Diurnal Variations in Cardiovascular Function and Glucose Regulation in Normotensive Humans
To define the physiological relationships between cardiovascular function, glucose regulation, and insulin secretion, we submitted nine young normotensive subjects to ambulatory blood pressure monitoring and blood sampling at 20-minute intervals for 24 hours to measure glucose, insulin, C peptide, cortisol, and growth hormone. Subjects ingested three identical carbohydrate-rich meals in the morning (8:30 am), early afternoon (2 pm), and evening (8 pm). On the following day, they underwent an intravenous glucose tolerance test for quantification of insulin sensitivity. Significant postmeal increases in systolic pressure averaging 18±10 mm Hg in the morning, 18±8 mm Hg in the early afternoon, and 26±19 mm Hg in the evening were observed. Postprandial variations in diastolic pressure and heart rate were significant only for the morning meal. The magnitude of the postprandial increases in systolic pressure was correlated with the amount of insulin secreted in the morning but not later in the day. Pulses of growth hormone consistently occurred 3 to 4 hours after the morning and midday meals, as well as after the onset of sleep. Our findings indicate that under normal conditions, there is a quantitative relationship between postprandial insulin secretion and blood pressure.
Epidemiological and clinical studies have shown that white patients with essential hypertension are insulin resistant compared with normotensive subjects.1 2 The frequent association between insulin resistance and hypertension is part of a commonly observed clinical presentation called syndrome X.3 In hypertensive subjects, the reduction in insulin sensitivity is found even in the absence of obesity.4 Some recent studies5 6 7 8 9 have emphasized the impairment in the vasodilator properties of insulin as a primary cause for the relationship between hypertension and insulin resistance, whereas others10 11 have proposed that insulin resistance is only a hemodynamic consequence of high BP. Despite the evidence pointing to the existence of a fundamental link between insulin sensitivity and BP, the interactions among BP, glucose regulation, and insulin action under the normal physiological condition of meal ingestion have not yet been defined in detail.
A number of reports have claimed that meal presentation and ingestion consistently elevates cardiac output and HR in normotensive subjects.12 13 14 Findings on postprandial BP changes have conflicted.15 16 17 These studies, as well as the majority of studies on insulin resistance and hypertension, were based on isolated BP and HR measurements. Therefore, it has not been possible to distinguish cardiovascular effects related to behavioral activation preceding food intake from those associated with metabolic changes. Moreover, casual BP readings have been shown to frequently reflect reactivity to the clinical setting rather than basal cardiovascular function18 and to be uncorrelated with fasting insulin levels in both normotensive and hypertensive subjects.19 20 During the past few years, the availability of portable devices for ambulatory BP and HR monitoring has improved our knowledge of BP variability under physiological and pathological conditions and has revealed the existence of a normal diurnal variation in BP and HR, with higher daytime than nighttime levels. This day-night variation appears to be primarily caused by the combined effects of postural changes and the sleep-wake transitions.21 The possible effects of psychological and behavioral anticipation of meal presentation and ingestion per se on daytime BP variability have not yet been examined during continuous monitoring.
A number of studies have shown that blood glucose levels in response to a mixed meal are markedly higher in the late afternoon or evening than in the morning and that both the size of the responses and their diurnal variation are more pronounced when the meal has a high carbohydrate content.22 23 24 The reduced glucose tolerance in the evening appears to reflect both a decrease in insulin sensitivity and a reduction in insulin secretion25 and could be partially mediated by circadian variations in circulating concentrations of cortisol, a counterregulatory hormone.26 Putative associations between metabolic and cardiovascular responses to a meal challenge are thus likely to depend on the time of day and to be more apparent for meals with a high carbohydrate content.
We therefore designed the present study to examine in detail the chronobiology of BP and HR, glucose regulation, and counterregulatory hormone secretions in healthy, nonobese, normotensive subjects during a 24-hour cycle during which three identical high-carbohydrate meals were ingested. In each subject, insulin sensitivity was measured in a separate experiment for examination of the putative relationships between postmeal variations in cardiovascular parameters and insulin sensitivity.
