1
Insulin Sensitivity in Postmenopausal Women
Exercise and Health
INSULIN SENSITIVITY AMONG HEALTHY, NONOBESE POSTMENOPAUSAL WOMEN WITH VARIOUS DEGREES OF SKELETAL MUSCLE MASS
Eric Goulet1,2, Christine Lord2, Jean Chaput3, Mylène Aubertin-Leheudre2,4, Florian Bobeuf4, Mélissa Labonté2,4, Isabelle J. Dionne1,2,4
1Department of Physiology and Biophysics/University of Sherbrooke, J1H 5N4
2Research Centre on Aging/ University of Sherbrooke/ J1H 4C4
3Department of Social and Preventive Medicine/LavalUniversity, Québec, Québec, Canada, G1K 7P4
4Faculty of Physical Education and Sports/University of Sherbrooke, Sherbrooke, Québec, Canada, J1K 2R1
ABSTRACT
Eric D.B. Goulet, Christine Lord, Jean P. Chaput, Mylène Aubertin-Leheudre, Florian Bobeuf, Mélissa Labonté, Isabelle J. Dionne. Insulin Sensitivity Among Healthy, Nonobese Postmenopausal Women with Various Degrees of Skeletal MuscleMass.JEPonline 2008;11(1):9-17. Aging is associated with a decrease in skeletal muscle mass and insulin sensitivity (IS). This reduction in muscle mass may decrease oxidative capacity and IS. Decreased IS increases proteolysis and reduces protein synthesis. Hence, it is believed that IS and skeletal muscle mass in the elderly are related. However, no study has examined this issue in healthy, nonobese postmenopausal women. Fifty-eight healthy postmenopausal women were distributed into four quartiles of muscle mass index (MMI). IS was determined using the quantitative IS check index (QUICKI), fat-free mass (FFM), fat-mass (FM) and trunk FM (TFM) using DXA and resting energy expenditure (REE) using indirect calorimetry. Total energy and nutrient intakes were also measured. MMI but not IS, FM, TFM, BMI, bodyweight, REE or total energy and nutrient intakes differed significantly among quartiles. No correlation was found between IS and MMI. Our results suggest that IS is not influenced to a great extent by skeletal muscle mass nor plays a substantial role in muscle loss in healthy, nonobese postmenopausal women.
Key Words:Aging, Insulin Resistance, Fat-Free mass, Fat Mass, Elderly Women
INTRODUCTION
Sarcopenia (1) and the decrease in insulin sensitivity (IS) (2) are two hallmark features of the aging process. Sarcopenia increases the risk for falls, fractures, and disability (3). A reduced IS increases the risk of developing glucose intolerance, dyslipidemia, atherosclerosis, and hypertension. Singly or collectively, these factors may contribute to the development of type II diabetes and cardiovascular diseases (4).
Because the decrease in skeletal muscle mass and IS occurs concurrently during aging, a question arises as to whether both are physiologically related. On the one hand, sarcopenia has been associated with abnormalities of the electron transport system of mitochondria (5), which reduces the oxidative capacity of skeletal muscle (6) that could contribute to a decrease in IS in the elderly (7). On the other, insulin plays a pivotal role in the control of protein balance, with a demonstrated ability to increase whole-body (8) and muscle protein synthesis (9) and depress muscle proteolysis (10). In contrast, it has been shown that a reduction in IS decreases the rate of whole-body (11) as well as muscle protein synthesis (12) and enhances the rate of muscle proteolysis (13).
These observations suggests a relationship between the level of IS and skeletal muscle mass in the elderly. To our knowledge, however, we are aware of no study that has yet examined this possibility. Shedding some light on this issue would be relevant for the health professionals. In fact, the demonstration of a relationship between IS and skeletal muscle mass in the elderly would help them refining their exercise and nutritional prescription for this particular population.
The purpose of this investigation was to examine whether there is a link between the level of IS and muscle mass index (MMI) in healthy, nonobese postmenopausal women after controlling for the confounding effect of trunk fat mass (TFM) and total fat mass (FM).
