The pleiotropic effects of SGLT-2 inhibitors and GLP-1 agonists hold the potential to target cardiorenal, hepatic and metabolic disorders using a disease model that replicates key features of MetS. Previous studies have primarily focused on obesity-related metabolic dysfunction when examining the effects of WDs in dogs. However, there is a lack of comprehensive studies on the biological and metabolic impacts of WDs independent of obesity. This is significant as multiple reports, including analyses by Romero-Corral et al. (2010) and Shi et al. (2020) on large adult populations, have established that individuals with normal weight MetS are at an increased risk of cardiovascular diseases, metabolic dysregulation, and higher mortality rates compared to both obese individuals with and without MetS. In addition, most clinical investigations on the effectiveness of dapagliflozin and empagliflozin for cardiovascular and renal outcomes primarily include non-obese subjects (McMurray et al., 2019; Wheeler et al. 2020; Butler et al., 2021; Oyama et al., 2022; EMPA-KIDNEY Collaborative Group, 2023). Collectively, these findings provide a compelling rationale for establishing a novel model of metabolic dysfunction that operates independently of obesity. Once established, this disease model will pave the way for in-depth studies on the pharmacodynamics of therapeutic drugs targeting cardiorenal metabolic health, including SGLT-2 inhibitors and GLP-1 agonists.
Our study maintained isocaloric conditions to isolate the effect of the diet’s composition from obesity as a confounding factor. It builds on preliminary data from Lyu et al. (2022), which showed a tendency towards elevated glucose concentrations in ten healthy Beagles under an isocaloric high-fat diet for six weeks. To the best of our knowledge, our research represents the first comprehensive characterization of the biological effects of a WD model, independent of obesity. By inducing MetS without causing weight gain, we have successfully developed a non-invasive, inducible, and potentially reversible preclinical model in just a few weeks. For ethical reasons and considerations related to animal welfare, it is important to emphasize that our objective was not to induce clinical symptoms of MetS in our study. Therefore, most of the observed changes reported herein remained within physiological limits. Overall, the WD was well tolerated with no reported adverse events during the study. Minor digestive issues appeared when transitioning from a regular diet to the WD, likely due to the absence of a proper weaning period between diets. However, these issues were resolved within a few days.
Hematological parameters consistently remained within normal physiological limits, showing no clinically meaningful changes. The most notable variations were observed in metabolic parameters. Specifically, the WD induced a statistically significant increase in fasting blood glucose levels, approaching the upper physiological limit. This resulted in an overall increase of approximately 20% in blood glucose concentrations compared to baseline. Interestingly, this observation was accompanied by a trend towards an increase in peak insulin concentrations, which is consistent with previous literature in dogs (Tvarijonaviciute et al., 2012b; Moinard et al., 2020) and indicative of a perturbation in insulin sensitivity that ultimately leads to insulin resistance (Cavaghan et al., 2000). The decrease in circulating glucagon concentrations may be symptomatic of a physiological feedback mechanism to maintain glucose homeostasis in response to increased fasting blood glucose (Rix et al., 2019).
Our dietary intervention also resulted in significant changes to serum chemistry parameters. These fluctuations, although still within physiological limits, demonstrate the ability of our model to greatly influence metabolism and homeostasis. Specifically, we observed a decrease in serum bicarbonate levels, which is in line with low-grade metabolic acidosis (Burger and Schaller, 2023). This is important because a recent meta-analysis, which included data from over 30,000 patients, found an association between MetS, lower bicarbonate, and a higher risk of metabolic acidosis (Lambert et al., 2023). Concurrently, there was a measurable increase in chloride serum concentrations, which may be attributed to hyperchloremic acidosis (Sharma et al., 2023) and/or the onset of MetS (Kimura et al., 2016). In addition, the WD induced marked reductions in both phosphorus and potassium levels, both of which have been linked to an increased risk for MetS (Kalaitzidis et al., 2005; Stoian and Stoica, 2014; Sun et al., 2014).
