Skeletal Muscle Phosphocreatine Recovery after Submaximal Exercise in Children and Young and Middle-Aged Adults

Abstract

Context: Elderly subjects have reduced mitochondrial function. However, it remains unclear whether the decline in mitochondrial function begins earlier in the life span.

Objective: The objective of the study was to determine skeletal muscle mitochondrial oxidative phosphorylation by ³¹phosphorous-magnetic resonance spectroscopy (MRS) across a variety of age groups.

Design: This was a cross-sectional study of 121 healthy normal-weight and overweight individuals from age 8 to 55 yr.

Setting: The study was conducted at a single university medical center in Boston, MA.

Participants: Participants included 68 children and 53 adults from the Boston community.

Interventions and Main Outcome Measures: Phosphocreatine (PCr) recovery was evaluated by ³¹phosphorous-MRS after submaximal exercise. Subjects were also evaluated with anthropometric measurements, metabolic profiles, and measures of physical activity.

Results: PCr recovery determined by ³¹phosphorous-MRS is positively associated with age in univariate analysis in a cohort of individuals aged 8–55 yr (r = +0.55, P < 0.0001). Stratification of subjects into four age groups (prepubertal and early pubertal children, pubertal and postpubertal children < 18 yr, young adults aged 18–39 yr, and middle aged adults aged 40–55 yr) demonstrates prolongation of PCr recovery with increasing age across the four groups (P < 0.0001 by ANOVA). The relationship between PCr recovery and age remains strong when controlling for gender; race; ethnicity; body mass index; measures of physical activity and inactivity; and anthropometric, nutritional, and metabolic parameters (P < 0.004).

Conclusions: Skeletal muscle PCr recovery measured by ³¹phosphorous-MRS is prolonged with age, even in children and young adults.

Skeletal muscle phosphocreatine recovery after submaximal exercise is increased with age even in children and young adults.

Mitochondria are subcellular organelles involved in basic energy homeostasis including carbohydrate and fatty acid metabolism. A role for altered mitochondrial function in aging has been hypothesized. Muscle strength and speed decline with aging (1). This decline is associated with a reduction in muscle fibers (2) as well as a reduction in the rate of oxidative metabolism within skeletal muscle ex vivo (3). In vivo, the rates of both mitochondrial oxidation and phosphorylation, as assessed by ¹³carbon and ³¹phosphorous nuclear magnetic resonance spectroscopy (MRS), are reduced by 40% in the sedentary elderly (aged 61–84 yr) compared with sedentary young adults (4). The decline in mitochondrial function with age is further supported by the increase in mitochondrial DNA mutations in subjects over the age of 50 yr compared with those under the age of 40 yr (5).

Because exposure to oxidizing agents increases in a continuous manner over time, we hypothesize that mitochondrial function is not reduced only in the elderly but rather is inversely related to age in a continuous manner, beginning in late childhood. In this study, we used ³¹phosphorous-MRS to assess phosphocreatine (PCr) recovery rate in skeletal muscle after submaximal exercise in subjects ranging in age from 8 to 55 yr to evaluate the relationship between age and PCr recovery rate from childhood to middle adulthood.

Subjects and Methods

Study subjects

Sixty-eight pediatric participants, aged 8–17 yr and 53 adult participants, aged 18–55 yr, were recruited from the Boston metropolitan area through advertisements in local media and physician referrals. The pubertal status (Tanner stage) of the pediatric population was determined by clinical examination by a pediatric endocrinologist (A.F.). All participants were otherwise healthy without known endocrine dysfunction. Participants were on no medications known to alter glucose metabolism, including hormonal therapies. In addition, participants had no known diagnosis of diabetes mellitus or medical conditions known to affect weight or glucose metabolism.

The current study was approved by the Partners, Massachusetts General Hospital, and Massachusetts Institute of Technology institutional review boards. Written informed consent or parental consent and participant assent were obtained from all participants. The investigation was conducted according to the principles of the Declaration of Helsinki.

