Abstract
This study analyzed the type and frequency of physical activity that most effectively reduces the homeostatic model assessment of insulin resistance (HOMA2-IR) among adults (≥19 years) in Asia. We used national representative data from 1645 men and 2272 women who participated in the Korea National Health and Nutrition Examination Survey in 2015 were included in the analysis. The effects of different types and frequencies of physical activity on HOMA2-IR were investigated using a multiple regression analysis. Compared with no activity, moderate-to-vigorous physical activity (MVPA) ≥5 times per week (β: −0.214, p≤0.0198) and walking and MVPA ≥5 times per week (β: −0.183, p≤0.0049) were negatively associated with HOMA2-IR. In the subgroup analysis, the strongest effect was observed among overweight men. Additionally, walking plus MVPA ≥5 times per week had the strongest effect on men with a higher-than-recommended daily calorie intake (β: −0.350, p≤0.0030). Therefore, in conclusion, the appropriate type and frequency of physical activity can help reduce HOMA2-IR in South Korean men, especially those who are overweight and/or have a higher-than-recommended daily calorie intake.
Significance of this study
What is already known about this subject?
The homeostatic model assessment of insulin resistance (HOMA2-IR) is convenient to assess IR.
Both abdominal obesity and IR associated strongly with type 2 diabetes are commonly associated with physical activity (PA).
PA has been associated with reductions in the risks of mortality and morbidity associated with chronic diseases, prevention of obesity, cardiovascular disease, cancer, and enhanced insulin sensitivity.
What are the new findings?
In this study, walking plus moderate-to-vigorous physical activity (MVPA) ≥5 times per week was associated with the lowest HOMA2-IR levels, followed by MVPA ≥5 times per week; by contrast, generally walking alone did not affect the HOMA2-IR.
Among men with overweight body mass index (BMI), walking plus MVPA, and only walking were also associated with lower HOMA2-IR.
Walking plus MVPA ≥5 times per week and MVPA ≥5 times per week were effective.
A reduction in the HOMA2-IR could be better achieved by exercising the most appropriate type and frequency of PA according to the subject’s individual characteristics.
How might these results change the focus of research or clinical practice?
Our study can suggest the most appropriate type and frequency of PA for reducing HOMA2-IR, and to make particularly specific inferences and recommendations for individuals with an overweight BMI. Therefore, it can help prevent the IR through PA.
Introduction
Diabetes is an important public health issue worldwide.1 Considering the financial effects of diabetes, numerous countries are taking steps to comply with the goals of the WHO Non-Communicable Disease Global Action Plan 2013–2020.2
Most cases of diabetes are considered type 2, the risk of which is closely related to insulin resistance (IR).3–5 Currently, many IR measurement methods are available. Although the gold standard euglycemic-hyperinsulinemic clamp test and intravenous glucose tolerance test are among the most typical methods for assessing IR,6 7 they are relatively time consuming and costly.8 By contrast, the homeostatic model assessment of IR (HOMA2-IR) is much more convenient because it can be used when only 2 fasting glucose and fasting insulin concentrations are available.9
Abdominal obesity has a major influence on IR,10 such that even those with normal weight exhibited an increase in IR if they had developed abdominal obesity, which make the carotid artery’s endogenous membrane dreadful.11 In brief, abdominal obesity directly affects IR. Notably, both of these variables associated strongly with type 2 diabetes are commonly associated with physical activity (PA).12 13 In previous studies, PA has been associated with reductions in the risks of mortality and morbidity associated with chronic diseases, as well as the prevention of obesity,14 15 cardiovascular disease,16 cancer,17 18 and enhanced insulin sensitivity.19 Conversely, insufficient PA can lead to an increased IR.20 Despite the obvious links between PA and chronic disease prevention, few studies have investigated the association between PA and HOMA2-IR. Therefore, this study aimed to examine the effects of the type and frequency of PA on IR, using data from the Korea National Health and Nutrition Examination Survey (KNHANES). Specifically, we evaluated the influences of type and frequency of PA on IR and differences in the body mass index (BMI) and daily caloric intake. Furthermore, this study attempted to suggest the type and frequency of PA that could most effectively reduce the individual HOMA2-IR.