METHODS
Subjects participating in the current investigation were selected from a pool of 100 women who had participated in various studies conducted in our laboratory. Each subject had to meet all the following criteria to be included in this study: (a) caucasian, age 50 to 75 yrs (mean age: 64.4 ± 5.4yrs, range: 53-75 yrs); (b) fasting plasma glucose < 110 mg/dl; (c) no hypertension (< 140/90 mmHg); (d) sedentary (not having participated in a regular exercise program over the past 6 months); (e) nonobese (body mass index (BMI) < 30 kg/m2); (f) stable bodyweight (± 2 kg) for the last 6 months prior to the study; (g) non-smoker; (h) moderate drinker (< 15 g of alcohol/day); (i) absence of menses for the past 12 months; (j) not having used hormonal replacement therapy for 1 yr; and (k) not taking any medication that could alter metabolism. Fifty-eight healthy postmenopausal women met the criteria. They were selected to participate in this study. Older women were studied as opposed to men, since they have been shown to live longer, suffer from more physical disability (14), have a greater prevalence of sarcopenia (3) and display a lower IS (15), thereby increasing the pertinence of our study. All volunteers signed an informed consent form approved by the Ethics Committee of the ResearchCenter on Aging of the University of Sherbrooke.
Overview of the Experiment
Subjects arrived at the ResearchCenter on Aging at 7:00 am after an overnight fast. A venous blood sample was drawn and used to determine glucose and insulin levels. After blood collection, subjects underwent measurement of resting energy expenditure (REE) and body composition. Before leaving the laboratory, subjects were given verbal and written instructions on how to complete a 3-d food record and use an electronic food scale.
Measurement of Resting Energy Expenditure
Subjects were instructed to assume the supine position on a hospital bed, remain silent, and perform a minimum of movements. Each subject was fitted with a face mask that was linked to a gas analyzing system (CCM/D metabolic cart, Medical Graphics Corporation, USA). The analyzer had been calibrated according to the manufacturer's instructions. Subjects were instructed again to maintain a normal breathing pattern and remain as immobile as possible until the completion of the test. All measurement periods were conducted in a thermally neutral and quiet room. Resting energy expenditure (kcal/day) was assessed during 30 min. However, only the last 15 min of the data from the test were used to estimate REE (16), which was calculated using the equation of Weir (17). The coefficient of variation (CV) for REE measured with the CCM/D metabolic cart is 1.5% in the laboratory.
Measurement of Body Composition
Height without shoes was measured using a wall-mounted stadiometer. Bodyweight was measured to the nearest 100 g using an electronic scale (Model 707, Seca, Germany) with subjects dressed in a hospital gown. Fat mass and fat-free mass (FFM) were measured using dual-energy x-ray absorptiometry (DXA, Lunar Prodigy, General Electric, USA). The laboratory test-retest CVs for the measures of FM and FFM using the DXA are 4.9% and 1.1%, respectively. Trunk fat mass was determined from a region extending from the top of the iliac crests to the top of the shoulders with the arms excluded, as previously described by Clasey et al. (18). Skeletal muscle mass was estimated from the appendicular FFM (legs and arms FFM) values using the formula developed by Kim et al. (19): [appendicular FFM (kg) x 1.19] - 1.01. Skeletal muscle mass determined with this formula correlates strongly and significantly with the skeletal muscle mass measured with magnetic resonance imaging(r = 0.98). Muscle mass index was then computed with the following formula: skeletal muscle mass/height (m2). Subjects were distributed into four quartiles of MMI created using SPSS software as follows:a) ≤ 6.89 kg/m2; b) 6.90 ≤ 7.53 kg/m2; c) 7.54 ≤ 8.16 kg/m2; and d) 8.17 kg/m2.