³¹Phosphorous-MRS protocol

Mitochondrial function was evaluated using ³¹phosphorous-MRS to assess PCr resynthesis after exercise at the Massachusetts General Hospital Athinoula A. Martinos Center for Biomedical Imaging as previously reported (6). Subjects were placed in a 60-cm bore, Siemens 3.0 T Tim Trio system with an operating frequency of 49.879 MHz for ³¹phosphorous-spectroscopy. Scans were performed in the late afternoon and subjects were instructed to refrain from caffeine intake and vigorous physical activity during the afternoon before the study. All subjects were allowed to acclimate to the scanner and were evaluated for comfort and lack of pain limiting activity level before the initiation of the protocol. Maximum voluntary contraction (MVC) for the quadriceps muscles was measured using standard weights with the subjects in a bilateral leg-extension exercise apparatus before commencement of the exercise protocol. MVC was determined as the maximum amount of weight that could be lifted once. The MVC was used, as in prior studies (7), to estimate the exercise load to control for subject variability in strength and thus allow for sufficient PCr depletion during the 3-min exercise period in this diverse cohort of subjects. Depletion of PCr by the end of exercise was confirmed in all subjects with visual inspection of the PCr curve. The leg-extension exercise for MVC was identical to the leg extension for the submaximal exercise used to determine PCr resynthesis.

A baseline resting spectra was obtained by averaging spectra collected every 15 sec for 2 min. The baseline spectra were used to correct for partial saturation and to provide PCr concentrations based on the β-ATP peak. Subjects were then instructed to perform submaximal exercise for 3 min at a constant load by performing bilateral quadriceps contractions at 0.5 Hz (every 2 sec) followed by a 5-min recovery period. An 8-cm diameter radiofrequency surface coil, tuned for ³¹phosphorus, was fastened proximal to the superior aspect of the patella over the anteromedial aspect of the right thigh, keeping the coil’s axis perpendicular to the z-axis of the magnet, allowing for isolation of the quadriceps muscle. PCr recovery was determined by fitting the time course of PCr concentration during recovery after exercise with an exponential curve [β3 + β2 (1 − e^−t/τ))] to generate τPCr, which describes mitochondrial phosphorylation potential with software generated by M.I.H. (Mirtech, Inc., Brockton, MA). All scans demonstrated a reduction in muscle pH to less than 6.95 at the end of exercise and/or an end of exercise ratio of PCr to inorganic phosphate (Pi) of less than 4, factors indicating adequate participant effort.

Biochemical assessment

Oral glucose tolerance tests were performed using standard procedures with a 75-g glucose load. Glucose and lipid levels including triglycerides were determined using standard methodology in the Massachusetts Institute of Technology clinical laboratory. Serum insulin was measured using a RIA (Diagnostic Products Corp., Los Angeles, CA) or a chemiluminescence immunoassay (Access immunoassay system, Beckman Coulter, Chaska, MN). In same-sample comparisons, interassay correlation was excellent (r = 0.99), using identical linear scales, without systematic differences in the results of the assays by Bland Altman analysis (8).

Nutrition assessment

Calorie and macronutrient intake were assessed by 3- to 4-d food records, which were reviewed by nutrition staff. Analysis was performed with the Nutrition Data Systems (Minneapolis, MN). Midarm circumference was measured in all children as previously reported (9,10,11). Dual-energy x-ray absorptiometry scans were performed on Hologic 4500 densitometer (Hologic, Inc., Waltham, MA) to evaluate lean mass in all adult participants as previously reported (12).

Physical activity assessment

Self-reported levels of physical activity and inactivity in the children were assessed with pediatric-specific modifications made to the Modifiable Activity Questionnaire (13). The physical activity level of the adults was evaluated by use of an accelerometer with daily step count averaged over a period of 7 d. The MVC was used to calculate the exercise load and thus individualize the exercise protocol to allow for adequate work load and PCr depletion in a cohort diverse in age, size, and fitness level.

Statistical analysis

Continuous variables were tested for normality of distribution with the use of Wilk-Shapiro test and examination of the histogram distribution. Variables that were normally distributed were compared using the Student’s t test, and variables that were not normally distributed were compared using the nonparametric Wilcoxon rank sums test. Nominal variables were compared using the χ² test. Statistical significance across groups was compared using ANOVA. Univariate regression analysis was performed comparing τPCr with demographic and anthropometric variables as well as age using the Pearson correlation coefficient. Multivariate regression analysis with standard least squares modeling was also performed for τPCr. Variables were chosen for entry into multivariate modeling based on potential physiological interactions with the PCr recovery time constant. Independent models consisting of age, gender, race, ethnicity, body mass index (BMI), anthropometric, metabolic, dietary, and physical activity measures as well as MVC were evaluated. Statistical analysis was performed using JMP statistical database software (SAS Institute, Inc., Cary, NC). Statistical significance was determined as P < 0.05.

