Eleven-year cat study maps metabolic shifts with aging

An 11-year longitudinal study of 21 research-colony cats found that aging was associated with both cumulative and nonlinear changes in blood and fecal metabolism, including shifts in energy handling, lipid composition and inflammatory markers. The cats were sampled annually from one year of age through 11 years, allowing researchers to compare young adult, mature adult and early senior stages within the same animals. The results identified multiple age-related trajectories rather than a single uniform pattern, with especially clear changes emerging around six to eight years of age.

The animals were adult domestic shorthairs housed at Massey University’s Centre for Feline Nutrition in New Zealand. Nine were male and 12 female, and 19 were neutered. Eleven cats received a commercially available complete and balanced diet throughout the sampling period, while the other 10 participated in earlier prospective studies and received either extruded or retorted diets; diet format and participation in other nutritional experiments therefore varied among the animals. Water was available ad libitum. Plasma and fecal samples were collected at the end of each November, and blood sampling followed an estimated fast of 12 to 18 hours. The researchers measured serum biochemistry alongside metabolomics and lipidomics, using gas chromatography-tandem mass spectrometry for targeted metabolites and liquid chromatography-quadrupole time-of-flight mass spectrometry for lipids.

After quality control, the analysis retained 64 plasma metabolomic variables, 278 plasma lipid variables, 66 fecal metabolomic variables, 401 fecal lipid variables and 23 serum biochemical markers. Repeated-measures analysis of variance with simultaneous component analysis was used to isolate age-related patterns while accounting for sex, diet and weight. The method combines mixed-effects modeling with principal component analysis, allowing linear and curved trajectories to be distinguished while reducing the influence of unrelated variation. Individual metabolites and lipids were then examined with linear mixed models, with false-discovery-rate correction used to limit false positives across the many measured features.

The most prominent combined pattern was a plateau during the first five years, a sharp decline around year six and continued change through year 11. Fecal features contributed most strongly to this overall trajectory, with 25 of the leading features coming from the fecal lipidome and eight from fecal metabolomics. Some datasets followed an approximately linear course, consistent with cumulative alteration, while others formed U-shaped or inverse-U-shaped curves. In plasma, the first age-related component followed a U-shaped pattern and explained 41.7% of the age-associated variation. Lactic acid, aspartic acid, xanthine, hypoxanthine, uracil, phosphoric acid and 2-hydroxyisobutyric acid were higher toward the beginning and end of life but lower during midlife. Glucose and fructose showed the inverse pattern, with higher concentrations in middle age. Nine of the 10 leading plasma metabolites retained significant associations after outlier removal and multiple-testing correction; glutamic acid did not.

Plasma lipids followed a different course, remaining comparatively stable from years one through seven before shifting downward from year eight. The leading component accounted for 40.9% of age-associated variance in this dataset. Declining phosphatidylcholines and phosphatidylethanolamines drove much of the pattern, while some longer-chain phosphatidylcholines increased. Desmosterol rose nearly quadratically across the 11 years. Most of the 10 leading plasma lipids were strongly associated with age at p values below 10^-6 before adjustment, although removal of outliers weakened the associations for two quadratic models. The investigators interpreted the changing phospholipid profile as consistent with possible alterations in fatty-acid synthesis, beta-oxidation and membrane lipid composition, but noted that direct measurements of hepatic or mitochondrial fat oxidation would be required to test that explanation.

Several findings were compatible with altered energy metabolism and possible mitochondrial dysfunction, without establishing either as a cause of aging. Lactic acid was higher and glucose lower at older ages than during the middle years, a pattern the authors said may be consistent with a greater reliance on anaerobic glycolysis. Two hydroxybutyric acids and 2-hydroxyisobutyric acid also followed nonlinear age-related patterns. At the same time, fecal coenzyme Q8, Q9, Q10 and Q11 had inverse-U-shaped trajectories, with higher concentrations between approximately years two and nine. Coenzyme Q10 is involved in mitochondrial energy production and antioxidant processes, but the fecal measurements do not directly demonstrate mitochondrial injury. Differences in fasting duration may also have influenced circulating metabolites, and the investigators cautioned against treating the observed associations as direct evidence of altered mitochondrial function.

