This prospective, population-based study represents the most comprehensive assessment of the metabolomic profile of future T2D risk in the Chinese population. There were strong positive associations of BCAA, apolipoprotein B/apolipoprotein A1, triglycerides, and VLDL particle size, and inverse associations of omega-3 fatty acids, HDL particle size and cholesterol concentrations in large HDL particles. The associations of several of the biomarkers most strongly related to future T2D risk were more extreme among participants with central obesity. When combined with traditional risk predictors, including glycaemia, circulating metabolic biomarkers significantly improved prediction of T2D over an average 8-year period.
The associations of BCAAs with incident T2D were among the strongest observed, and were qualitatively, and broadly quantitatively, consistent with previous study findings4,5,12,13. For example, a meta-analysis with ~ 1500 cases of incident T2D from seven individual prospective, predominantly Western population, studies, found adjusted RRs for T2D of 1.36, 1.36 and 1.35 per 1 SD higher isoleucine, leucine and valine, respectively4. Similarly, in a nested case–control study in China, comprising ~ 1500 incident T2D cases and a similar number of controls, there were positive associations of leucine/isoleucine and valine concentrations with T2D, with adjusted RRs comparing top vs. bottom quartiles of 1.75 and 1.54, respectively12. A genetic association study, including almost 50,000 T2D cases, found higher genetically-predicted BCAA concentrations were associated with increased T2D risk, suggesting a causal relationship21. A separate study, using genetic variants associated with BCAA and with insulin resistance, suggested insulin resistance leads to higher circulating BCAA concentrations, rather than the converse22. In combination, these findings suggest insulin resistance increases BCAA concentrations, which precede and contribute to T2D. This is consistent with persistence of the associations of BCAA in the present study after exclusion of T2D cases diagnosed during the first years of follow-up, and with previous descriptions of the trajectory from normoglycaemia to T2D23, highlighting a potentially valuable role for BCAA as markers of future T2D risk.
Our study showed strong inverse associations of omega-3 fatty acids with T2D risk. A large individual participant data meta-analysis, based on ~ 65,000 participants from 20 prospective studies (of mainly European ancestry) and > 16,000 cases of incident T2D, found qualitatively similar associations24. When analyses were limited to circulating fatty acids, individuals with combined omega-3 fatty acid, or docosahexaenoic acid, concentrations in the top, compared with the bottom, quintile had 23% and 24%, respectively, lower T2D risk. Prior investigations of the associations of fatty acids with T2D in Chinese populations are limited, but the described meta-analysis showed no clear heterogeneity across populations24. Although there are plausible mechanisms to support a protective effect of omega-3 fatty acids24, the causal relevance of the observed associations remains uncertain. However, the potential to influence omega-3 fatty acid levels through dietary intervention highlights the need for further investigation.
The large number of significant independent associations observed between circulating metabolic biomarkers and incident T2D risk in the present study in part reflects the focus of the metabolomics platform on lipid and lipoprotein measures, and correlations between these. The present study provides, for the first time, detailed characterisation of the relevance of lipoprotein size and subclass particle concentrations to T2D risk in a Chinese population. As shown in previous Western population studies5,25, we observed higher T2D risk among participants with higher concentrations of large VLDL particles and lower concentrations of large HDL particles, smaller mean HDL particle size and large mean VLDL particle size, as well as higher TG levels and lower HDL-cholesterol levels. This is consistent with an insulin resistant state26, which is a well-established component of the causal relationship between adiposity and T2D27. The observed stronger associations of certain metabolic biomarkers with T2D risk among centrally obese CKB participants may reflect greater prominence of insulin resistance in T2D aetiology among this population subgroup22,26,28,29. Although similar heterogeneity was not observed across BMI strata, the relative leanness of the study population prevented separate examination of the associations of metabolic biomarkers among participants with general obesity (i.e., BMI ≥ 30 kg/m2, observed in ~ 4% of the total CKB population30). At the same time, however, the population’s leanness provides a unique opportunity to expand our understanding of the aetiology of T2D among less adipose individuals and populations. In so doing, it valuably demonstrates the relevance of insulin resistance throughout the full adiposity range.