Methods
Study population
This study used data from the KNHANES, which has been conducted by the Korean Centers for Disease Control and Prevention to assess the health and nutritional status of the Korean population since 1998. This national, cross-sectional, population-based survey evaluates the health and nutritional statuses of non-institutionalized Korean civilians using health interviews and examinations, as well as nutrition surveys. Survey stratification is based on 16 metropolitan cities and provinces, as well as administrative divisions and dwelling units.
The 2015 KNHANES baseline data sets included 3381 and 3999 individuals, respectively. Among the 7380 total eligible respondents, subjects younger than 19 years of age and those with missing data for variables used in this study were excluded. Finally, 3917 subjects, including 1645 men and 2272 women, were included in this study.
Measures
Dependent variable
This study aimed to investigate differences in the HOMA2-IR, a simple and useful indicator of IR, according to the type and frequency of PA among South Korean adults. The HOMA2-IR was the dependent variable. Although many methods for quantifying the measured IR are available, the HOMA2-IR is considered accurate and is widely used in clinical studies.21 In this study, we used the HOMA calculator version 2.2.3 (available from http://www.dtu.ox.ac.uk).22 This method can be easily calculated based on measured fasting plasma blood glucose (3.5–25.0 mmol/L) and fasting plasma insulin (20–400 pmol/L) concentrations.
HOMA2-IR, an indicator of IR, is relatively simple and also provides accurate values. It has been used in many clinical studies and is a reliable measure. Hence, it is believed that the use of HOMA2-IR is still being updated.
Variables of interest
In this study, the independent variable was PA, which was classified by type plus frequency for analysis. The PA criteria included exercise during leisure time but excepted PA during work. The PA type was subcategorized into walking and moderate-to-vigorous physical activity (MVPA). MVPA included vigorous intensity, moderate intensity, strength, and flexibility exercises, all of which were used to examine the effect of PA on the HOMA2-IR. With respect to exercise frequency, the subjects were also divided into 3 groups: none, ≥5 times per week, and <5 times per week. We combined these 2 variables of PA type plus frequency, which was categorized as follows: no activity, walking <5 times, walking ≥5 times, MVPA <5 times, MVPA ≥5 times, walking and MVPA <5 times, and walking and MVPA ≥5 times.
Covariates
The characteristics of the individual study subjects were stratified by sex and included in this study as covariates. The individual characteristics included age (19–29, 30–39, 40–49, 50–59, 60–69, and ≥70 years), educational level (elementary school graduate or below, middle school graduate, high school graduate, and college graduate or higher), marital status (single, once married, and married), income status (low, mid-low, mid-high, and high), job (white collar, pink collar, blue collar, and unemployed), smoking (current, former, and never), alcohol intake (yes or no), BMI (average, overweight, and obese), diabetes (yes or no), daily calorie intake (less than recommended, recommended, and more than recommended), and family history of chronic diseases (yes or no). The subjects were classified into 3 groups by BMI—average (<23 kg/m2), overweight (23–24.9 kg/m2), and obese (≥25 kg/m2)—according to the WHO criteria for adults.23 24 The subjects were also classified into 3 groups by daily caloric intake—less than recommended, recommended, and more than recommended—in alignment with the criteria of the National Research Council in the USA.25
Statistical analysis
The general characteristics of the study subjects were analyzed using t-test and analysis of variance with regard to comparisons of the differences in median HOMA2-IR. The HOMA2-IR values were log transformed for parametric tests. A multiple regression analysis was used to analyze the associations between the HOMA2-IR and the type and frequency of PA. Additionally, a sex-stratified analysis was performed to determine whether different patterns exist between men and women. Statistical analyses were conducted using the SAS software, V.9.4 (SAS Institute). All p values in this study were two sided and considered significant at <0.05.