Results

Clinical characteristics of study subjects

One hundred twenty-one participants were recruited between the ages of 8 and 55 yr. Participants were 51% male and 49% female. Fifty-two percent of participants were Caucasian, 29% were African-American, and 17% were of Hispanic origin. Of the 68 pediatric participants (<18 yr of age), 33 participants were of normal weight at less than the 85th percentile for BMI based on age and sex, and 35 participants were overweight at the 85th percentile or greater. Of the 53 adult participants (≥18 yr of age), 20 were of normal weight (BMI < 30 kg/m²) and 33 were obese (BMI ≥ 30 kg/m²) (Table 1).

Table 1.

Baseline characteristics of study subjects

	All	Prepubertal children	Pubertal children	Young adults	Middle aged adults	P
n	121	23	45	22	31
Age (yr)	23.8 (14.7)	10.2 (1.4)	14.3 (2.0)	25.9 (6.7)	46.3 (4.7)	<0.0001
Gender (male/female)	62/59	15/8	24/21	7/15	16/15	0.15
Race						0.94
Caucasian	63 (52%)	13 (57%)	23 (51%)	12 (55%)	15 (48%)
Non-Caucasian	58 (48%)	10 (44%)	22 (49%)	10 (46%)	16 (52%)
Ethnicity						0.61
Hispanic	20 (17%)	4 (17%)	8 (18%)	5 (23%)	3 (10%)
Non-Hispanic	101 (84%)	19 (83%)	37 (82%)	17 (77%)	28 (90%)
Anthropometric parameters
BMI (kg/m2)	27.3 (7.9)	22.3 (6.7)	26.0 (7.5)	31.6 (7.9)	30.0 (6.7)	<0.0001
Normal weight/overweighta	53/68	11/12	22/23	7/15	13/18	0.57
Waist to hip ratio	0.88 (0.09)	0.90 (0.08)	0.86 (0.08)	0.88 (0.08)	0.91 (0.11)	0.08
Metabolic parameters
Total cholesterol (mg/dl)	162 (35)	164 (32)	146 (28)	174 (42)	175 (33)	0.0006
Triglycerides (mg/dl)	82 (59)	69 (45)	76 (52)	95 (61)	92 (73)	0.35
HDL cholesterol (mg/dl)	48 (11)	52 (12)	43 (8)	47 (10)	52 (13)	0.002
LDL cholesterol (mg/dl)	100 (29)	99 (28)	90 (23)	111 (35)	106 (30)	0.02
Fasting blood glucose (mg/dl)	81.0 (15.2)	77.5 (5.4)	77.5 (7.4)	84.8 (30.7)	85.9 (9.4)	0.04
Fasting insulin (μIU/ml)	8.5 (8.5)	6.2 (4.4)	11.5 (11.7)	10.0 (7)	5.0 (3.4)	0.004
HOMA-IR	1.8 (2.2)	1.2 (0.9)	2.3 (2.4)	2.5 (3.2)	1.1 (0.9)	0.03
Nutritional parameters
Caloric intake (kcal/d)	1927 (606)	1728 (335)	1919 (526)	1945 (645)	2064 (807)	0.38
Daily caloric intake/kg	30.2 (14.3)	43.2 (16.4)	30.9 (12.6)	24.1 (12.3)	24.9 (9.3)	<0.001
Carbohydrates (%)	51 (8)	51 (7)	52 (7)	50 (8)	49 (9)	0.55
Protein (%)	17 (4)	16 (3)	17 (4)	18 (4)	18 (4)	0.28
Fat (%)	33 (6)	34 (6)	33 (5)	33 (5)	34 (7)	0.90
Saturated fatty acids (%)	11 (3)	12 (3)	12 (2)	11 (2)	11 (3)	0.20
Trans-fatty acids (%)	2 (1)	2 (1)	2 (1)	2 (1)	2 (1)	0.43
Glycemic load	200 (76)	179.6 (44)	207.9 (74.9)	206.5 (103)	196 (81)	0.62
Physical activity parameters
Hours of vigorous activity perweek		3.6 (2.5)	3.8 (2.0)			0.82
Accelorometer (×1000 steps/d)				8.0 (2.9)	11.9 (5.5)	0.12
MRS parameters
τPCr (sec)	41.2 (19.9)	27.4 (11.7)	36.1 (13.2)	48.0 (23.7)	54.1 (20.9)	<0.0001
MVC (lb)	83.8 (40.4)	42.6 (21.6)	81.7 (21.9)	105.2 (11.5)	105.0 (38.3)	<0.0001

