Association of Questionnaire-Based Physical Activity Analysis and Body Composition Dynamics in Type 2 Diabetes: A Cross-Sectional Study

Introduction

The composition of the body primarily consists of bone, muscle, and fat, with a balanced proportion being the fundamental prerequisite for maintaining optimal health. Due to the effects of abnormal glucose metabolism in T2D patients, bone metabolism and lipid metabolism are altered, and the body composition is out of balance.^1,2 In T2D, in the presence of insulin resistance, skeletal muscle metabolic dysfunction leads to an increase in protein catabolism and a decrease in synthesis, causing a decrease in muscle strength and a decrease in muscle mass, which in turn leads to the development of sarcopenia.^3,4 Studies have shown that diabetic osteoporosis (DOP) can be comorbid in approximately 50% of diabetic patients,⁵ which increases the risk of falls and fractures.⁶ Seventy-two percent of the T2D population has high abdominal fat deposition, and hypertension, dyslipidemia, and insulin resistance are the main factors that cause abdominal visceral adipose tissue (VAT) accumulation.^7,8 T2DM patients with abdominal obesity exhibit more severe lipid metabolism disorders, and the degree of insulin resistance and dyslipidemia becomes increasingly pronounced as the body mass index rises.⁹

Previous studies have shown that insufficient physical activity and sedentary behavior are the main factors leading to obesity and insulin resistance in T2D.^10,11 Insufficient long-term Physical Activity (PA) has a direct impact on bone mineral content and skeletal muscle mass,^12,13 and is closely related to the morbidity and mortality of T2D.¹⁴ Studies on physical activity and body composition in T2D patients have received much attention, but whether there is a quantitative relationship between physical activity and body composition in T2D patients has yet to be investigated. Therefore, this study investigated whether there is a relationship between different levels of physical activity and changes in body composition.

Materials and methods

Participants

T2D patients hospitalized from June 2020 to July 2022 in the Department of Endocrinology and Metabolism of a tertiary hospital in Nantong city were sampled via convenience sampling.

The inclusion criteria for individuals were as follows: (1) met the World Health Organization (WHO) 1999 diagnostic criteria;¹⁵ (2) aged 25–75 years; (3) had diabetes for a duration of ≥6 months; (4) had a glucose-lowering regimen unchanged for 3 months prior to enrollment; and (5) voluntarily participated in this study and maintained the same basic lifestyle for 6 months.

Exclusion criteria: (1) Type 1 diabetes, gestational diabetes, and other special type; (2) history of metabolic diseases such as hyperthyroidism, hypothyroidism, and rheumatoid arthritis; (3) combination of serious diabetic complications, such as diabetic ketoacidosis, diabetic hyperosmolar coma; (4) the presence of long-term glucocorticoids and sex hormones; (5) patients with pacemakers or metal plates installed in the body.¹⁶ This study was reviewed and approved by the Ethics Committee of the Second Affiliated Hospital of Nantong University in accordance with the Declaration of Helsinki. In addition, all participants provided informed consent when recruited for the study.

Data Collection

The International Physical Activity Questionnaire (IPAQ-S) was used in this study to assess physical activity levels and sedentary behavior, which was developed by the International Consensus Group on Physical Activity Measurement in 2001,¹⁷ with 7 items and 4 dimensions, 3 categories of physical activity, namely, high intensity and moderate intensity, and low intensity, and 3 categories of physical activity and sedentary behavior. The questionnaire asked patients to review the number of days they had been physically active in the past 7 days, the length of activity per day, and the length of sedentary time per day; only activities and sedentary time over 10 min were considered relevant physical activities and sedentary time. The metabolic equivalent value of high-intensity physical activity was 8.0 MET, moderate-intensity physical activity was 4.0 MET, and low-intensity physical activity was 3.3 MET.¹⁸ Physical activity levels are reported as medians and quartiles. Individuals’ weekly level of physical activity of a certain intensity was calculated as follows: metabolic equivalent (MET), corresponding to physical activity × weekly frequency (d/w) × time per day (min/d); total physical activity level (MET-min/w), high-intensity physical activity level + moderate-intensity physical activity level + low-intensity physical activity level. The patients were grouped according to physical activity level: low, moderate, or high.