Nine healthy, young volunteers (mean age, 24.5±2.5 years; 6 men, 3 women) agreed to participate in this study. None had a personal or family history of hypertension, diabetes, or psychiatric illness. All of them had normal weight for height (body mass index, 22.7±2.1 kg/m2), and none took any medication during the 3 months before the experiment. The women did not use hormonal contraception and were studied in the follicular phase of the menstrual cycle. Shift workers, subjects with sleep complaints, and subjects who had experienced recent (<3 months) transmeridian travel were excluded from the protocol. Before inclusion in the study, all subjects had a physical examination and routine laboratory tests, including an oral glucose tolerance test interpreted with the standard criteria of the National Diabetes Data Group.27
The protocol was approved by the institutional Review Committee. All subjects gave written informed consent and were paid for their participation in the study. All procedures used were in accordance with institutional guidelines. The investigations were performed in the Sleep Laboratory of Erasme Hospital, Université Libre de Bruxelles (Belgium). Each investigation involved 3 consecutive nights of hospitalization followed by a half-day outpatient admission on the 4th day. The first 2 nights (days 1 to 2) served to habituate the subjects to the laboratory environment. Bedtimes were chosen to be as similar as possible to the subject's usual sleep-wake cycle, but the subjects were required to be recumbent in total darkness by midnight and awake in normal indoor light by 7:45 am.
In the morning after the 2nd night of habituation, ie, on day 3, the subjects were equipped with an ambulatory BP monitor that is based on an auscultatory technique (Accutracker II, Suntech Medical Instruments Inc). The device, fitted to the nondominant arm, was programmed to take BP and HR readings every 10 minutes for 25 hours. Around 10 am, a catheter for blood sampling was inserted into an antecubital vein and kept patent by a slow drip of heparinized saline for collection of blood samples at 20-minute intervals over the next 26 hours. Hormonal data collected during the first 2 hours were discarded to eliminate artifactual changes related to the stress of venipuncture. Thus, for each volunteer, the first data point was collected at noon. During the night, blood sampling was performed from the adjacent room through a hole in the wall to avoid disturbing the subject's sleep. The subjects ate four exactly identical 700-kcal meals (60% carbohydrate, 25% protein, and 15% fat; same type and amount of food in each meal) at 8:30 am (ie, 3.5 hours before the beginning of blood sampling and BP monitoring), 2 pm, and 8 pm on day 3 and 8:30 am on day 4. The meals were spaced by at least 5.5 hours to allow for glucose to return to basal levels before ingestion of the next meal. During the daytime period, the subjects were free to ambulate in the room, but they were not allowed to lie down or nap. However, physical activity was severely limited by the intravenous sampling line.
At the end of the experiment, ie, shortly after noon on day 4, the subjects were discharged and then readmitted as outpatients on the following morning, after they had fasted overnight for at least 12 hours, to undergo an intravenous glucose tolerance test modified by an injection of tolbutamide as described by Bergman et al.28
All samples from the same subject were analyzed in duplicate in the same assay. Plasma glucose was measured by a glucose analyzer (STAT 2300, Yellow Springs Instrument Co) with a coefficient of variation of less than 2%. Insulin levels were determined by a commercially available radioimmunoassay (Pharmacia Diagnostics AB) with a limit of sensitivity of 18 pmol/L and intra-assay coefficient of variation averaging 5%. Plasma C peptide levels were determined by a previously described radioimmunoassay29 with an intra-assay coefficient of variation averaging 6%. GH levels were analyzed by a commercially available radioimmunometric assay (Medgenix). The limit of detection was 0.2 μg/L, and the intra-assay coefficient of variation averaged 14.6% in the range of 0.2 to 1.0 μg/L, 4.1% in the range of 1.0 to 5.0 g/L, and 1.9% for concentrations above 5.0 μg/L. Plasma cortisol levels were measured by radioimmunoassay (Diagnostics Products) with a limit of detection of 27 nmol/L and an average intra-assay coefficient of variation of 5%.
Sleep Recording and Analysis
Staging of polygraphic sleep recordings was performed by an experienced rater following the criteria of Rechtschaffen and Kales.30 Sleep onset and morning awakening were defined, respectively, as the first and last 20-second intervals scored as stage II, III, IV, or rapid eye movement (REM) sleep. The sleep period was calculated as the time between sleep onset and morning awakening. Sleep efficiency was calculated as (Sleep Period−Total Duration of Wakefulness)/Total Time in Bed.