Prepubertal group includes prepubertal and early pubertal children. Pubertal group includes pubertal and postpubertal children less than 18 yr old. Young adult group includes subjects 18–39 yr old and middle-aged group includes subjects 40–55 yr old. The data are presented as mean (sd). For determination of significance for race, ethnicity, and weight status, a χ² was performed. For conversion from milligrams per deciliter to millimoles per liter for total, HDL, and LDL cholesterol, multiply by 0.0259, and to convert from milligrams per deciliter to millimoles per liter for triglycerides, multiply by 0.0113. For conversion from milligrams per deciliter to millimoles per liter for glucose, multiply by 0.0555. For conversion from microinternational units per milliliter to picomoles per liter for insulin, multiply by 6.945. For conversion from pounds to kilograms, multiply by 0.454. HDL, High-density lipoprotein; LDL, low-density lipoprotein. 

^aNormal weight and overweight were determined by BMI for age and gender less than 85th percentile or 85th percentile or greater for participants less than 18 yr old and by BMI of less than 30 or 30 kg/m² or greater for participants 18 yr old or older. 

The participants were divided into age categories as prepubertal and early pubertal children, Tanner stage 1 and 2 (group 1); pubertal and postpubertal children, Tanner stages 3–5 (group 2); young adults, aged 18–39 yr of age (group 3); and middle-aged adults, aged 40–55 yr of age (group 4). The characteristics of the four groups are presented in Table 1. Relative differences in total caloric intake per day were observed across the age groups, but these differences were not statistically significant. However, total caloric intake corrected for body weight differed among the groups as anticipated and consistent with the recommended daily allowance (14). Macronutrient intake including carbohydrates, protein, or fat (total fat, saturated fat, or trans-fatty acids) expressed as percentage of energy intake did not differ significantly across groups. Results of physical activity assessments are shown and are not different among the groups of adults and children (Table 1).

Relationship of τPCr to anthropometric and metabolic parameters

τPCr was not associated with anthropometric measurements including BMI (r = +0.06; P = 0.53) or waist circumference (r = +0.15; P = 0.09) in our cohort. τPCr was associated with fasting glucose (r = +0.18; P = 0.05) but was not associated with homeostasis model assessment insulin resistance index (HOMA-IR; r = +0.03; P = 0.75). τPCr was not associated with timing in the menstrual cycle of our adult female subjects (r = −0.26; P = 0.26).

τPCr was associated with MVC (r = +0.34; P = 0.0002); however, this relationship was no longer significant after controlling for age in multivariate modeling. Conversely, the relationship between τPCr and age remained significant when MVC was added to the multivariate model (see below).

Relationship of τPCr to age

τPCr was positively associated with increasing age by univariate analysis (r = +0.55, P < 0.0001; Fig. 1A). When the participants were divided into their respective age categories, τPCr increased across groups (P < 0.0001 by overall ANOVA; Fig. 1B) as well as between all groups (P < 0.05 between all groups by Student’s t test; Fig. 1B).

Figure 1.

A, τPCr is positively associated with age on univariate analysis (R = +0.55; P < 0.0001) among all subjects (n = 121). B, τPCr is higher across increasing age groups. Subjects were divided into prepubertal/early pubertal children (n = 23); pubertal/postpubertal adolescents less than 18 yr old (n = 45); young adults 18–39 yr old (n = 22); and middle-aged adults 40–55 yr old (n = 31). Groups are significantly different by overall ANOVA (P < 0.0001). All groups are significantly different from each other by Student’s t test (P < 0.05).

Multivariate regression analysis of the relationship between τPCr and age

Multivariate regression analyses of the relationship between τPCr and age of the participants was performed controlling for demographic, anthropometric, physical activity, dietary, and metabolic parameters. Age was robustly and significantly associated with τPCr after controlling for gender, race, ethnicity, and BMI (β = 0.79; P < 0.0001; R² for overall model: 0.32; P for overall model: < 0.0001). Furthermore, age continued to have the strongest association with τPCr when weight category (<85th percentage or ≥85th percentage BMI for children and <30 or ≥30 kg/m² for adults) was used instead of absolute BMI (β = 0.75; P < 0.0001; R² for overall model: 0.32; P for overall model: <0.0001). Controlling for physical activity or MVC did not significantly affect the primary model and age continued to have the strongest association with τPCr (P < 0.004). The relationship between age and τPCr remained highly significant controlling for weight, fasting glucose, insulin, HOMA-IR, and dietary intake, including total calories or fat (total or saturated fat) intake. The relationship between τPCr and age remained significant in the pediatric group when controlling for midarm circumference as a measure of lean body mass and in the adult group when controlling for lean body mass by dual-energy x-ray absorptiometry.