Recent prospective analyses among ~ 65,000 UK Biobank (UKB) participants examined the associations of 139 of the biomarkers considered herein (measured using the same NMR-metabolomics platform) with incident T2D (n = 1719) recorded during almost 12 years’ follow-up, adjusting for sociodemographic factors, fasting time, smoking, alcohol drinking and general and central adiposity25. Overall, the associations of 98 biomarkers were qualitatively consistent in the two study populations, including significant positive associations of 53 biomarkers with T2D risk and inverse associations of 27 biomarkers. However, the observed associations of several biomarkers appear more extreme in the CKB population, including BCAA (e.g., leucine HR 1.82 vs. 1.19 and valine 2.05 vs. 1.31 per 1 SD increment), apolipoprotein B/apolipoprotein A1 (1.79 vs. 1.09), and relative omega-3 fatty acid concentration (0.72 vs. 0.92). This is perhaps unexpected given the higher mean BMI in UKB (26.9 kg/m2) than in CKB (23.9 kg/m2 in subcohort participants). It is possible that these differences in the strength of the associations reflect, in part, ethnic differences in the typical pathophysiology of T2D19. Further studies directly comparing associations of metabolic biomarkers with T2D between ethnically diverse populations are needed, and may reveal novel insights into T2D aetiology.
The ability to identify individuals at greatest risk of T2D is vital for appropriate targeting of preventative interventions. Advances in “omics” research have stimulated interest in their potential for improving prediction of T2D risk over and above the traditional risk prediction models which frequently over-estimate actual risk31. An established risk prediction model in Chinese adults20 showed good discriminatory ability in CKB, with a c-statistic of 0.86, better than in the population in which it was developed (c-statistic 0.7720) and comparable to the performance of established models in other populations31. This strong discriminatory ability of established T2D risk prediction models presents challenges in identifying biomarkers capable of improving risk prediction. Thus, while addition of selected circulating metabolic biomarkers to the traditional T2D risk prediction model further improved its performance (c-statistic 0.91), the improvement was modest. Of note, however, although previous studies of mostly Western populations have observed enhanced discriminatory ability of T2D risk prediction models after inclusion of metabolic biomarkers, the degree of improvement was generally less marked5,7,11,12,15,32, with unclear generalisability to other populations. The few studies in China that have assessed this have frequently included limited biomarkers (e.g., restricted to amino acids33 or lipids34). The present study highlights the potential relevance of including biomarkers from diverse molecular pathways for improved risk prediction. Moreover, the standardised, targeted, high-throughput metabolomics platform used35,36 highlights the translational potential of the current study findings to clinical settings.
Our study had several strengths. It is among the largest Chinese population studies investigating prospective associations of circulating metabolic biomarkers with incident T2D12,32,33,34,37,38, and the largest to simultaneously investigate biomarkers across multiple diverse molecular pathways. Moreover, we employed an established targeted and validated metabolomics platform39,40, quantifying biomarker concentrations and enabling direct comparison with other studies. Furthermore, limited use of lipid-lowering medications in the study population reduced potential biases. However, the study had limitations. First, incident T2D was limited to diagnosed cases, although any associated misclassification would be expected to result in underestimation of associations of biomarkers with T2D. Second, repeat biomarker measurements were not available, preventing adjustment for intra-individual variation, again likely underestimating the strength of associations. Third, use of non-fasting blood samples may have increased inter-individual variation in biomarker concentrations. However, the analyses were adjusted for fasting time, as well as dietary factors, and there was no clear heterogeneity in associations across fasting time strata (data not shown). Fourth, lack of external validation of the risk prediction model incorporating metabolic biomarkers may have resulted in over-estimation of the model’s discriminatory ability. Finally, the observational nature of the study precludes conclusions regarding causality of observed associations.
Overall, the present study demonstrates highly significant associations of multiple circulating metabolic biomarkers from diverse molecular pathways with risk of future T2D in a relatively lean Chinese adult population. It highlights the ability of high-throughput, comprehensive, targeted NMR-metabolomic profiling to improve prediction of T2D beyond established risk factors (including glycaemia), demonstrating the potential clinical value of this approach in identifying those individuals most likely to benefit from early targeted T2D prevention efforts. Understanding of these associations is arguably of particular importance in China, where diabetes prevalence has escalated rapidly over recent decades, and continues to rise2.
Wise (formerly TransferWise) is the cheaper, easier way to send money abroad. It helps people move money quickly and easily between bank accounts in different countries. Convert 60+ currencies with ridiculously low fees - on average 7x cheaper than a bank. No hidden fees, no markup on the exchange rate, ever.
How to access the offer?
1- Click here
2- Select “Register''
3- Enter your email address, create a password, and select your country of residence
4- Fill out the required personal information, and the free first transfer offer will be applied automatically.
Benefits of the Multi-Currency Account:
- Free to create online
- Hold 50+ currencies
- Get multiple local bank details in one account (including EU, UK, US)
- Convert currency at the real exchange rate, even on weekends
- Spend whilst travelling on the Wise debit card without high conversion fees
Wise International Transfers:
- $1.5 billion saved by customers every year
- Send money to over 60 target currencies
- Lower fees for larger transfers
- No hidden fees. No bad exchange rates. No surprises.
- Send your money with a bank transfer, or a debit or credit card