Multimorbidity Patterns and Associated Factors Among Middle-Aged and Older Adults in China: Evidence from the CHARLS Study

Using data from the 2018 CHARLS study, this research identifies distinct multimorbidity patterns among Chinese adults aged 45 and older, revealing that discordant multimorbidity (involving different disease systems) is more prevalent than concordant forms, is associated with poorer health outcomes and higher healthcare utilization, and is characterized by central clusters of cardiometabolic, arthritis, and digestive diseases.

The Gist

Imagine your body as a bustling city. Usually, we think of health problems as single construction sites: "Oh, there's a pothole on Main Street (high blood pressure)," or "There's a broken streetlight on 5th Avenue (arthritis)."

But for many older adults, the city isn't just dealing with one or two issues. It's facing a multimorbidity crisis—multiple construction sites happening all at once.

This study, based on data from over 8,000 middle-aged and older adults in China, asks a crucial question: Are these construction sites related to each other, or are they a chaotic mix of unrelated problems?

Here is the breakdown of the research in simple terms, using some creative analogies.

1. The Two Types of "City Chaos"

The researchers sorted the sick patients into two main groups based on how their diseases interact:

• The "Concordant" Group (The Synergistic Neighborhood):
  Imagine a neighborhood where all the houses are built with the same faulty wiring. If one house has an electrical fire, it makes sense that the neighbors might too.

  • What it is: This group has diseases that share the same root causes or treatment plans. For example, having high blood pressure, diabetes, and high cholesterol all together. They are all "cardiometabolic" (heart and metabolism) issues.
  • The Analogy: It's like having a garden where everything is a type of rose. They all need the same water, the same sun, and the same fertilizer. It's complex, but the gardener knows exactly what to do.

• The "Discordant" Group (The Mixed-Use District):
  Now imagine a neighborhood where one house is on fire, the next door has a leaking roof, the third has a broken foundation, and the fourth is being invaded by termites. These problems are totally different systems.

  • What it is: This group has a mix of diseases from different body systems. For example, someone with arthritis (joints), asthma (lungs), and stomach issues (digestion).
  • The Analogy: This is like a city where the plumbing, the electrical grid, and the traffic lights are all failing at once. The plumber can't fix the traffic light, and the electrician can't fix the roof. It requires a whole team of different experts to coordinate, and it's much harder to manage.

2. Who is in Which Group?

The study found that most people (about 8 out of 10) fell into the Discordant (mixed-up) category.

Who is more likely to be in the "Mixed-Up" group?

• Women: They were twice as likely to have this complex mix compared to men.
• Rural Residents: People living in the countryside were more likely to have this mix than city dwellers.
• Smokers: Both current and former smokers were more likely to have this chaotic mix.
• Active People: Surprisingly, people who did high-intensity exercise were more likely to have discordant multimorbidity. (Think of it as: maybe they are active because they are trying to manage their complex health issues, or perhaps the intensity of their activity interacts with their specific mix of conditions).

3. The Cost of the "Mixed-Up" City

The study compared the "Rose Garden" (Concordant) people with the "Mixed-Up District" (Discordant) people. The results were clear:

• The "Mixed-Up" group felt worse: They were more likely to be depressed, have trouble doing daily tasks (like bathing or walking), and rate their health as poor.
• The "Mixed-Up" group used more resources: They visited doctors more often, went to the hospital more frequently, and were less satisfied with the care they received.

Why? Because treating a "Rose Garden" is straightforward. You treat the heart, and you treat the diabetes, and the guidelines often overlap. But treating a "Mixed-Up District" is a nightmare of conflicting instructions. One doctor might say "take this pill," while another says "don't take that pill because it hurts your stomach." It's a logistical puzzle that leaves patients feeling overwhelmed and the healthcare system stretched thin.

4. The Five "City Clusters"

Using a special statistical tool called Latent Class Analysis (think of it as a super-smart sorting algorithm), the researchers found that even within the "Mixed-Up" group, people tended to fall into specific patterns:

1. The Heart & Metabolism Club: The most common group (mostly concordant).
2. The Joints & Stomach Crew: A mix of arthritis and digestive issues.
3. The Lungs & Joints Crew: A mix of breathing problems and joint pain.
4. The "Everything is Broken" Crew: A complex mix of heart, joints, stomach, and lungs.
5. The High Blood Pressure & Joints Crew: A specific pairing of hypertension and arthritis.

5. What Does This Mean for Us?

The main takeaway is that not all multiple diseases are created equal.

• For Doctors: We need to stop treating patients like they just have a list of separate problems. If a patient has a "Mixed-Up" city, they need a Conductor, not just a group of solo musicians. They need a primary care doctor who can coordinate between the heart specialist, the lung doctor, and the stomach doctor to make sure the treatments don't fight each other.
• For Patients: If you have a mix of different types of diseases, it's normal to feel more overwhelmed and use more healthcare. You aren't "failing" at managing your health; the system is just struggling to handle the complexity of your specific "city."

In short: Having one type of disease cluster is like managing a busy kitchen. Having a discordant mix is like managing a kitchen, a construction site, and a traffic control center all at the same time. This study tells us that the "construction site" group is larger, sicker, and needs a much more coordinated, team-based approach to get better.