Discussion

This is the first study, to our knowledge, to investigate mitochondrial function in vivo in a large group of children and young and middle-aged adults and to demonstrate an age-related decline in skeletal muscle PCr recovery that begins in childhood. In our study, age was robustly associated with τPCr on univariate analysis and furthermore remained significantly associated with τPCr after controlling for gender, race, ethnicity, BMI, physical activity, dietary intake, lean body mass, and metabolic parameters in multivariate modeling. In addition, stratification of the subjects into increasing age groups demonstrated a progressive prolongation of τPCr across the age groups.

An age-related decline in ATP production and content in skeletal muscle has previously been demonstrated in rodent studies (15,16). Human studies have reported a reduced mitochondrial ATP production rate in healthy, lean sedentary elderly individuals compared with BMI- and activity level-matched young adults (4). However, the cause and tempo of mitochondrial decline in humans has not been elucidated. Our data are the first to demonstrate a progressive decline in skeletal muscle τPCr starting in childhood and continuing through middle adulthood. Unlike the decline in lean body mass and bone mineral density (17) or increase in visceral adiposity (18,19), which occurs after early adulthood, our data demonstrate a reduction in skeletal muscle PCr recovery that occurs on a continuum from childhood through early adulthood and into later adult years.

Evidence is accumulating regarding the association of altered mitochondrial function and the development of insulin resistance and type 2 diabetes mellitus, even among subjects without known mitochondrial disorders (20,21). Therefore, age-related comorbidities, including insulin resistance, may be associated with declining mitochondrial function with aging.

There is debate as to whether the reduction in mitochondrial oxidative capacity, previously reported in elderly individuals, is related to the aging process itself or is secondary to alterations in body composition and fitness level associated with aging. Prior studies demonstrated that difference in BMI may not be associated with difference in skeletal muscle mitochondrial function (21,22), and this is supported in our own study of 121 normal-weight and obese subjects. Furthermore, our data demonstrate that the association between age and τPCr remained significant, even after controlling for BMI and lean body mass as well as measures of physical activity and a measure of the exercise load achieved during testing. These data suggest that the effect of age on mitochondrial function may, at least partially, be independent of the changes in anthropometric and fitness level associated with aging. This is consistent with recent data demonstrating that neuronal mitochondrial function declines with aging by ¹³C/¹H-MRS and labeled glucose and acetate (23) and may suggest that there is a physiological decline in mitochondrial function in multiple organs with aging, independent of physical activity and fitness.

The cross-sectional nature of this study limits our ability to assess causality. Nonetheless, we phenotyped a large number of subjects using ³¹phosphorous-MRS and saw a robust relationship between age and τPCr. To our knowledge, this is the largest cohort of subjects in whom PCr recovery has been determined in vivo. Second, we did not confirm mitochondrial function of our subjects ex vivo because muscle biopsies to assess mitochondrial function were not feasible in this large cohort. However, ³¹phosphorous-MRS is a well-established technique to assess skeletal mitochondrial function in vivo (24), and our specific technique to measure PCr recovery after submaximal exercise in skeletal muscle has previously been validated by corroboration with mitochondrial DNA content (6). MVC was associated with τPCr, but this was a result of the anticipated strong association of MVC with age. Importantly, the association between τPCr and age remained highly significant controlling for MVC. In addition, oral glucose tolerance testing was performed for the evaluation of glucose metabolism because hyperinsulinemic clamps were not feasible in this large, young cohort.

In conclusion, τPCr progressively increases with age, beginning as early as late childhood and continuing through early adulthood into middle age. Our data are consistent with, but significantly extend, prior data showing reduced mitochondrial function in the elderly, by demonstrating that τPCr increases with age, starting in childhood. Future studies will be necessary to corroborate the current findings, including longitudinal studies beginning in childhood. In addition, further studies are necessary to determine the pathophysiological significance of increasing τPCr that develops even among the young. Moreover, studies are needed to evaluate whether pharmacological and lifestyle interventions to improve mitochondrial function may improve or even prevent age-associated diseases such as insulin resistance and cardiovascular diseases.