BP and HR
Each individual profile of BP and HR was analyzed as previously described.21 Briefly, after artifactual data (pulse pressure <15 mm Hg or isolated increase/decrease >50% compared with the preceding measurement) were edited, the missing values were replaced by linear interpolation between the two adjacent values. Then, for each SBP, DBP, and HR profile, the overall 24-hour variation was quantified by building a best-fit curve based on periodogram calculations.31 The best-fit pattern can be unimodal, bimodal, or trimodal, with acrophases and nadirs defined as the times of occurrence of maxima and minima, respectively. The amplitude of the variation is defined as 50% of the difference between the highest acrophase and the lowest nadir and can be expressed in absolute units (ie, millimeters of mercury or beats per minute) or relative units (ie, percentage of the 24-hour mean level).
Insulin Secretory Rates
Insulin secretory rates were mathematically derived from plasma C peptide levels with a mathematical model for distribution and metabolism, with individual parameter values estimated from standard population values as previously described.32 33
Postmeal Glucose, Insulin, BP, and HR Responses
For each meal, the premeal levels of plasma glucose, serum insulin, and insulin secretory rates were calculated as the mean level during the hour preceding meal presentation. The postmeal response was then estimated as the area under the curve above the premeal level, and the duration of the response was calculated as the delay between meal presentation and the time when concentration returned to the premeal level. The onset of the metabolic response was defined as the timing of the first significant increase in glucose above the premeal level based on the same criteria as used for pulse identification.34 For BP and HR, the evaluation of meal-related changes was separated into two phases. First, changes in BP and HR occurring within ±20 minutes of the scheduled mealtime but before the onset of a significant elevation of plasma glucose (ie, anticipatory response) were examined by ANOVA for repeated measures, with time as the independent variable. The magnitude of periprandial changes in BP and HR occurring before detectable metabolic responses was quantified as the (positive or negative) difference between the maximum and minimum BP and HR levels. Second, the magnitudes of BP and HR changes coincident with or subsequent to the onset of the glucose response were calculated as the difference between the highest (or lowest) value observed during the first 40 minutes of the metabolic response and the last BP or HR value measured before the increase in plasma glucose.
Insulin sensitivity was determined from the intravenous glucose tolerance test data using the “minimal model” analysis of Bergman et al.28
Analysis of Cortisol and GH Profiles
The circadian variation of cortisol levels was evaluated by the periodogram method as described above for BP and HR profiles. Cortisol secretory rates were mathematically derived from the plasma cortisol concentration with a two-compartment model for cortisol distribution and metabolism as described elsewhere.22 A postmeal response in cortisol secretory rate was considered to be present if a significant secretory pulse identified by a computerized algorithm for hormonal pulse detection34 occurred within 60 minutes of meal presentation. The magnitude of the response was estimated by integration of the secretory rates over the duration of the postmeal pulse. Significant pulses of GH secretion were identified by a modification of the same algorithm for pulse detection34 with a threshold of three times the coefficient of variation of the assay in the relevant range of concentration. The amount of GH secreted in each pulse was estimated by deconvolution based on a one-compartment model for GH clearance with individually adjusted half-life periods (mean±SE: 17±1 minutes) and a volume of distribution of 7% of body weight as previously described.35
Effects of time of day on responses to meals were analyzed by ANOVA for repeated measures. Periprandial variations in BP and HR occurring before the onset of metabolic responses were tested by ANOVA for repeated measures over the period −20 to +20 minutes around the scheduled time of meal presentation. Similarly, variations in BP and HR coincident with or following the glucose elevation were also tested by ANOVA for repeated measures over the period from 0 to 40 minutes, where time 0 corresponds to the last BP or HR data measured before the onset of the metabolic response. Unless otherwise indicated, all group results are expressed as mean±SE.
Twenty-Four–Hour BP and HR Profiles
As illustrated in Fig 1⇓, BP and HR profiles generally conformed with previous descriptions,21 showing an abrupt decline around bedtime and a nocturnal nadir at midsleep. The quantitative characteristics of the profiles are summarized in Table 1⇓. Periodogram analysis showed the overall diurnal variation to be significant in 7 of 9 SBP profiles, 7 of 9 DBP profiles, and all HR profiles. Bimodal patterns, with a morning acrophase, an evening acrophase, and a nighttime nadir, were observed in 5 of 9 SBP profiles, 4 of 9 DBP profiles, and 7 of 9 HR profiles. The amplitude of the diurnal rhythm, when expressed as a percentage of the 24-hour mean level, was lowest for SBP, highest for DBP, and intermediate for HR.