The diabetes treatment regimens in this study included insulin injection therapy, oral insulinotropic agents, insulin sensitizers, bisphosphonates, α-glucosidase inhibitors, DDP-4 inhibitors, SGLT-2 inhibitors, GLP-1 receptor agonists, and lifestyle intervention alone. Lifestyle alone was defined as only dietary control or physical activity in the last three months without receiving any diabetes medication.

In this study, the body composition of T2D patients was determined using DXA from Hologic Discovery Wi (S/N 86856), which was introduced into the Department of Nuclear Medicine of our hospital. The body composition measurements included lumbar spine (L1~L4) bone mineral density (LSBMD), femoral neck bone mineral density (FNBMD), hip bone mineral density (H-BMD), and total body mineral density (T-BMD); trunk bone mineral salt content, lean tissue mass, and fat mass; total body bone mineral salt content, lean tissue mass, and fat mass; and extremity skeletal muscle mass (ASM). In this study, trunk fat percentage (%) = trunk fat mass (g)/ [trunk bone mineral salt mass (g) + trunk lean tissue mass (g) + trunk fat mass (g)]; whole body fat percentage (%) = whole body fat mass (g)/ [whole body bone mineral salt mass (g) + whole body lean tissue mass (g) + whole body fat mass (g); and extremity skeletal muscle mass index (ASMI) were used to assess limb skeletal muscle mass, ie, ASMI = limb skeletal muscle mass (kg)/height² (m²).

For the collection of blood samples, elbow venous blood was drawn from each patient by the ward nurse early the following morning for the determination of biochemical markers after an overnight fast of 8 hours. Fasting insulin levels (FINS, two-site sandwich immunoassay with direct chemiluminescence) were measured using a fully automated chemiluminescence immunoassay analyzer (UniCel DxI800, Beckman Coulter) with automated detection steps; fasting glucose levels (FPG, oxidase assay), triglyceride levels (TGS, colorimetric assay), total cholesterol levels (TC, cholesterol oxidase assay), LDL cholesterol, and LDL cholesterol were determined by the ward nurse. Total cholesterol (TC, cholesterol oxidase method), low-density lipoprotein cholesterol (LDLC, selective melting method), and high-density lipoprotein cholesterol (HDLC, enzyme-modified method) were measured with an automated biochemical instrument (Model 7600, Hitachi, Ltd). The serum N-terminal osteocalcin (N-MID), β-collagen specific sequence (β-CTX), and total type I precollagen amino-terminal extended peptide (TP1NP) were measured via electrochemiluminescence immunoassay; the levels of 25-hydroxyvitamin D and parathyroid hormone (PTH) were measured via liquid chromatography-tandem mass spectrometry; and glycosylated hemoglobin (HbA1c) was measured via ion exchange high-performance liquid chromatography. All biochemical indices were measured by professional physicians in the Department of Medical Laboratory of our hospital. In this study, the insulin resistance index and pancreatic β-cell secretion function were assessed by the Homeostasis Model Assessment (HOMA):¹⁹ insulin resistance index (HOMA-IR) = FPG × FINS/22.5; and the pancreatic β-cell secretion index (HOMA-β) = 20 × FINS/(FPG-3.5).

Statistical Analysis

The general data, biochemical indices and body composition parameters of T2D patients were analyzed according to physical activity level (H-PA, M-PA, and L-PA); one-way ANOVA was used for normally distributed data; the χ2 test was used for count data, which are presented as the mean ± standard deviation () and frequency or constitutive ratio; and the use of a scatter plot to demonstrate the differences in body composition parameters among patients with different physical activity levels and body composition parameters among patients with different levels of physical activity.

A multiple linear regression model was further developed to explore the differences in the mean values of the body composition parameters (LSBMD, FNBMD, HBMD, TBMD, ASMI, B-FAT, and T-FAT) between the M-PA and L-PA subgroups by stepwise correction for other clinical confounders (B [95% CI]), with the H-PA group serving as the reference.