Dietary Assessment
Energy and nutrient intakes were assessed using a 3-d food record. It has been shown that 3-d dietary records provide a good estimate of energy and nutrient intakes (20) in addition of having been proven a valid method in elderly without cognitive impairments (21). Subjects were provided a portable electronic food scale and a food log. Each subject was thoroughly instructed how to correctly use the scale and food log. Subjects were requested to record their food intakes over two consecutive weekdays and one weekend day. During these measurement periods, subjects were instructed to maintain their normal diet. Dietary analyses for daily energy, protein (animal and vegetal), fat and carbohydrate intakes were performed using the CANDAT software, version 6.0 (Candat System, Canada).
Insulin Sensitivity Measurement
Insulin sensitivity was assessed indirectly with the quantitative IS check index (QUICKI) using fasting plasma glucose and insulin values with the following formula: 1/{log [fasting insulin (mU/ml)] + [log fasting glucose (mg/dl)]} (22). In normal weight individuals, Katz et al. (22), Straczkowski et al. (23), and Yokoyama et al. (24) showed that this surrogate measure of IS correlates significantly (r = 0.48, 0.39 and 0.64, respectively) with that of the hyperinsulinemic-euglycemic clamp technique. Moreover, QUICKI has been demonstrated to be a good predictor of IS (25).
Assays
All plasma glucose values were measured with the chemistry analyser Vitros 950 using the glucose oxidase method (Johnson and Johnson, USA). The test-retest CV for glucose was 2.0%. Forty-three plasma insulin values were measured by a double-antibody radioimmunoassay (Intermedico, Canada) and the test-retest CV for insulin with this method was 8.0%. Fifteen plasma insulin values were measured using an ELISA kit (KAQ 1251, Biosource, Belgium). The test-retest CV with this method was 4.5%. All values of insulin measured were combined since the different techniques did not differ significantly from each other.
Statistical Analysis
All data were tested for normality of distribution using the Kolmogorov-Smirnov Test. Only fasting glucose and insulin levels were abnormally distributed and thus were log-transformed. In that the outcome of results was not change, the insulin and glucose data are reported in their original form. One-way ANOVA with Tukey's post-hoc tests was used to identify differences among the quartiles of MMI. Homogeneity of variances was tested using the Levene test. ANCOVA was used to determine whether there were differences among quartiles with respect to IS when controlling for FM and TFM. Pearson’s product-moment correlation analyses were performed to examine relationships between QUICKI or MMI with certain variables of interest. Significance was set at p < 0.05. Results are presented as mean ± SD. All statistics were performed using the SPSS software, version 9.0 (Chicago, USA).
RESULTS
Table 1 shows the physical characteristics of subjects among quartiles of MMI. With the exception of quartile 1 versus quartile 3, there was no difference in age between any of the quartiles.
As anticipated, MMI was significantly different among quartiles. Results show that there was no significant difference among quartiles of MMI regarding bodyweight, BMI, FM, TFM, and % body fat.
As indicated in Table 2, there were no differences among quartiles for IS and fasting plasma glucose and insulin levels.Daily energy intake as well as total animal, vegetal protein, carbohydrate and fat intakes were not different among quartiles of MMI, as shown in Table 3. Despite different levels of MMI, no significant difference in REE was noted among quartiles.
Table 1. Physical characteristics of subjects among the four quartiles.
Variables
/ Quartiles1 (n=14) / 2 (n=15) / 3 (n=15) / 4 (n=14)
Age (yrs)
/ 67.1 ± 6.21 / 65.3 ± 4.6 / 61.3 ± 5.0 / 64.1 ± 6.7Height (m)
/ 1.56 ± 0.05 / 1.59 ± 0.06 / 1.59 ± 0.04 / 1.58 ± 0.04Bodyweight (kg)
/ 61.8 ± 8.1 / 65.8 ± 8.6 / 66.5 ± 7.2 / 69.4 ± 7.9BMI (kg/m2) / 25.5 ± 2.7 / 26.1 ± 2.5 / 26.4 ± 2.7 / 27.7 ± 2.3
MMI (kg/m2) / 6.7 ± 0.22 / 7.3 ± 0.2 / 7.8 ± 0.2 / 8.5 ± 0.3
FM (kg) / 24.9 ± 6.7 / 26.2 ± 6.1 / 25.7 ± 6.9 / 26.7 ± 6.4
TFM (kg) / 11.4 ± 3.5 / 11.4 ± 3.3 / 11.9 ± 3.4 / 11.5 ± 3.4
Body fat (%) / 41.4 ± 6.0 / 41.1 ± 5.2 / 39.3 ± 6.9 / 39.7 ± 6.3
BMI: body mass index; MMI: muscle mass index; FM: fat mass; TFM: trunk fat mass.