The individual profiles shown in the right panels of Fig 1⇑ exemplify the variability of BP and HR in the high-frequency range. Rapid fluctuations of BP and HR around the times of meal presentation are apparent, but similar variability can be observed at other times of day, in the absence of identifiable stimuli. Nevertheless, as can be seen on the mean profiles shown in the left panels, robust short-term periprandial BP elevations, confirmed by the statistical analyses described below, emerged from the group data.
Cardiovascular Responses to Meal Presentation and Ingestion
Table 2⇓ gives the characteristics of the cardiovascular responses to the identical meals ingested in the morning, midday, and evening. Changes in SBP, DBP, and HR that coincided with or followed the onset of the postmeal increase in blood glucose levels were considered to represent cardiovascular effects related to meal ingestion (quantified in the top part of Table 2⇓), whereas changes in SBP, DBP, and HR that occurred around the scheduled meal time but before the elevation of blood glucose levels were considered to reflect cardiovascular responses to meal presentation (quantified in the bottom part of Table 2⇓). Fig 2⇓ illustrates the mean profiles of SBP, DBP, and HR referenced to the timing of basal premeal plasma glucose level, ie, the mean cardiovascular responses to meal ingestion. Most subjects exhibited a significant rise in SBP in response to meal ingestion, with maximal SBP occurring 20 to 30 minutes after the beginning of the elevation of plasma glucose. This increase in SBP was of similar magnitudes for the morning and midday meals and tended to be larger for the evening meal. Changes in DBP and HR following meal ingestion were significant only after the morning meal. Cardiovascular changes preceding the beginning of the glucose response were most consistently observed at midday and in the evening and, at these times, were significant at the group level for both SBP and DBP (bottom part of Table 2⇓). These anticipatory BP changes were of magnitudes similar to those occurring following the elevation in blood glucose. Changes in HR before the onset of the metabolic response were less consistent and reached significance for the evening meal only. No significant anticipatory response of either BP or HR could be detected for the morning meal, possibly because of a masking effect of the large postawakening increase in BP and HR occurring coincidentally. Examples of anticipatory cardiovascular responses are shown in Fig 3⇓.
Glucose, Insulin, Cortisol, and GH Responses to Meals
As evident in the profiles illustrated in Fig 4⇓, the glucose responses and, to a lesser extent, the insulin responses to identical meals varied markedly according to the time of day (Table 3⇓). For plasma glucose, the maximal increment, area under the curve, and duration of the response were larger at midday and in the evening than in the morning. For serum insulin, effects of time of day were significant only for the duration of the postmeal response. The maximal insulin secretory rate after meal ingestion did not vary significantly with time of day. However, the amount of insulin secreted was higher at midday and in the evening than in the morning because of a more prolonged response.
Meal ingestion resulted in significant pulses of cortisol secretion at midday in 8 of 9 subjects and in the evening in 5 of 9 subjects. Confirming previous observations,20 the secretory response was larger at midday (2940±340 nmol·min, n=8) than in the evening (1840±340 nmol·min, n=5, P<.05).
The mean GH profile (Fig 4⇑) revealed a consistent temporal relationship between GH secretory pulses and the timings of meals. Two large secretory pulses occurred consistently 4 to 5 hours after both the morning and midday meals. In individual profiles, significant GH pulses were detected in 7 of 9 subjects at 273±24 minutes after breakfast (amount of GH secreted: 19.4±5.0 μg) and at 277±23 minutes after lunch in 8 of 9 subjects (amount of GH secreted: 19.9±6.8 μg). Sleep onset occurred on average 222±10 minutes after the evening meal, ie, before a similar postprandial GH pulse could be observed while the subjects were still awake. As expected,35 a large pulse of GH secretion occurred within 1 hour of sleep onset in all subjects.