The data were statistically analyzed using SPSS 22.0 and GraphPad Prism 9.0.2; The alpha values were taken as 0.05 as the test level, and the p values were all two-sided probabilities, with p<0.05 indicating a statistically significant difference.

Results

Clinical Characteristics of Patients

Table 1 categorizes 615 patients with T2D into three groups according to their level of physical activity: H-PA, M-PA, and L-PA; 173 (28.2%), 232 (37.7%), and 210 (34.1%) of these patients, respectively. Differences among the three groups were statistically significant in terms of the number of females, age, duration of diabetes mellitus, history of previous falls, regular consumption of seafood, consumption of strong tea and coffee, and administration of alpha-glucosidase inhibitors, SGLT-2 inhibitors, and GLP-1 receptor agonists (p<0.05); however, no differences existed in the other clinical fundamentals (p>0.05).

Table 1 Comparison of Basic Data Among Subgroups of Physical Activity Level

Table 2 shows the differences in biochemical indices and body composition parameters among the three subgroups after grouping by physical activity. The β-CTX, TP1NP, HbA1c, B-FAT and T-FAT were significantly greater (p<0.05), and the 25(OH)D, LSBMD, FNBMD, HBMD, TBMD, and ASMI were significantly lower (p<0.001) with decreasing physical activity. However, in terms of lipids, lnHOMA-IR and lnHOMA-β did not significantly differ (p>0.05). In contrast, Figure 1 presents the comparison of bone mineral density, muscle, and fat parameters across tertiles of physical activity levels in Table 2 as a scatter plot.

Table 2 Comparison of Biochemical Indexes and Body Composition Parameters of Physical Activity Level Among Subgroups

Figure 1 Scatter plots of body composition indices in three subgroups with high level of physical activity(H-PA), moderate level of physical activity(M-PA) and Walking. (a) lumbar spine bone mineral density (LSBMD); (b) femoral neck bone mineral density (FNBMD); (c) hip bone mineral density (H-BMD); (d) total bone mineral density (T-BMD); (e) appendicular skeletal muscle index (ASMI); (f) body fat pertencage (B-FAT); (g) trunk fat pertencage (T-FAT).

Table 3 shows that the total physical activity level of 2708 (2033, 3674) MET-min/w was significantly greater in males than in females (p<0.001); however, for moderate-intensity physical activity, the physical activity level was significantly greater in females than in males (p=0.026). Similarly, the average sedentary time per day in the total population was 6.76 ± 2.08 h, with females spending more time sedentary per day than males (t = −3.557, p<0.001). Similarly, in terms of body composition parameters, both bone density and muscle content were significantly greater in men than in women, while the opposite was true for fat content (p<0.001).

Table 3 Comparison of Physical Activity Level, Sedentary Time and Body Composition Parameters in Different Sex Patients

Table 4 categorizes 615 type 2 diabetic patients according to age group criteria. Specifically, high-intensity and moderate-intensity physical activity, as well as overall activity levels, were greater in T2D patients aged 45–59 years than in patients in other age groups (p<0.05); in terms of body composition parameters, bone mineral density progressively decreased with patient age (p<0.01), whereas there was no significant difference between the B-FAT and T-FAT groups (p> 0.05).

Table 4 Comparison of Physical Activity Level, Sedentary Time and Body Composition Parameters in Different Age Groups

Mean Differences in Body Composition Parameters Among the Three Physical Activity Groups

Tables 5 and 6 compare the mean differences in BMD, muscle and fat mass between M-PA and L-PA patients by stepwise correction for confounders, using H-PA as a reference (B [95% CI]). After correcting for patients’ general information, lifestyle information, biochemical indices, and sedentary duration, the mean differences in LSBMD, FNBMD, HBMD, TBMD, ASMI, B-FAT, and T-FAT between the H-PA and L-PA groups were −0.098 (−0.136 to −0.061) g/cm², −0.048 (−0.075 to −0.021) g/cm², −0.063 (−0.090 to −0.035) g/cm², −0.035 (−0.060 to −0.009) g/cm², −0.178 (−0.327 to −0.029) kg/m², 1.487 (0.587 to 2.386)%, and 1.905 (0.785 to 3.026)%, respectively. Physical activity levels were independently and positively correlated with LSBMD, FNBMD, HBMD, TBMD, and the ASMI and independently and negatively correlated with the B-FAT and T-FAT in T2D patients.