Results
The 2015 baseline general characteristics of the study subjects are shown in table 1 according to sex, along with the relationship between HOMA2-IR and each of the covariates. The 1645 male subjects reported the following: 153 (9.3%), no activity; 180 (10.9%), walking <5 times per week; 179 (10.9%), walking ≥5 times per week; 65 (4.0%), MVPA <5 times per week; 83 (5.0%), MVPA ≥5 times per week; 258 (15.7%), walking plus MVPA <5 times per week; and 727 (44.2%), walking plus MVPA ≥5 times per week. The corresponding median (IQR) HOMA2-IR scores were 1.015 (0.601–1.477), 0.917 (0.607–1.456), 0.933 (0.569–1.403), 0.913 (0.649–1.447), 0.799 (0.551–1.244), 0.890 (0.601–1.276) and 0.835 (0.559–1.261), respectively. The 2272 female subjects reported the following: 200 (8.8%), no activity; 320 (14.1%), walking <5 times per week; 295 (13.0%), walking ≥5 times per week; 98 (4.3%), MVPA <5 times per week; 73 (3.2%), MVPA ≥5 times per week; 374 (16.5%), walking plus MVPA <5 times per week; and 912 (40.1%), walking plus MVPA ≥5 times per week. The corresponding median (IQR) HOMA2-IR scores were 0.927 (0.572–1.363), 0.932 (0.627–1.425), 0.834 (0.611–1.311), 0.826 (0.500–1.222), 0.922 (0.577–1.346), 0.845 (0.565–1.211) and 0.814 (0.583–1.205), respectively. In addition, the dependent variable of this study, HOMA2-IR, was calculated by levels of triglyceride (TG) and high-density lipoprotein cholesterol (HDL-C). The results of the regression analysis revealed that among men, both TG (β: 0.07, p≤0.0001) and HDL-C (β: −0.90, p≤0.0001) are associated with IR with statistical significance. Among women, both TG (β: 0.18, p≤0.0001) and HDL-C (β: −0.64, p≤0.0001) again showed a statistically significant relationship with IR.
Table 2 describes the relationships of HOMA2-IR with the combined type plus frequency. Among men, the log HOMA2-IR decreased when individuals performed MVPA ≥5 times per week (β: −0.214, p≤0.0198) and walking plus MVPA ≥5 times per week (β: −0.183, p≤0.0049). No statistically significant results were observed among women. Regarding BMI, both men with overweight (β: 0.323, p≤0.0001) and obese BMI (β: 0.763, p≤0.0001) exhibited increased log HOMA2-IR values than men with average BMI. Women with overweight (β: 0.328, p≤0.0001) and obese BMI (β: 0.662, p≤0.0001) also exhibited increased log HOMA2-IR values than women with average BMI. When diabetes was considered, statistically significant results were observed in both men (β: 0.304, p≤0.0001) and women (β: 0.318, p≤0.0001) with regard to HOMA2-IR.
Table 3 presents the associations between PA and the log HOMA2-IR by BMI and daily calorie intake. Men with an overweight BMI generally exhibited significant results with respect to the combined type plus frequency of PA. When the PA type and frequency were evaluated concurrently, the highest scores were obtained in the group that reported walking plus MVPA ≥5 times per week (β: −0.425, p≤0.0006), followed by those who performed walking ≥5 times per week (β: −0.376, p≤0.0105), MVPA ≥5 times per week (β: −0.368, p≤0.0318) group, walking plus MVPA <5 times per week (β: −0.310, p≤0.0334), and walking <5 times per week (β: −0.243, p≤0.0385). Men with an obese BMI also exhibited significant associations of HOMA2-IR with PA type plus frequency, but only in the MVPA ≥5 times per week (β: −0.294, p≤0.0250) group. Among women, by contrast, statistically significant results were only observed among those with an average BMI who reported walking plus MVPA ≥5 times per week (β: −0.135, p≤0.0238).
We further observed that men who consumed more than the recommended daily calorie intake exhibited significant associations in the analysis. Highest scores were found in those in the walking plus MVPA ≥5 times per week (β: −0.269, p≤0.0468) group, followed by the walking plus MVPA ≥5 times per week (β: −0.350, p≤0.0030) group. Among women, no significant results were found.
Discussion
This study analyzed the associations of the HOMA2-IR with the type and frequency of leisure-time PA using data from a nationally representative sample of Korean adults. The results revealed that in men, the combined type and frequency of PA correlate with negative changes in HOMA2-IR. Specifically, the performance of MVPA ≥5 times per week and walking plus MVPA ≥5 times per week reduced HOMA2-IR levels.