Relationships Between Metabolic and Cardiovascular Responses to Meals
The relationships between the parameters characterizing postmeal cardiovascular responses (ie, changes in SBP, DBP, and HR) and the parameters characterizing postmeal metabolic responses (ie, maximal increment, area under the curve, and duration of response for glucose, serum insulin, and insulin secretory rate; amount secreted for cortisol and GH) were examined for each meal separately. A highly significant correlation, illustrated in Fig 5⇓, was observed between SBP postmeal increase and the amount of insulin secreted after the morning meal (r=.94, P<.0001). After the midday meal, the correlation between the postmeal SBP increase and the amount of insulin secreted in response to the meal was no longer significant (r=.55) because of one outlying subject. This subject had a large anticipatory SBP increase and, perhaps as a result, a small postmeal SBP elevation despite a high insulin secretory response. When this subject was excluded from the calculation, the correlation between postmeal SBP increase and amount of insulin secretion after the midday meal became highly significant (r=.93, P<.0001). After the dinner meal, there was no correlation (r=.00) between postmeal increases in SBP and the insulin secretory response.
Overall Profiles of Plasma Cortisol and GH and Relationship to BP and HR
As shown in Fig 4⇑, plasma cortisol levels exhibited the classic circadian rhythmicity of corticotropic activity. Periodogram calculations detected a significant rhythm in all nine subjects, with a mean amplitude of 81±5%, a nocturnal nadir (44±11 nmol/L) in the early part of the night (00:15±37 minute), and a sharp early morning rise culminating in an acrophase of 350±11 nmol/L at 9:11±34 minutes. The amplitude of the diurnal variation in DBP, but not SBP or HR, tended to correlate with the amplitude of the cortisol rhythm (r=.70, P=.08 for the absolute amplitudes; r=.73, P=.06 for the relative amplitudes).
When mean profiles of HR and GH (Figs 1 and 4⇑⇑) were compared, it became apparent that the peaks of GH secretion were concomitant with falls in HR, primarily following sleep onset. This was reflected in a significant correlation between the relative amplitude of the 24-hour HR profile and the total amount of GH secreted during the 24 hours (r=.81, P<.01). Thus, the subjects who secreted the largest amounts of GH had the most pronounced nighttime HR decrease.
Insulin Sensitivity Index
Our young, healthy, lean subjects exhibited a high insulin sensitivity index of 15.6±1.9 (10−4/min)/(μU/L). There were no significant correlations between 24-hour mean SBP, DBP, and HR levels and insulin sensitivity index. There were also no consistent correlations between the quantitative characteristics of postmeal DBP or HR responses and insulin sensitivity index. However, the postmeal increase in SBP was inversely correlated with the insulin sensitivity index (r=−.75, P=.05) for the evening meal but not for the morning or midday meals.
The use of BP monitoring during a 24-hour period permitted us to demonstrate that in healthy, nonobese normotensive subjects, the ingestion of carbohydrate-rich meals induces short-term increases in SBP, irrespective of the timing of meal ingestion. Moreover, for the first time, a correlation between the increase in SBP and the amount of insulin secreted in response to a meal was observed. Thus, the subjects who secreted the lowest amounts of insulin had the smaller postmeal rises in SBP. These findings demonstrate the existence of a relationship between postprandial cardiovascular and metabolic variables under normal conditions.
Several previous reports12 13 14 15 16 17 have described the normal hemodynamic response to the ingestion of mixed meals and oral glucose. These studies were based on either infrequent casual BP and HR measurements or on short-term monitoring of intra-arterial pressure. A 20% to 40% increase in cardiac output, a 10% to 20% fall in peripheral vascular resistance, and an increase in HR (ranging in magnitude from 3 to 16 beats per minute) have been clearly demonstrated in young, healthy volunteers receiving oral glucose, liquid meals, or mixed solid meals.12 13 14 15 16 17 All except one14 of these previous studies were limited to the morning period. Our finding of a significant increase in HR after ingestion of the morning meal is thus in agreement with these earlier data. However, in our subjects, changes in HR after meal ingestion in the early afternoon and evening were not significant, suggesting that the positive findings of previous morning studies may not be readily generalized to the rest of the day.
