Background
Identifying and reducing cardiometabolic risks driven by obesity remains a major healthcare challenge [
1]. Metabolic syndrome (MetS) is associated with an increased risk of cardiovascular disease, type 2 diabetes, and all-cause mortality [
2,
3]. Abdominal obesity is associated with low-grade inflammation and has been proposed as a driver for metabolic syndrome [
4]. Body weight loss may improve the factors of MetS [
5]; however, weight loss-induced improvements have proven difficult to maintain since substantial weight regain often occurs within the first year [
6]. Therefore, investigations of treatment strategies that can maintain, or even reduce, metabolic syndrome, abdominal fat, and low-grade inflammation in currently healthy persons with obesity to prevent future cardiometabolic disease are warranted [
7].
MetS denotes a cluster of common risk factors and was intended as an early measure for cardiometabolic disease risk [
8]. However, the dichotomous design of MetS has its limitations, and it is debated whether different definitions of MetS add predictive value when adjusted for its individual factors [
9,
10]. The newer metabolic syndrome severity z-score (MetS-Z) combines weighted contributions of all MetS factors into a single continuous measure [
11]. Studies have shown that individuals within the fourth quartile of MetS-Z scores (> 0.675) had a hazard ratio (HR) of 5.1 for coronary heart disease with more than 11 years of follow-up [
12,
13] and 17.4 for future diabetes with a median follow-up of 8 years compared to those from the first quartile of MetS-Z scores. However, MetS-Z has not been investigated in randomized clinical trials comparing treatments during weight loss maintenance in people at risk of future cardiometabolic disease.
High-sensitivity C-reactive protein (hsCRP) is an established biomarker of inflammation [
14] and is commonly elevated in persons with obesity [
15]. The relationship between hsCRP and the risk of cardiovascular disease is well documented; hsCRP levels of < 1 mg/L, 1–3 mg/L, or > 3 mg/L can be used to classify the risk of cardiovascular risk as low, intermediate, or high (in combination with traditional cardiovascular risk factors) [
14,
15].
Exercise and glucagon-like peptide–1 receptor agonists (GLP-1 RAs) may be different strategies in the primary and secondary prevention of MetS, abdominal fat, and inflammation [
7].
A meta-analysis has shown that moderate-to-vigorous aerobic exercise for at least 12 weeks can improve the factors of MetS [
16], and a study of self-reported physical activity has shown that exercise was associated with reduced inflammation markers in 10 years of follow-up [
17]. The potential anti-inflammatory effects of exercise might, in part, be due to reduced visceral fat independent of total body weight loss [
4,
18]. However, determining the effects of exercise interventions is often complicated by high study heterogeneity and, importantly, varying adherence to intervention protocols [
19]. Thus, studies that assess the effects of exercise adherent to intervention protocols are limited.
The GLP-1 RA, liraglutide, approved for obesity therapy, induces weight loss and improves glycemic control and cardiovascular risk factors (e.g., lipid profile and blood pressure) [
20,
21]. GLP-1 RAs are also suggested to lower inflammation due to direct anti-inflammatory effects on various tissues and immune cells and partly because of the weight loss seen with GLP-1 RA treatment [
22,
23].
We recently showed that a diet-induced 12% weight loss was maintained after one year with either exercise or liraglutide treatment. Combining the two treatments led to additional weight loss, while the placebo group regained body weight [
24]. In the present study, we investigated the effects of actually performed moderate-to-vigorous exercise, liraglutide 3.0 mg/day, or the combination of exercise and liraglutide on MetS-Z, abdominal obesity, and the inflammation marker hsCRP in a one-year maintenance period following a diet-induced weight loss.
Discussion
Identifying and managing the risk of cardiometabolic disease associated with obesity remains a major healthcare challenge. Metabolic syndrome, abdominal obesity, and low-grade inflammation constitute risk factors for future cardiometabolic disease. Therefore, we investigated improvements in metabolic syndrome, abdominal obesity, and low-grade inflammation during exercise, a glucagon-like peptide 1 receptor agonist, or the combination of the two following an eight-week low-calorie diet.
The diet-induced weight loss reduced MetS-Z, abdominal obesity, and inflammation marker hsCRP. After one year, the combination of exercise and liraglutide treatment reduced MetS-Z, android fat percentage, and hsCRP compared to placebo. Exercise treatment maintained MetS-Z and hsCRP and reduced android fat percentage compared to placebo. Liraglutide treatment reduced Mets-Z and android fat percentage while maintaining hsCRP compared to placebo. Placebo treatment was associated with maintenance of the diet-induced reductions in MetS-Z, hsCRP, and android fat percentage, even though 50% of the weight lost during the low-calorie diet was regained in the placebo group, while MetS prevalence and fat masses increased again. In addition, we have previously reported that the placebo group became sedentary one year after the initial weight loss [
28].
