0021-972X/78/4706-1183S02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1978 by The Endocrine Society

Vol. 47, No. 6 Printed in U.S.A.

Effect of Age on Fasting Plasma Levels of Pancreatic Hormones in Man* DAVID BERGER, ROBERT C. CROWTHER, JOHN C. FLOYD, JR., SUMER PEK, AND STEFAN S. FAJANS Department of Internal Medicine, Division of Endocrinology and Metabolism and the Metabolism Research Unit, The University of Michigan, Ann Arbor, Michigan 48109 ABSTRACT. The effect of age and adiposity on fasting plasma levels of pancreatic polypeptide (hPP), glucagon (IRG), insulin (IRI) and glucose was examined in 263 healthy subjects between the ages of 20-69 yr. Mean plasma levels of hPP rose continuously from the third through the seventh decades. Mean plasma levels of IRG rose within the third and fourth decades but failed to rise further thereafter. Mean plasma levels of IRI did not change with age. Mean plasma levels of glucose rose by approximately 2 mg/dl • decade. The correlations of age with hPP, IRG, glucose, and adiposity were 0.47, 0.35, 0.25 (all P < 0.01) and 0.15 (P < 0.05), respectively. When adjustments were made for adiposity, the corre-

W

E HAVE reported that the mean fasting plasma level of immunoreactive human pancreatic polypeptide (hPP) is 51 pg/ml in young healthy adults between the ages of 20-34 yr (1-3). Subsequently, evidence was obtained in these and in additional subjects that fasting plasma hPP correlated with age (4). In the present study, we have examined the effect of age on plasma levels of hPP in a larger number of subjects who span a greater range of age and have added similar observations with respect to plasma levels of immunoreactive glucagon (IRG), imunoreactive insulin (IRI), and glucose. We also have examined relationships among these and other var-

lations of age with hPP, IRG, and glucose remained. Adiposity correlated with IRI, IRG, and glucose but when age correction was made, only the correlation of adiposity with IRI persisted. We conclude that: 1) age has a significant effect on fasting plasma levels of hPP and IRG; 2) the patterns of the age-related changes in hPP and IRG are not the same, suggesting that there are differences in the mechanism^) by which age influences plasma levels of these two pancreatic hormones; and 3) age should be considered in the interpretation of fasting plasma levels of hPP and IRG. (J Clin Endocrinol Metab 47: 1183, 1978)

iables and the relative importance of age and body weight on hPP, IRG, IRI, and glucose. Materials and Methods

Sixty young healthy adults volunteered for this study. They ranged in age from 20-34 yr. All were in good health and had no family history of diabetes mellitus. Additional healthy subjects were recruited from among faculty and staff members of the University of Michigan who had responded to an invitation for routine health appraisal. Subjects with chronic or acute disease, who were taking medications, or who had a family history of diabetes mellitus were excluded. Of the 348 faculty and staff members and 60 young adults sampled, 226 males and 37 females, ranging in age from 20-69 yr, were included in the analysis (203 faculty and staff and 60 young adults). Received March 6, 1978. Address requests for reprints to: John C. Floyd, Jr., Subjects reported to the laboratory after an overM.D., C7009 University Hospital, Ann Arbor, Michigan night fast of 8-16 h (mean, 14 h). On the basis of 48109. demographic data, which included body weight, * This work was supported in part by USPHS Grants these subjects were judged to have been consuming AM-02244, AM-00888, and TI AM-05001 from the NIAMDD and 5P11-GM-15559 from the National Insti- diets of composition similar to that of most of the tutes of General Medical Sciences; the American Diabetes American middle class. Measurements of height Association-Michigan Affiliate; the Upjohn Co. (Kala- and weight were followed by a 15- to 30-min period mazoo, MI); Pfizer, Inc. (New York, NY); and the Eli when subjects were seated, during which time hisLilly Co. (Indianapolis, IN). It was presented in part at the 37th Annual Meeting of the American Diabetes As- torical information was recorded. Informed consent was obtained from all subjects. Samples of periphsociation. 1183

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 07:55 For personal use only. No other uses without permission. . All rights reserved.

