Highlights
- Latent Class Analysis (LCA) identified four distinct clinical phenotypes among Chinese individuals with type 2 diabetes (T2D) and incident chronic kidney disease (CKD).
- The young-onset group (Class 1) incurred the highest longitudinal healthcare costs, with a mean per-patient-per-year (PPPY) cost of US$6,087.
- Middle-aged individuals with fewer baseline comorbidities (Class 3) experienced significant cost increases at the onset of CKD, driven largely by outpatient and psychiatric care.
- Early clinical intervention and risk stratification are essential to mitigate the substantial economic burden associated with diabetic kidney disease, particularly in early-onset populations.
Background
The global prevalence of type 2 diabetes (T2D) continues to rise, bringing with it an escalating burden of chronic kidney disease (CKD). CKD is one of the most significant and costly complications of diabetes, often leading to end-stage renal disease (ESRD), cardiovascular events, and premature mortality. In the Hong Kong healthcare system, which provides comprehensive public services through the Hospital Authority, managing the intersection of T2D and CKD represents a major economic challenge.
Standard clinical models often fail to capture the heterogeneity of patients with T2D. Recent shifts in precision medicine advocate for identifying specific patient subtypes or “trajectories” that might benefit from tailored interventions. Understanding the economic implications of these diverse clinical profiles is crucial for health policy experts and clinicians to prioritize resources effectively. The recent study by Du et al., published in Diabetologia, utilizes Latent Class Analysis (LCA) to bridge the gap between clinical phenotypes and longitudinal healthcare expenditure in a large-scale Chinese cohort.
Key Content
Methodological Framework: The Hong Kong Diabetes Register
The synthesis of evidence centers on a longitudinal study of 2,886 individuals from the prospective Hong Kong Diabetes Register (HKDR). Established in the 1990s, the HKDR provides a robust data infrastructure for evaluating long-term outcomes in Chinese patients. The cohort for this analysis spanned 2007 to 2019, accounting for 109,784 person-years of follow-up. The researchers employed 14 clinical and demographic variables to perform LCA, a person-centered statistical approach that identifies unobserved subgroups (latent classes) within a population based on shared characteristics.
Characterization of the Four Latent Classes
The LCA revealed four distinct classes, each with unique demographic profiles, comorbidity burdens, and medication usage patterns at the time of CKD incident:
- Class 1: Young Onset (18.3%) – This group had a mean age of 44.4 years at T2D onset. Despite their youth, they exhibited moderate comorbidities (25.6% with moderate/high Elixhauser Comorbidity Index [ECI]) and heavy medication usage (90.2% on ≥3 medications).
- Class 2: Old-Age Onset (21.2%) – Patients with a mean onset age of 66.9 years. They had similar comorbidity scores to Class 1 but slightly less intensive medication regimens at baseline (70.7% on ≥3 medications).
- Class 3: Middle-Aged, Low-Comorbidity (33.9%) – Onset at approximately 54.2 years. This group was characterized by the lowest baseline comorbidity (14.0% moderate/high ECI) and minimal medication use (15.6% on ≥3 medications).
- Class 4: Middle-Aged, Moderate Comorbidity (26.5%) – Similar onset age to Class 3 (54.1 years) but with significantly higher baseline treatment intensity (98.9% on ≥3 medications) and moderate comorbidities.
Differential Healthcare Costs and Resource Utilization
The average healthcare cost across the cohort was US$4,395 PPPY. However, the distribution was highly skewed across the latent classes. Class 1 (Young Onset) was the most economically impactful, with costs reaching US$6,087 PPPY. This group’s high expenditure was multifaceted, involving inpatient, outpatient, and emergency services, reflecting the aggressive nature of early-onset diabetes and its rapid progression to renal complications.
A significant finding was the “cost spike” observed in Class 3. Although these individuals appeared clinically “healthier” at baseline (fewer comorbidities and medications), their costs surged to US$4,260 PPPY upon the onset of CKD. Interestingly, this increase was primarily driven by psychiatric care and outpatient services, suggesting that the transition from a low-medication state to a CKD diagnosis may carry a significant psychological and diagnostic burden.
Clinical Outcomes and CKD Incidence
The study reported a CKD incidence of 26.29 per 1,000 person-years. The longitudinal cost evaluation used hierarchical generalized linear mixed models (HGLMM), which confirmed that the economic burden of CKD is not static but fluctuates significantly based on the timing of the diagnosis and the underlying patient phenotype. The young-onset class consistently remained the most expensive trajectory throughout the follow-up period.
Expert Commentary
The findings from Du et al. highlight a critical paradigm shift in diabetic care: the necessity of early and aggressive management for young-onset T2D. From a biological perspective, young-onset T2D is often associated with a more aggressive decline in beta-cell function and a higher lifetime exposure to hyperglycemia, which accelerates microvascular damage in the kidneys. The high PPPY cost of US$6,087 for this group underscores the “long-term cost of failure” in early prevention.
The data regarding Class 3 (middle-aged, low baseline risk) provides a cautionary tale for clinicians. Patients who seem well-controlled with few medications may fall into a “monitoring gap.” When these patients eventually develop incident CKD, the sudden increase in healthcare utilization—including psychiatric services—suggests that our current healthcare delivery systems may not be adequately preparing “lower-risk” patients for the psychosocial and physical impact of chronic complications. There is a clear need for integrating mental health support into diabetic renal care, particularly for those who experience a sudden shift in their health status.
Furthermore, the use of the Elixhauser Comorbidity Index in this study reinforces its utility as a predictive tool for healthcare expenditure, though the LCA approach proves superior by integrating age-at-onset and treatment intensity into the predictive matrix. One limitation of the study is its focus on a single geographic region (Hong Kong); however, given the high prevalence of T2D in Asian populations globally, these findings are highly relevant to other urbanized East Asian cohorts.
Conclusion
This latent class trajectory analysis provides a sophisticated roadmap for understanding the economic landscape of T2D and CKD. The evidence clearly identifies young-onset patients as the highest-cost subgroup, necessitating prioritized screening and the use of nephroprotective agents (such as SGLT2 inhibitors and GLP-1 RAs) earlier in their disease course. Additionally, the sudden cost rise in middle-aged patients with previously low comorbidity profiles suggests that vigilance must be maintained even in seemingly stable populations. Future research should focus on whether early intervention with novel therapeutics can flatten these high-cost trajectories and improve the quality of life for these high-risk phenotypes.
References
- Du Y, Zhang M, Li AQY, et al. Differential healthcare costs in individuals with type 2 diabetes and incident chronic kidney disease in Hong Kong: a latent class trajectory analysis. Diabetologia. 2026; PMID: 41848900.
- Chan JCN, Lim LL, Luk AOY, et al. The Hong Kong Diabetes Register: 25 years of research and emerging evidence. Lancet Diabetes Endocrinol. 2014;2(12):967-79. PMID: 25128038.
- Kong APS, Lau ESH, Luk AOY, et al. Secular trends in mortality and cardiovascular-renal complications in type 2 diabetes in Hong Kong, 2001-2016. PLoS Med. 2020;17(11):e1003363. PMID: 33206649.
