Abstract Background Heart failure (HF) with preserved ejection fraction (HFpEF) is a major public health problem worldwide, especially among the elderly. It is a complex syndrome characterised by multiple underlying comorbidities whose specific contribution is virtually impossible to disentangle in humans. As a result, treatment for HFpEF is still largely based on a "one-size-fits-all" approach, ignoring the specific pathophysiological substrate in each patient. Purpose we sought to identify the specific contribution of each HFpEF comorbidity to the development of this complex syndrome and to identify its earliest determinants and underlying mechanisms. Methods We used mouse models of ageing, obesity, hypertension, sleep apnoea, and hyperglycaemia to study the natural course of the syndrome and investigate the main determinants of HFpEF in each group separately and collectively. We followed the mice for up to 2.5 years and generated 100,000 individual data points (without omics), including echocardiography, vascular echography, lung ultrasound, histology and gene expression analyses in right and left ventricle, lung, liver and kidney. We developed new ad hoc statistical models to analyse all the longitudinal data. To identify the parameters that best determine the risk of HFpEF we used two complementary approaches: we first investigated features that could discriminate between mice with and without HFpEF and next, we investigated the changes associated with increased risk of HFpEF (time-to-HFpEF) among those animals that developed the syndrome. In addition, we used mass spectrometry proteomics to identify changes in protein expression and post-translational modification associated to HFpEF in each comorbidity. Results Age was the major contributor to HFpEF. Animals with HFpEF showed an increase in proteins involved in glycolysis, glycogen catabolism, and glutathione metabolism compared to those that did not develop HF, suggesting a change in metabolic substrate. The main parameters associated with HFpEF at old age started changing already at middle age. All comorbidities contributed to HFpEF through routes that were partially common but that were also characterised by distinct features. Changes in the right ventricle and in large vessels were strongly associated with the development of HFpEF in some comorbidities. Neither myocardial inflammation nor fibrosis were associated with HFpEF. Female mice followed a rather common route to HFpEF, including an association with kidney fibrosis and myocardial hypertrophy, while the development of the syndrome in mice was more comorbidity-determined. Conclusions Mice with different comorbidities follow at least partially different routes to HFpEF that are further affected by sex-dependent differences. Our results suggest that patients would benefit from a more tailored approach depending on their specific underlying pathology and gender.Main changes associated with HFpEF