1184

BERGER ET AL.

eral venous blood were then drawn, placed into heparinized tubes, and refrigerated immediately. Plasma was separated within 2 h after drawing and stored at -20 C until assayed. Determination of hPP Plasma hPP was measured by a double antibody RIA. Standard hPP (lot 615-10543-200) and rabbit anti-hPP serum (lot 615-1054B-248-18) were the generous gifts of Dr. Ronald Chance of the Eli Lilly Co. (Indianapolis, IN). The antiserum does not cross-react with insulin, glucagon, somatostatin, gastrin, secretin, or GH (4). The sensitivity of the assay is 0.3 ±0.1 pg/tube (mean ± 2 SD). The slope of the logit dose-response curve is —2.159 ± 0.13 (mean ± SE) and the 50% inhibition point is 54 ± 0.1 (mean ± SD) pg/tube (45 assays). Interassay variability has been estimated with two samples of plasma. Mean ± SD for the two samples in 15 assays were 44 ± 26 and 2120 ± 545 pg/ml. For the same two samples, intraassay variability was (mean ± SD for 10 determinations) 53 ± 10 and 1862 ± 220 pg/ml. The recovery of hPP added to plasma pools is 109 ± 2% (3 determinations). The plasma levels reported herein were obtained from 30 separate assays; data from 15 of the assays were those used in calculating the interassay variability given above. Interassay variability was unlikely to have made spurious the results of statistical analyses involving plasma levels of hPP for the following reasons. 1) The 203 faculty and staff members were sampled on or near their birthdates, which occurred over a 3-month period. As their samples arrived in the laboratory during these 3 months, they were run in the next assay. Assays were run every 1-2 weeks and the 203 samples were distributed among 11 consecutive assays. Among these 11 assays there were no statistical differences of the mean ages of the subjects per assay. 2) During the 3 months of faculty sampling, 33 young adults, aged 20-29 yr (not otherwise reported here), were randomly assigned to the 11 assays; their fasting plasma hPP was 56 ± 5 pg/ml (mean ± SE). The 60 young adults (20-29 yr) included in this report were randomly sampled and assayed in 19 assays in 6 months immediately preceding the 3 months of faculty and staff collection; their fasting plasma hPP was 54 ± 4 pg/ml (mean ± SE). The hPP assay gave nearly identical mean results for the age range of 20-29 yr over the 9 months of this study. 3) The age-hPP correlation was statistically significant among the plasma of 125 of the 203 faculty members when they were assayed in a single assay at a later date.

JCE & M • 1978 Vol47 • No6

Determination of IRG Plasma IRG was measured by a double antibody RIA method described previously (5,6) using a beef and pork glucagon standard (Eli Lilly and Co., lot 258-2348-167-1). The antiserum, G-9-I (Ann Arbor, MI), reacts with the carboxyl-terminal region of glucagon and not with the amino-terminal region. It has negligible cross-reactivity with glucagon-like materials present in jejunal extracts (5, 6). Determination of IRI Plasma IRI was measured by a modification of the method of Morgan and Lazarow (7) using a human insulin standard (Eli Lilly and Co., lot 516734-B-33) (8). Determination of glucose Plasma glucose was measured by a hexokinase method (9) using a Gilford 3500 spectrophotometer and reagent kits purchased from the Worthington Biochemical Corp. (catalog no. 27615). Statistical analysis RIA results were computed using a program (RIANAL) described by Duddelson, Midgley, and Niswender (10). Statistical analyses were performed using a computer program (MIDAS) developed by the Statistical Research Laboratory of the University of Michigan (11). Statistical comparisons were made by one-way analysis of variance using Scheffe's method to evaluate intergroup comparisons (12). The distributions of plasma hPP, IRG, IRI, and glucose were examined. The distribution of hPP was markedly skewed. Tests of the shape of the distribution (12) suggested that hPP was log normally distributed. Thus, logarithms of hPP were used in the analysis of variance, linear regression, and correlation analyses which involved levels of hPP. To examine the influence of any given variable on the correlation among other variables, partial correlation coefficients were computed (12). By this method it is possible to remove the effect of the fixed variable, allowing one to examine the residual correlations among the remaining free variables. To examine the possibility that combined age and weight factors were present, two-way analyses of variance were performed for plasma hPP, IRG, IRI, and glucose using decade of age and quartiles of percentage of ideal body weight as the factors in a two-way model with replications (12).

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 07:55 For personal use only. No other uses without permission. . All rights reserved.

AGE AND PLASMA hPP, IRG, IRI, AND GLUCOSE

1185

Results 103

1

CO

co

oo

oo

i>

2-10 107 0-11 111 8-12 113 2-11 109

So

CO

00

o

bo

o co r1 CM

Effect of age on fasting plasma levels of pancreatic hormones in man.

0021-972X/78/4706-1183S02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1978 by The Endocrine Society Vol. 47, No. 6 Printed in U...
713KB Sizes 0 Downloads 0 Views