Studies of the response in BP to oral glucose loading and mixed meals have given conflicting results. Increases, decreases as well as increases, and an absence of response have been reported. In some studies, SBP and DBP remained unchanged,14 16 and in others, SBP was unchanged or increased and DBP was decreased.12 13 15 Our findings clearly show an increase in SBP but not in DBP after ingestion of high-carbohydrate meals in the morning, midday, and evening. Again, it must be noted that whereas our study design allowed for the characterization of BP responses at different times of day, most previous studies were performed in the morning, when our findings showed the increase in SBP tending to be of lesser magnitude. Interestingly, the tendency for lower postmeal SBP responses in the morning than later in the day paralleled the pronounced difference in the size of the glucose response. Diurnal variations in postmeal responses in glucose and insulin secretion have been previously observed22 23 24 36 and were purposely enhanced in the present study by the rich carbohydrate content of the meals.
None of the previous studies of cardiovascular reactivity to meal ingestion involved a methodology that permitted the distinction between the changes in BP and HR anticipating the elevation of blood glucose and insulin levels and those paralleling or following the metabolic response. The former are less likely to be directly involved in the link between insulin resistance and hypertension than the latter. Anticipatory BP and HR increases were observed most consistently in the evening and were rarely present in the morning.
The amount of insulin secreted and the postmeal increment in SBP were positively correlated in the morning and, to a lesser extent, in the early afternoon but not in the evening. Several studies have indicated that glucose tolerance, insulin secretion, and insulin sensitivity are decreased in the latter part of the day compared with the morning and early afternoon.23 24 25 Factors underlying these evening alterations in the parameters of glucose regulation may be responsible for the absence of a quantitative correlation between SBP and insulin secretion at that time of day. The intriguing finding of a negative correlation between insulin sensitivity (which was measured in the morning) and the magnitude of the SBP rise in the evening but not in the morning or early afternoon will need confirmation in a larger subject population.
We measured plasma cortisol and GH levels on all samples to examine the possible roles of these counterregulatory hormones in modulating metabolic and cardiovascular responses. As in previous studies,22 37 postmeal cortisol elevations were detected, particularly after lunch. No relationship between postmeal cortisol secretion, glucose and insulin responses, and BP or HR variations could be shown. Consistent effects of meal ingestion on the 24-hour profile of plasma GH have not been previously identified. The use of high-carbohydrate meals and of a sensitive immunoradiometric assay probably facilitated the demonstration of consistent preprandial GH pulses. The occurrence of GH pulses during troughs of plasma glucose levels is likely to reflect the well-known inhibition of GH secretion by glucose, followed by a stimulation induced by falling glucose levels. A similar time course of plasma GH levels can be observed after ingestion of oral glucose.37
In conclusion, this study shows that under normal conditions, meal ingestion is associated with a consistent increase in SBP that is directly related to the amount of insulin secreted. The findings document a relationship between insulin secretion and BP in normotensive, nonobese, young subjects.
Selected Abbreviations and Acronyms
|DBP||=||diastolic blood pressure|
|SBP||=||systolic blood pressure|
This study was supported by a grant from ZENECA Pharmaceutical, Belgium, and by grants DK-41814, DK-31842, and DK-20595 from the National Institutes of Health, Bethesda, Md. We thank Mireille L'Hermite-Balériaux for the determination of GH levels, Françoise Pignez for secretarial help, Annette Fiasse for technical support, and Rood SA Belgium, which provided us with Accutracker II devices.
Reprint requests to Jean-Paul Degaute, MD, Hypertension Unit, Hopital Erasme, Université Libre de Bruxelles, 808 route de Lennik, B-1070 Brussels, Belgium.
- Received December 27, 1995.
- Revision received February 22, 1996.
- Revision received June 11, 1996.
Modan M, Halkin H, Almog S, Lusky A, Eshkol A, Shefi M, Shitrit A, Fuchs Z. Hyperinsulinemia: a link between hypertension, obesity and glucose intolerance. J Clin Invest. 1985;75:809-817.
Reaven GM. Role of insulin resistance in human disease. Diabetes. 1988;37:1595-1607.
Natali A, Santoro D, Palombo C, Cerri M, Ghione S, Ferrannini E. Impaired insulin action on skeletal muscle metabolism in essential hypertension. Hypertension. 1991;17:170-178.
Baron AD, Brechtel-Hook G, Johnson A, Hardin D. Skeletal muscle blood flow: a possible link between insulin resistance and blood pressure. Hypertension. 1993;21:129-135.
Vollenweider P, Randin D, Tappy L, Jequier E, Nicod P, Scherrer U. Impaired insulin induced sympathetic neural activation and vasodilatation in skeletal muscle in obese humans. J Clin Invest. 1994;93:2365-2371.