1Quartile 1 different from quartile 3 (p < 0.05); 2All quartiles significantly different from one another (p < 0.05).
Table 2. Metabolic variables among the four quartiles.
Variables*
/ Quartiles1 (n=14) / 2 (n=15) / 3 (n=15) / 4 (n=14)
QUICKI / 0.387 ± 0.026 / 0.363 ± 0.034 / 0.376 ± 0.026 / 0.379 ± 0.035
Insulin (mU/ml) / 5.1 ± 1.9 / 7.7 ± 3.8 / 6.0 ± 2.5 / 6.6 ± 5.4
Glucose (mg/dl) / 71.6 ± 29.1 / 60.1 ± 41.8 / 63.6 ± 39.5 / 60.3 ± 37.1
*None of the values differ significantly from one another.
Table 4 shows the correlations between QUICKI or MMI with variables of interest. QUICKI did not correlate with MMI but correlated significantly and positively with age, and significantly and negatively with bodyweight, BMI, FM and TFM. No relationships were observed between QUICKI and REE, total energy and protein intakes (including animal protein). On the other hand, MMI correlated significantly and positively with bodyweight and BMI but no correlations were observed to determine whether there is a relationship between IS and MMI in healthy with age, FM, TFM, REE, total energy and protein intakes (including animal protein).
DISCUSSION
Our objective was, nonobese postmenopausal women. This was the first study to examine this particular issue in this population. Using QUICKI as a surrogate measure of IS, we showed no difference in IS between quartiles of MMI in the cohort studied. Moreover, there was no correlation between IS and MMI. Therefore, these results suggest that in healthy, nonobese postmenopausal women, the action of insulin in the fasting state is not influenced by skeletal muscle mass nor does it play a substantial role in its loss during aging
Table 3.Nutrition-related components among the four quartiles.
Variables*
/ Quartiles1 (n=14) / 2 (n=15) / 3 (n=15) / 4 (n=14)
Energy intake (kcal/kg BW/day) / 35.2 ± 11.2 / 33.4 ± 9.5 / 34.0 ± 8.1 / 29.1 ± 12.3
Protein intake (g/kg BW/day) / 1.4 ± 0.6 / 1.4 ± 0.5 / 1.5 ± 0.5 / 1.3 ± 0.4
Animal protein intake (g/kg BW/day) / 1.0 ± 0.6 / 0.8 ± 0.4 / 0.9 ± 0.5 / 0.8 ± 0.4
Vegetal protein intake (g/kg BW/day) / 0.7 ± 0.2 / 0.7 ± 0.3 / 0.6 ± 0.2 / 0.5 ± 0.2
CHO intake (g/kg BW/day) / 4.2 ± 1.5 / 3.6 ± 1.6 / 3.7 ± 0.8 / 3.6 ± 1.6
Lipid intake (g/kg BW/day) / 1.4 ± 0.8 / 1.5 ± 0.4 / 1.5 ± 0.6 / 1.0 ± 0.6
BW: bodyweight
CHO: carbohydrate
*None of the values differ significantly from one another.
Variables / QUICKIr / MMI
r
MMI / - 0.10 / ---
Age / 0.34* / -0.23
Bodyweight / -0.35* / 0.29*
BMI / - 0.36* / 0.28*
FM / - 0.34* / 0.03
TFM / - 0.46* / 0.00
REE / 0.27 / 0.03
TEI / 0.17 / -0.21
TPI / 0.14 / -0.15
TAPI / 0.07 / -0.17
Table 4. Correlations between certain variables of interest.