Table 5 Mean Difference in Bone Mineral Density Between Subgroups at the Physical Activity Level (B[95% CI])

Table 6 Mean Differences in Physical Activity Levels of Skeletal Muscle and Fat Between Subgroups (B[95% CI])

Discussion

In this study, the total physical activity level of the 615 study subjects was 2531 (1872, 3308) MET-minutes/week. Among the different levels of physical activity, high, medium and low accounted for 28.2%, 37.7% and 34.1%, respectively, of the total number of study subjects. Patients with T2D are mostly characterized by a low level of physical activity, and there is insufficient physical activity.²⁰

Studies have proven that physical inactivity is a risk factor for lower bone density. A lack of physical activity leads to a reduction in direct stimulation of bone tissue, muscle pulling, and muscle vibration, which reduces bone loading and affects bone strength and bone biomechanical properties.²¹ Primary osteoporosis markers (β-CTX and TP1NP) as well as HbA1c were significantly greater (p<0.05), and secondary osteoporosis markers 25(OH)D and BMD parameters (LSBMD, FNBMD, HBMD, and TBMD) were significantly lower (p<0.001) with decreasing physical activity levels in this study; even after adjusting for confounders, the relationship remained after adjustment for confounders. Studies have confirmed that various forms of physical activity has positive effect on big rotor and femoral neck bone mineral density, the biggest benefit of lumbar spine.²² Studies have shown that as type 2 diabetic patients age, inflammatory factor and oxidative stress products increase in the body, prolong the duration of the disease, leading to insufficient insulin secretion, decreased synthesis of osteoblast nucleosides and collagen, and decreased calcium content in the bones.²³ In the present study, the bone mineral density decreased significantly (p<0.001) as the age of the study subjects increased. More than one-third of them were elderly people for whom walking was an acceptable exercise, had a low risk of accidents, and could be practiced in a variety of settings throughout the year, as opposed to high-intensity or moderate-intensity physical activity. Walking or jogging to achieve high enough mechanical stress to create a counterforce that stimulates the bones is effective in maintaining BMD levels.²⁴ In this study, women had significantly lower BMD levels than men did (p<0.001). For perimenopausal and postmenopausal women, a dramatic decrease in estrogen levels in the body affects osteoblast proliferation and collagen synthesis, predisposing them to high-transformation osteoporosis.²⁵ In women older than 50 years of age in the United States, 30 minutes of low-intensity physical activity was associated with a 12% lower risk of lumbar spine osteoporosis than was sedentary activity.²⁶ Therefore, it is recommended that older women with T2D be encouraged to engage in multiple forms of low-intensity physical activity to ensure that their BMD increases or at least remains unchanged.²⁷

With the progression of T2D, the body’s skeletal muscle function and muscle mass gradually decrease, and the prevalence of sarcopenia is much greater than that in the nondiabetic population.²⁸ Studies have shown that stimulating skeletal muscle through exercise can effectively improve muscle mass and strength and is an effective way to treat sarcopenia.^29,30 The skeletal muscle content of the extremities decreased significantly (p<0.001) as the level of physical activity decreased in this study. After correcting for age, sex, and sedentary duration, physical activity level remained an independent influence on limb skeletal muscle mass (95% CI: −0.327 to −0.029). Exercise of various intensities has been proven to improve myocyte mass and muscle protein synthesis; high-intensity resistance exercise has more prominent advantages in improving muscle mass and muscle function^31,32 and is currently the most effective intervention for the treatment of diabetic sarcopenia.³³ Liao³⁴ conducted a 12-week intervention study in which elastic band resistance training was used in elderly women with muscle loss. An intervention study revealed an increase in total skeletal muscle mass and significant beneficial effects on muscle mass and physical function in the intervention group. Resistance exercise in patients with diabetes mellitus combined with sarcopenia was effective at increasing skeletal muscle type II fibers and reducing skeletal muscle cell apoptosis.³¹ Additionally, aerobic exercise is an effective intervention for improving skeletal muscle mass in the body. In a Japanese study, 42 patients with T2D were treated with aerobic exercise (aerobics) twice a day, and the rate of decrease in skeletal muscle mass index in the intervention group (46.7%) was significantly lower than that in the control group (85.2%) compared with that in the control group that did not perform the exercise, which effectively prevented the decrease in skeletal muscle mass.³⁵