Baron AD, Brechtel G. Insulin differentially regulates systemic and skeletal muscle vascular resistance. Am J Physiol. 1993;256:E61-E67.
Steinberg HO, Brechtel G, Johnson A, Fineberg N, Baron AD. Insulin-mediated skeletal vasodilatation is nitric oxide dependent, a novel action of insulin to increase nitric oxide release. J Clin Invest. 1994;94:1172-1179.
Baron AD, Brechtel G, Johnson A, Fineberg N, Henry DP, Steinberg HO. Interaction between insulin and norepinephrine on blood pressure and insulin sensitivity: studies in lean and obese men. J Clin Invest. 1994;94:2453-2462.
Anderson EA, Mark AL. The vasodilator action of insulin: implications for the insulin hypothesis of hypertension. Hypertension. 1993;21:136-141.
Kelbæk H, Munck O, Christensen NJ, Godtfredsen J. Central haemodynamic changes after meal. Br Heart J. 1989;61:506-509.
Jern S. Effects of acute carbohydrate administration on central and peripheral hemodynamic responses to mental stress. Hypertension. 1991;18:790-797.
Lipsitz LA, Ryan SM, Parker A, Freeman R, Wei JY, Goldberger AL. Hemodynamic and autonomic nervous system responses to mixed meal ingestion in healthy young and old subjects and dysautonomic patients with postprandial hypotension. Circulation. 1993;87:391-400.
Grollman A. Physiological variations in the cardiac output in man, II: the effect of ingestion of food on the cardiac output, pulse rate, blood pressure, and oxygen consumption in man. Am J Physiol. 1929;89:370-399.
Degaute JP, van de Borne Ph, Linkowski P, Van Cauter E. Quantitative analysis of the 24-hour blood pressure and heart rate patterns in young men. Hypertension. 1991;18:199-210.
Van Cauter E, Shapiro ET, Tillil H, Polonsky KS. Circadian modulation of glucose and insulin responses to meals: relationship to cortisol rhythm. Am J Physiol. 1992;262:E467-E475.
Jarret RJ. Rhythms in insulin and glucose. In: Krieger DT, ed. Endocrine Rhythms. New York, NY: Raven Press Publishers; 1979:247-258.
Lee A, Ader M, Bray GA, Bergman RN. Diurnal variation in glucose tolerance: cyclic suppression of insulin action and insulin secretion in normal weight, but not obese, subjects. Diabetes. 1992;41:742-749.
Plat L, Byrne MM, Sturis J, Polonsky KS, Mockel J, Féry F, Van Cauter E. Effects of morning cortisol elevation on insulin secretion and glucose regulation in man. Am J Physiol. 1996;270:E36-E42.
National Diabetes Data Group. Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes. 1979;28:1039-1057.
Bergman RN, Prager R, Volund A, Olefsky JM. Equivalence of the insulin sensitivity index in man derived by the minimal model method and the euglycemic glucose clamp. J Clin Invest. 1987;79:790-800.
Faber OK, Binder C, Markussen J, Heding LG, Naithani VK, Kuzuya H, Blix P, Horwitz DL, Rubenstein AH. Characterization of seven C-peptide anti-sera. Diabetes. 1978;27(suppl 1):170-177.
Rechtschaffen A, Kales A. A Manual of Standardized Terminology Techniques and Scoring System for Sleep Stages of Human Subjects. Washington, DC: US Government Printing Office; 1968. National Institutes of Health publication 204.
Van Cauter E. Method for characterization of 24-h temporal variation of blood components. Am J Physiol. 1979;237:E255-E264.
Van Cauter E, Mestrez F, Sturis J, Polonsky KS. Estimation of insulin secretion rates from C-peptide levels: comparison of individual and standard kinetic parameters for C-peptide clearance. Diabetes. 1992;41:368-377.
Polonsky KS, Licinio-Paixao J, Given BD, Pugh W, Rue P, Galloway J, Karrison T, Frank B. Use of biosynthetic human C-peptide in the measurement of insulin secretion rates in normal volunteers and type I diabetic patients. J Clin Invest. 1986;77:98-105.
Van Cauter E. Estimating false-positive and false-negative errors in analysis of hormonal pulsatility. Am J Physiol. 1988;254:E786-E794.