Insulin sensitivity level was similar among quartiles despite significant differences in MMI. Several studies that examined the effect of aging or abdominal fat on IS provide interesting secondary data arguing against the role of the reduction in skeletal muscle mass in the aging-associated decline in IS and glucose tolerance. Imbeault et al. (26) looked at the role played by aging per se on insulin and glucose responses in 200 young and middle-aged women during an oral glucose tolerance test (OGTT). Younger women had 4 kg more FFM than older women, and when visceral fat was controlled for, no differences in glucose and insulin responses were noted between groups.
Korth et al. (27) looked at the insulin and glucose responses of younger and older men and women during an OGTT. Waist circumference, a surrogate measure of abdominal obesity, was similar among individuals and despite the difference in FFM among groups (young men: 72 kg; older men: 55 kg; young women: 55 kg; older women: 33 kg), no significant differences in the insulin and glucose responses were noted among groups. Finally, Paolisso et al. (28) found no difference in IS, as measured with the glucose clamp technique, between healthy middle-aged (45 yrs old) and centenarian men and women. Fat-free mass was 10% lower in centenarian but the waist to hip ratio was similar between groups. If muscle mass had played a prominent role in the control of glucose metabolism in these studies, differences in IS should have been observed. Our results, derived from a static measure of IS, are therefore in agreement with those investigations that used more direct measures of IS and suggest that skeletal muscle mass plays, at best, only a minor role in the age-related decline in IS.
In light of these observations, it appears that it is the accumulation of abdominal fat that favors the decline in IS with aging while the amount of skeletal muscle mass plays a secondary role. This is consistent with the large body of evidence accumulated over the past decade indicating that the accumulation of abdominal fat is a major predictor of insulin resistance (29). The present results are in accordance with this observation in that the lack of difference in IS was associated with no difference in TFM among quartiles of MMI. Additionally, there was a significant relationship between IS and TFM in this study. From a metabolic standpoint, results of the present study indicate that it is preferable to limit the gain in abdominal fat rather than attempting to gain muscle mass when the goal is to maintain or improve IS during aging.
Research shows that insulin resistance exists in regard not only of glucose metabolism but also of protein metabolism. Chevalier et al. (11) demonstrated a significant correlation between whole-body protein anabolic response and IS during a hyperinsulinemic-euglycemic isoaminoacidemic clamp. Because we observed no difference in IS among quartiles, our results indicate that insulin action in the cohort studied did not contribute significantly to the loss of skeletal muscle mass. However, we did not directly assess muscle protein metabolism using tracers. It cannot be totally excluded that despite an apparently equal sensitivity of glucose metabolism to insulin, the sensitivity of protein metabolism to insulin of individuals in each quartile of MMI was reduced in proportion to the differences in skeletal muscle mass. Unfortunately, we did not assess muscle protein metabolism and thus did not address this question.
Many factors play a role in the erosion of the skeletal muscle mass occurring during aging (1). A lack of protein consumption is one of them (30). In the present study, we observed that levels of protein intake, including the consumption of animal and vegetal protein, did not differ among quartiles of MMI, thereby ruling out the effects of this modulator. The level of physical activity also substantially affects skeletal muscle mass level. The subjects in the present study were sedentary and, therefore, this factor is unlikely to have played a role.
Our study has some strengths and weaknesses that merit discussion. The use of QUICKI limits our ability to interpret the findings. The particular question we were interested in could have been more adequately addressed if individuals would have undergone a glucose challenge in the form of hyperinsulinemic-euglycemic clamp, an OGTT or an intravenous glucose tolerance test (IVGTT). These tests (i.e., simulating a fed state) may detect abnormalities of glucose metabolism not possible to detect with QUICKI. The cross-sectional nature of the study precludes establishing causality. The small sample size may have provided insufficient power to detect differences in IS among groups. Despite this, the results are useful for meta-analysis purposes, as these types of analyses control for sample size. Also, on the positive side, this is the first study to examine the relation between IS and skeletal muscle mass among elderly women and, as a result, the results may incite further investigations on this topic.