Consistent with the definition of MetS by the National Heart, Lung, and Blood Institute (NHLBI), our diet induced a significant elevation of SBP by approximately 10 mmHg. The increase in SBP appeared to be independent of any notable changes in renin and angiotensin peptides. Interestingly, SBP was not found to increase in a previous canine study focusing on obesity-related cardiac dysfunction and MetS (Tropf et al., 2017), again supporting our rationale for studying the effect of western diets independently of obesity. Our study also found mild increases in NT-proBNP, although mostly within the reference range. We suspect that the increase in circulating natriuretic peptides occurred secondarily to the increase in SBP, as previously reported in the literature (Hussain et al., 2022; Jang et al., 2023), but it could also be indicative of cardiac stress (Bayes-Genis et al., 2023). Notably, some dogs showed NT-proBNP concentrations exceeding 900 pmol/L, a threshold commonly associated with structural heart disease in canines (Singletary et al., 2012; Wilshaw et al., 2021).
Total cholesterol increased by approximately 45% after ten weeks. Importantly, in line with the definition of MetS, dogs fed the isocaloric WD model experienced a significant reduction in HDL-cholesterol, along with an increase of LDL-cholesterol (of around 25%). These shifts occurred independently of any corresponding alterations in serum triglyceride concentrations. While surprising, this finding is consistent with earlier research from Lahm Cardoso et al. (2016) which showed a strong correlation between body condition scores (BCS) and triglyceride levels in dogs, with values approaching the upper limit of 200 mg/dL in dogs with a BCS of 8 or above (classified as "overweight" or "obese" in our study).
Our results on redox status align with previous human studies (Matsuzawa-Nagata et al., 2008; Boden et al., 2017; Aleksandrova et al., 2021). Specifically, we observed significant increases in TOS and d-ROMs at BAS2. In contrast, the effect on antioxidant markers was more nuanced and generally mild, with levels of CUPRAC, FRAP, TEAC, and Thiol remaining stable at BAS2. This is in line with the variable impact of dietary fat on systemic antioxidative stress markers in dogs. Some studies have shown no effect of carbohydrate and fat concentrations on oxidative stress biomarkers (Chiofalo et al., 2020), while others have reported an increase in antioxidant capacity, but no effect on oxidative stress markers (Vecchiato et al., 2023).
Our study highlights the comprehensive changes in the metabolome induced by the WD, including biological pathways related to general metabolism, complex lipids, and biogenic amines. These observations underscore the potential relevance of this model in studying MetS and its associated health complications. Notably, all the metabolites detected in our study were classified according to MSI Level 2 standards (Sumner et al., 2007). The positive correlation of nicotinamide to the baseline diet (BAS1) in both general metabolism (urine) and biogenic amines (urine) suggests that dogs had lower amounts of this essential form of vitamin B3 after ten weeks of feeding with a WD (BAS2) compared to their standard diet. This is consistent with earlier findings from Qu et al. (2022) reporting perturbations in nicotinamide metabolism in dogs fed a high-fat diet for twelve weeks. Nicotinamide plays a crucial role in various metabolic pathways, particularly in energy production and DNA repair (Surjana et al., 2010; Amjad et al., 2021). Similarly, the positive correlation of glycerol to BAS2 in general metabolomics (urine) indicates that glycerol concentrations were increased during feeding with the WD. Glycerol is a key component of triglycerides and is involved in energy metabolism, especially in lipid breakdown and synthesis (Frühbeck et al., 2014). This elevation is likely related to an increased metabolism of triglycerides caused by the WD, indicating a potential shift in lipid metabolism. The positive correlation of tartaric acid (2,3-dihydrobutanoic acid) with BAS2 in multiple classes (general metabolomics in urine, biogenic amines in urine, and biogenic amines in serum) indicates that tartaric acid levels increased during the WD phase. These changes are possibly associated with the increased catabolism of the antioxidant ascorbic acid and accompany variations in oxidative stress markers highlighted above (Bánhegyi et al., 2004).