Evidence for an inflammatory trajectory came from serum proteins and fecal metabolites. The albumin-to-globulin ratio declined almost linearly with age, accompanied by lower albumin and higher globulin concentrations. The ratio reflects the balance between liver-derived albumin and a broader group of globulins that includes acute-phase proteins and immunoglobulins, so the authors regarded the pattern as supportive of the inflammaging hypothesis. The first component of the serum analysis explained 62.6% of age-related variation and declined strongly over time. Fecal histamine and several N-acylethanolamines also increased in older cats. N-acylethanolamines participate in inflammatory, endocannabinoid and appetite-related signaling, while gut microorganisms can produce histamine. These observations support an association between age and inflammatory or immune-related metabolic change, but they do not show which process came first.

The fecal metabolome followed a stepped decline, with many compounds high during the first five years and lower from years six through 11. The leading component explained 51.5% of age-related variance. Purine and pyrimidine derivatives, including thymine, hypoxanthine, xanthine and uracil, decreased, as did several compounds related to amino-acid metabolism. Tryptophan metabolism showed contrasting patterns across the two sample types: plasma kynurenine increased with age in the combined analysis, while fecal anthranilic acid and 3-hydroxyanthranilic acid declined. Histamine and 5-aminovaleric acid, a potential product of microbial lysine catabolism, increased. The researchers proposed that these differences might reflect changing interactions among the host, diet and gut microbial metabolism, but the study did not measure microbial composition or microbial function directly and therefore could not establish a causal role for dysbiosis.

Fecal lipids followed a U-shaped trajectory. Sphingomyelins and lysophospholipids had higher values at younger and older ages, while coenzyme Q derivatives and bile acids formed inverse-U-shaped patterns. Five bile acids associated with age were linked to the fecal pattern of the primary feline bile acid taurochenodeoxycholic acid, and no secondary bile acids were associated with age. Two triglycerides included among the leading features increased in year 11, but their associations with age were not significant after correction, and their residual distributions showed substantial statistical problems. The authors noted that sphingomyelins and lysophospholipids have been associated with intestinal inflammation in cats with chronic enteropathy, while altered bile-acid excretion could be consistent with age-related changes in fat digestion or oxidation. Because fecal compounds can originate from diet, microbiota, intestinal secretions and shed epithelial cells, their biological source cannot be assigned from concentration measurements alone.

The study’s principal strength was its repeated follow-up of the same cats over more than a decade, with both plasma and feces sampled and multiple omics platforms used alongside routine chemistry. Its limitations constrain how broadly the findings can be applied. Only 21 healthy research cats were included, making this a small and relatively controlled cohort whose results may not represent the wider domestic-cat population. Diet was not identical for every animal, activity and food intake were not measured, and the microbiome was not characterized. Follow-up ended at 11 years, close to the beginning of the senior life stage, so the investigators could not determine whether the changes after six to eight years continued, accelerated or plateaued. Cognitive and physical-function assessments were also absent. The results therefore identify candidate biological markers and mechanisms for further testing, rather than providing a validated test for biological age or proof that an early intervention could alter the trajectories.

Subject of Research: Longitudinal metabolomic and lipidomic changes associated with aging in 21 research-colony domestic cats

Article Title: Metabolomic and lipidomic signatures of aging in cats reveal shifts in energy metabolism and inflammation: an 11-year longitudinal study

Article References: Hartvigsson, O., Ross, A. B., Moesta, A., Heiser, A., Lee, E., Butowski, C. F., Thomas, D. G., Laxalde, J., Bermingham, E. N., & Moniot, D. (2026). Metabolomic and lipidomic signatures of aging in cats reveal shifts in energy metabolism and inflammation: an 11-year longitudinal study. Frontiers in Veterinary Science, 13, Article 1953163. https://doi.org/10.3389/fvets.2026.1953163

Image Credits: AI Generated

DOI: 10.3389/fvets.2026.1953163

Keywords: feline aging, longitudinal study, metabolomics, lipidomics, energy metabolism, inflammaging, mitochondrial dysfunction, gut microbiome, tryptophan metabolism, domestic cats, Metabolomic, lipidomic