In the subgroup analysis, significant associations were found between HOMA2-IR and the combined type and frequency of PA among men with overweight BMI. Specifically, HOMA2-IR was most reduced by walking plus MVPA ≥5 times per week, followed by walking ≥5 times per week and MVPA ≥5 times per week. Among men with obese BMI, the performance of MVPA ≥5 times per week group had reduced HOMA2-IR level. In addition, among men who consumed more than the recommended daily calorie intake, walking plus MVPA ≥5 times per week most effectively reduced HOMA2-IR, followed by MVPA ≥5 times per week. By contrast, only 1 variable, walking plus MVPA ≥5 times, affected HOMA2-IR of women in the average BMI group.
Our study results are consistent with those of previous studies, in which participation in overall and vigorous PA was found to associate with significantly high insulin sensitivity.19 Additionally, in a prospective cohort study examining the association between regular vigorous exercise and the subsequent incidence of non-insulin-dependent diabetes mellitus, PA was identified as a probable preventive method.12 Another study also found that MVPA affects the prevention of type 2 diabetes.26
One particular focus of this study was the potential influence of walking, one of the lightest and most basic physical activities, on IR, given that this form of exercise has been identified as one of the most effective in terms of a wide range of health benefits.27 28 Notably, previous studies have reported that walking may reduce the incidence of premature death,29 chronic disease,27 while brisk walking was shown to improve insulin sensitivity.26 Moreover, another study found that low-intensity exercise, including walking, effectively reduces HOMA-IR; however, the study focused solely on patients with diabetes and their treatment, with variables including other types of low-intensity PA and hospital diet.30 As previous studies of walking have yielded various results, further investigation is needed to clarify the associations between various walking strengths and HOMA-IR.
In this study, MVPA ≥5 times per week was associated with the lowest HOMA2-IR levels, followed by walking plus MVPA ≥5 times per week; by contrast, generally walking alone did not affect HOMA2-IR. A statistically significant result was found only in men with overweight BMI. Although these results are consistent with those of previous studies reporting steady exercise as an effective means of reducing IR,19 they also present MVPA, rather than walking, as an associative factor with HOMA-IR. However, among subjects with a particularly high (ie, overweight) BMI, walking was also associated with a lower HOMA2-IR. In general, although walking is known to be beneficial to health,27 28 it generally did not improve IR in this study.
Our results allow us to suggest the most effective PA type and frequency for reducing HOMA2-IR, and to make particularly specific inferences and recommendations for individuals with an overweight BMI. Appropriate PA is known to have various beneficial health effects, and certain PAs affect the HOMA2-IR. Therefore, a reduction in the HOMA2-IR could be better achieved by exercising the most appropriate type and frequency of PA according to the subject’s individual characteristics. However, we note that statistically significant associations were only observed among male subjects in this study and further studies are needed to investigate relevant associations among women. In conclusion, the findings of this study suggest that PA is negatively associated with the HOMA2-IR.
This research had some limitations. For example, the PA type and frequency categories were not stratified into more detailed subcategories. With respect to PA type, this study aimed to investigate whether walking alone could reduce IR; therefore, all the exercises except walking were combined into a single group. In addition, we attempted to investigate the duration and other frequently used variables for a more detailed classification of PA, such as information on heart rate, oxygen uptake. Due to data limitations, however, the correlation of HOMA2-IR with such variables could not be determined. Moreover, this study featured a cross-sectional design; therefore, we could not exclude a potential bidirectional effect.
Conclusion
Among Korean adults, and especially among men, participating in MVPA was associated with a reduction in the HOMA2-IR, regardless whether walking was practiced. Additionally, among men with an overweight BMI and those who consumed above the recommended daily calorie intake, walking plus MVPA ≥5 times per week appeared to be most effective for reducing the HOMA2-IR. These findings infer the potential benefits of performing PA in a manner that could effectively reduce the HOMA2-IR.
Footnotes
Contributors JK and ECP designed the study, collected the data, performed the statistical analysis, and drafted the manuscript. JK, ECP, WK, SAL, and DC contributed to the discussion and reviewed and edited the manuscript. ECP is the guarantor of this work, and as such had full access to all study data. ECP assumes responsibility for the integrity of the data and the accuracy of the data analysis.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Patient consent for publication Not required.