, S. E. Forhan, and G. M. Grodsky, Endocrinology 108, 730 (1981).
Biological TraceElement Research
Vof.24, 1990