A greater diversity of fatty acids was correlated with BAS1, especially in stool, indicating a wider range of fatty acid profiles in the baseline diet. This diversity is essential for energy production and cell membrane structure (Hishikawa et al., 2014). We suspect that the higher diversity of fatty acids in BAS1 is due in part to the presence of more soluble dietary fiber, which is fermented by intestinal microbiota into short chain fatty acids (Sivaprakasam et al., 2016). After the WD diet, there was an increase in saturated fatty acids, specifically FA 17:0 in stool lipidomics and PC 18:0. High levels of saturated fatty acids have been linked to negative health outcomes, such as cardiovascular disease (Siri-Tarino et al., 2010; Hooper et al., 2020). Furthermore, palmitoleic acid, an omega-7 monounsaturated fatty acid, showed a positive correlation with BAS2 in stool general metabolomics. Palmitoleic acid (16:1n7) increases lipolysis, glucose uptake, and glucose utilization for energy production in white adipose cells (Bolsoni-Lopes et al., 2014; Cruz et al., 2018). The increase in saturated fatty acids, specifically FA 17:0 in BAS2, could suggest a shift in lipid metabolism and/or be a result of the consumption of butter and ruminant fats (Alves et al., 2006). While there is a decrease in fatty acid diversity and an increase in saturated fatty acids associated with BAS2, it remains unclear whether these changes are related to MetS per se, or simply changes in stool composition caused by differences in diet composition.
In both general metabolomics (GC-MS) and LC-MS assays, several unidentified metabolites were detected. For GC-MS, this was due to spectral library matches failing to identify metabolites below the 800 threshold. Advanced data processing techniques, such as Parallel Factor Analysis, could be employed to deconvolve data and obtain cleaner spectra (Amigo et al., 2008; Giebelhaus et al., 2022a). However, this would require a separate and dedicated study. Additionally, the bioamines assay detected several non-amine compounds due to its ability to detect compounds without an amine group. With LC-MS, the presence of unidentified metabolites could possibly be attributed to biotransformation of known metabolites, which involves the addition or removal of specific chemical moieties such as (de)-glycosylation, (de)-methylation, (de)-amination, and (de)-hydroxylation. These transformations often occur during metabolic processes (Giebelhaus et al., 2022b). To identify these metabolites, biotransformation analysis techniques and exploration of additional libraries and databases would be necessary. However, this is beyond the scope of this study.
This study presents several limitations that are worth mentioning. First, the experiment was limited in size and did not address the potential for reversibility of the model, specifically regarding the metabolic impacts of transitioning back to a standard diet. Due to the non-invasive nature of our model, this study was primarily descriptive and did not provide in-depth mechanistic insights into how WDs affect cardiovascular-kidney-metabolic health. The study also lacks some functional data, such as the time-dependent effects of the WD on intestinal permeability and fecal microbiome composition. This was partly deliberate, as those effects have been extensively characterized previously in the literature (e.g., Moinard et al., 2020). Previous studies published by our consortium in dogs (Ward et al., 2021; Ward et al., 2022; Sotillo et al., 2023; Schneider et al., 2023) were conducted in the context of an activated RAAS. No elevation in PRA-S or angiotensin peptides was noted in the current study, which likely explains why many data points on aldosterone were below the lower limit of quantification, preventing statistical analysis of this biomarker. Additionally, our findings related to serum, urine and fecal metabolomics could be confounded by factors other than changes in diet composition, including differences in processing between a commercial and a home-cooked diet. Determining whether metabolites are changing in response to altered metabolism or simply due to changes in dietary composition using untargeted metabolomics can be challenging. Therefore, mechanism-based hypothesis testing on the metabolome would require a targeted or semi-targeted approach that focuses on specific classes of metabolites.
In summary, our isocaloric WD, designed to mimic the NHANES diet, effectively replicated key characteristics of MetS. These included elevated BP, increased fasting glucose levels, and reduced HDL-cholesterol, all independent of abdominal obesity. Additionally, the WD induced an increase in natriuretic peptide levels, along with a mild state of metabolic acidosis and significant changes in the serum, urine, and fecal metabolome. Our findings underscore the utility of this model for investigating the metabolic effects of novel antidiabetic therapies in the context of obesity-independent MetS, which will be presented in a separate manuscript. Furthermore, this research enables future translational studies that could have potential benefits for both human and veterinary medicine.