The insulin-like growth factor (IGF) axis and its binding proteins have been implicated in glucose metabolism and cardiovascular disease. Prior work established that higher circulating IGFBP-1 and IGFBP-2 are markers associated with lower risk of incident type 2 diabetes. However, how these proteins relate to cardiovascular outcomes has been less clear. This analysis used the UK Biobank resource to examine associations of IGFBP-1 and IGFBP-2 with incident diabetes, macrovascular disease and mortality, integrating plasma proteomic measures and imaging data.
Approximately 50,000 UK Biobank participants with plasma proteomic measurements for IGFBP-1 and IGFBP-2 were included in the analyses. The study combined disease outcome data, routine biochemical markers, and cardiac and vascular imaging from UK Biobank. Multivariate Cox regression models were used to assess associations of IGFBP-1 and IGFBP-2 quartiles with incident diabetes, incident macrovascular disease, and mortality outcomes. Associations with routine biomarkers and MRI-derived cardiac and aortic measures were also evaluated. The report is a preprint and methodological detail beyond these high-level descriptions (for example, specific covariates included in the multivariate models) was not reported in the source abstract.
Participants in the highest quartiles of IGFBP-1 and IGFBP-2 had substantially lower risk of developing diabetes during follow-up. The reported hazard ratios for incident diabetes were 0.31 for IGFBP-1 and 0.32 for IGFBP-2, indicating approximately a 68–69% lower hazard compared with lower quartiles. Both proteins were negatively associated with HbA1c, triglyceride/HDL ratio and abdominal adiposity, findings that align with their inverse associations with diabetes risk.
Despite their association with lower diabetes incidence, higher circulating IGFBP-1 and IGFBP-2 were paradoxically associated with increased risk of incident macrovascular disease and with higher all-cause and cardiovascular-related mortality. The study reports increased risks with hazard ratios of 1.81 and 2.39. The source abstract does not provide further granularity on which hazard ratio corresponds to which outcome or protein beyond the overall statement of increased risk. These discordant associations suggest that favourable relationships with glycemic and metabolic markers did not translate into reduced cardiovascular risk in this cohort.
Both IGFBP proteins correlated inversely with metabolic risk markers (lower HbA1c, lower triglyceride/HDL ratio, and less abdominal adiposity) but showed positive associations with cardiac injury and stress markers: higher NT-proBNP and higher troponin I. Imaging associations included larger cardiac chamber sizes and increased aortic dimensions. Thus, higher IGFBP-1 and IGFBP-2 levels were linked to biochemical and structural features that are typically interpreted as markers of cardiac stress or remodeling, which may help explain the increased cardiovascular risk observed.
The findings indicate a discordance in the relationships of IGFBP-1 and IGFBP-2 with metabolic and cardiovascular outcomes: protective associations for incident diabetes coexist with associations suggesting greater cardiovascular risk. This pattern raises the possibility of complex or context-dependent roles for these binding proteins in cardiometabolic pathophysiology. Clinically, the results imply that circulating IGFBP-1 and IGFBP-2 cannot be assumed to confer uniformly lower cardiometabolic risk despite favourable links to diabetes-related measures. The study highlights the need for mechanistic work to clarify whether IGFBP-1 and IGFBP-2 have direct causal effects on cardiac structure and function or whether they are markers of other processes that increase cardiovascular risk.
This report is a preprint and has not been peer reviewed. The abstract provides limited methodological detail (for example, the full set of covariates used in multivariate models and temporal aspects of follow-up were not specified in the source abstract). Causal inference cannot be drawn from reported associations. Ethical approval for UK Biobank data use was noted (NHS Research Ethics Service 11/NW/0382) and the study used application ID 547717. The authors declared no competing interests.
The authors reported funding support from the British Heart Foundation. No competing interests were declared. The full paper and supplementary material may contain additional details on methods, sensitivity analyses and subgroup assessments; those details were not reported in the source abstract.