Large reductions in MetS-Z, abdominal obesity, and hsCRP compared to placebo were seen in the combination group, providing large potential reductions in cardiometabolic risk. Furthermore, the combination group showed a reduction of android fat percentage that was about twice as large as the reduction seen in both the exercise and liraglutide groups, underlining the complementary effects of combined treatment.
Liraglutide treatment alone further reduced MetS-Z following the diet-induced reductions, largely due to reduced fasting glucose, an expected effect of liraglutide treatment [
29]. Similar decreases in android fat percentage were seen with exercise compared to the liraglutide group, but exercise did not further reduce MetS. Exercise has been found to decrease MetS, but these studies did not include an initial diet-induced weight loss phase [
16,
30]. These findings suggest that, in already weight-reduced individuals, treatment with GLP-1 RA might be helpful in reducing cardiometabolic risk further.
If the comparative risks, illustrated in Fig.
1, indeed reflect the risks of the participants, the liraglutide groups substantially reduced the risk of future diabetes and coronary heart disease. Clinically, the MetS-Z model might prove valuable in guiding the primary prevention of cardiometabolic disease.
The combination group reduced body weight due to fat mass loss with a preferential reduction of android fat rather than gynoid fat. Thus, during the entire trial, abdominal obesity of participants in the combination group was reduced by almost 8%-points while maintaining total lean mass. This finding contrasts with other weight loss strategies, including the low-calorie diet used in this trial, which often lead to large amounts of lost lean mass (e.g., 20–50% lost by bariatric surgery, 30–47% by GLP-1 RA treatment before weight loss) [
31]. Furthermore, in men, only the combination treatment was able to lower android fat percentage.
The liraglutide group reduced android fat percentage without changing body weight, suggesting a reduction of android fat percentage independent of weight loss. This is consistent with recent findings from a 36-week study examining changes to visceral fat estimated by magnetic resonance imaging in response to treatment with liraglutide [
32]. Regarding hsCRP, we observed a significant reduction in the combination compared to the placebo group in the per-protocol analysis. In the intention-to-treat analysis, this reduction was no longer significantly different from the placebo group. This suggests that adherent exercise in combination with liraglutide might be able to add improvements to low-grade inflammation.
Despite similar reductions in android fat in the exercise and liraglutide group, exercise did not significantly decrease hsCRP after one year. However, studies that demonstrate reduced inflammation as an effect of exercise do often not have an initial weight loss phase [
17,
33], which substantially reduced hsCRP in this study (from a median of 3.8 to 2.4 mg/L); thus, the possibility for additional improvement through physical activity alone might have been limited.
In a clinical setting, hsCRP levels higher than 3 mg/L indicate increased cardiovascular risk [
14,
15]. At inclusion in this study, the mean hsCRP level was above the upper limit of hsCRP and approached the lower limit of 1 mg/L in the liraglutide treatment groups at the end of the trial. When liraglutide treatment was combined with adherent exercise, hsCRP was reduced by more than 50% during the entire trial. Therefore, these hsCRP findings indicate that combination treatment can exert clinically meaningful reductions in low-grade inflammation after diet-induced weight loss. Regarding insulin resistance, adherent exercise was able to maintain the large reductions in HOMA-IR induced by the low-calorie diet.
A strength of this study is the longitudinal, randomized, placebo-controlled design with four separate groups to assess single and combined effects of treatments with exercise and liraglutide 3.0 mg/day. Another strength is the novelty of analyzing the effects of maintained interventions on a clinically relevant continuous metabolic syndrome score combined with assessments of abdominal obesity and inflammation, translating to potential cardiometabolic risk.
In this study, we present the findings from the participants who completed the trial according to the prescribed interventions to better observe the effects of actually performed exercise, often confounded by inadequate adherence [
19]. The limitation of this approach includes a possible selection bias which might have skewed the treatment estimate. A reason for not fulfilling the high demands of the per-protocol requirements may be the time consumed on exercise, which is a known barrier to exercise [
34]. In the present study, the per-protocol participants in the exercise groups performed an average of 2.5 h of exercise per week for a whole year. Importantly, we also present the intention-to-treat analysis, including the 36 participants not fulfilling the high per-protocol demands, which generally painted a similar picture in the placebo, exercise, and liraglutide groups compared to the per-protocol analysis. Thus, except for hsCRP, which only showed significant differences between the placebo and combination groups in the per-protocol population, there were no differences in the results between intention-to-treat or per-protocol analyses.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.