The prevalence and growth rate of overweight and obesity in China are the highest in the world.³⁶ In this study, the average BMI of the 615 T2D patients was 25.61±3.67 kg/m², the average waist circumference was 91.02±9.82 cm, the body fat content reached 30.99±6.67%, and the trunk fat content was also high. Overall, these patients reached the level of overweight or even obesity, reflecting the general phenomenon of overweight or obesity in T2D patients. Insufficient physical activity reduces the body’s energy consumption, which increases lipid deposition in the body and decreases insulin secretion. Exercise, on the other hand, can improve insulin secretion by decreasing adipokines and thus improving insulin secretion.³⁷ High-intensity physical activity promotes rapid glucose uptake and improves insulin resistance; long-term exercise training improves mitochondrial function and increases the expression of metabolic genes such as glucose transporter proteins.³⁸ In this study, the insulin resistance indices of the study subjects were high, and one-way ANOVA showed that as physical activity decreased, body fat content and trunk fat content increased significantly, while glycated hemoglobin (HbA1c) levels increased significantly (p<0.001); after correcting for various confounders, a decrease in physical activity was found to be an independent risk factor. An increasing number of studies have confirmed that physical activity is one of the most important measures for treating abnormalities in lipid metabolism. Various physical activities increase the body’s energy expenditure and reduce body weight, fasting insulin levels, and triglyceride (TG) synthesis in the liver.³⁹ A meta-analysis that included 48 studies indicated that among 2990 participants with at least three metabolic syndromes, the group that performed moderate-intensity aerobic exercise training showed significant improvement in lipid metabolism; reductions in TG, TC, and LDLC; and an increase in HDLC.⁴⁰ In the present study, none of the differences in lipid metabolism-related indices (TG, TC, HDLC, LDLC) were significantly different among the physical activity subgroups (p>0.05). This may be related to the use of GLP-1 receptor agonists in diabetic patients. Multiple studies have shown that GLP-1-receptor agonists promote weight loss. A network meta-analysis comparing the efficacy of different drugs for weight loss on the basis of 143 RCTS included suggests that GLP-1 receptor agonists can effectively promote weight loss of more than 5% on the basis of lifestyle management, which is superior to orlistat, SGLT2i, and metformin.⁴¹ It also improves blood sugar and lowers blood pressure and total cholesterol levels.⁴² But the TG concentration was greater than the normal value (2.45±1.89) mmol/L, which may be related to the dietary tastes of the participants in the study and dietary combinations.

Limitations

In this study, the population evaluated was representative of patients with type 2 diabetes, which limits the generalizability of the study to some extent. This study is only a cross-sectional study, and long-term follow-up can be further conducted in the future to pay attention to the dynamic change process of population composition.

Conclusions

Physical inactivity is an independent influence on body composition changes in patients with T2D; the less physically active and the more sedentary the person is, the lower the bone density, the less skeletal muscle content, and the more fat content. In the future, intervention studies can be further carried out for patients with type 2 diabetes mellitus, such as insufficient physical activity and sedentary lifestyle, to guide screening and control the occurrence of secondary complications such as osteoporosis, sarcopenia and obesity in patients with type 2 diabetes mellitus, and to improve the quality of life.