Technical Summary

1. Problem Statement

Multimorbidity (the coexistence of two or more chronic conditions) is a growing global health burden. While prevalence and risk factors are well-studied, there is no standardized method to classify multimorbidity patterns. Existing literature often focuses on single-index diseases (e.g., diabetes) rather than the general population. Furthermore, the distinction between concordant multimorbidity (conditions sharing pathophysiological mechanisms or management strategies, e.g., cardiometabolic diseases) and discordant multimorbidity (conditions from different disease systems) is under-explored regarding their specific predictors, health consequences, and structural patterns in the Chinese population. Understanding these distinctions is crucial for moving from single-disease guidelines to integrated, patient-centered care.

2. Methodology

Data Source:

• Dataset: 2018 wave of the China Health and Retirement Longitudinal Study (CHARLS), a nationally representative household survey of adults aged ≥45 years.
• Sample: 8,974 individuals with two or more of 14 specific chronic conditions (hypertension, dyslipidemia, diabetes, cancer, chronic lung disease, liver disease, heart disease, stroke, kidney disease, digestive diseases, memory-related diseases, psychiatric disorders, arthritis/rheumatism, and asthma).
• Exclusion: Participants with missing key variables or only one chronic condition were excluded.

Definitions:

• Concordant Multimorbidity: Defined as the coexistence of two or more cardiometabolic diseases (hypertension, diabetes, dyslipidemia, heart disease, stroke, liver disease, kidney disease) OR asthma with another chronic lung disease, without other conditions.
• Discordant Multimorbidity: Any other combination of diseases (mixing different disease systems).

Statistical Analysis:

1. Logistic Regression: Used to identify predictors (demographics, lifestyle) associated with discordant vs. concordant multimorbidity.
2. Health Outcome Analysis: Logistic regression models (univariable and multivariable) assessed the association between multimorbidity type and six outcomes: depression, ADL limitations, self-reported health, outpatient visits, hospitalization, and satisfaction with health services.
3. Latent Class Analysis (LCA): A model-based clustering method used to identify subgroups of disease combinations.
   • Model 1: Applied to the entire multimorbid sample to identify overall patterns.
   • Model 2: Applied only to the discordant subgroup to isolate complex patterns.
   • Model Selection: Optimal class numbers were determined using Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), log-likelihood, and entropy, prioritizing clinical interpretability and parsimony.

3. Key Contributions

• Systematic Classification: Provides a robust classification of multimorbidity in China into concordant and discordant types, moving beyond simple disease counts.
• Pattern Identification: Utilizes LCA to map specific disease clusters, revealing that while cardiometabolic diseases are dominant, they frequently co-occur with musculoskeletal and digestive issues in complex ways.
• Health Burden Quantification: Demonstrates that discordant multimorbidity is not just more common but is significantly associated with worse health outcomes and higher healthcare utilization compared to concordant types.
• Policy Relevance: Highlights the need to shift from specialty-oriented care to integrated, multidisciplinary management, particularly for the discordant group.

4. Key Results

Demographics and Predictors:

• Prevalence: Discordant multimorbidity accounted for 81.4% (n=7,306) of the sample, while concordant accounted for 18.5% (n=1,668).
• Predictors of Discordance: Individuals were significantly more likely to have discordant multimorbidity if they were:
  • Female (OR = 2.17).
  • Rural residents (OR = 1.89).
  • Smokers (former or current).
  • Engaging in high-intensity physical activity (counter-intuitively, compared to low/no activity).
• Note: Age, marital status, education, and alcohol consumption were not significant predictors in multivariable models.

Health Outcomes and Utilization:
Patients with discordant multimorbidity (vs. concordant) had significantly higher odds of:

• Depressive symptoms (OR = 1.82).
• ADL limitations (OR = 2.03).
• Poor self-rated health (OR = 1.50).
• Healthcare Utilization: Higher rates of outpatient visits, home-based care, and hospitalization, alongside lower satisfaction with local health services.

Latent Class Analysis (LCA) Findings:

• Overall Sample (5 Classes):
  1. Cardiometabolic Class (41.7%): Hypertension, dyslipidemia, heart disease, diabetes.
  2. Arthritis-Digestive Class (33.3%): High prevalence of arthritis and digestive diseases.
  3. Respiratory Disease-Comorbidity Class (12.6%): Arthritis + chronic lung disease/asthma.
  4. Multisystem Morbidity Class (6.7%): Complex mix of all systems (cardio, arthritis, digestive, respiratory).
  5. Arthritis-Hypertension Class (5.7%): Specific coexistence of hypertension and arthritis.
• Discordant Subgroup Only (4 Classes):
  1. Digestive Disease-Comorbidity (37.7%).
  2. Arthritis-Cardiometabolic (26.8%).
  3. Respiratory Disease-Comorbidity (21.5%).
  4. Multisystem Morbidity (14.0%).

5. Significance and Implications

• Clinical Practice: The findings suggest that concordant multimorbidity (e.g., cardiometabolic) may be managed more effectively within existing single-disease guidelines due to shared pathophysiology. In contrast, discordant multimorbidity presents a complex clinical challenge where treatment goals may conflict, requiring multidisciplinary collaboration and integrated care models rather than simple accumulation of specialist treatments.
• Public Health: The high prevalence of discordant multimorbidity, particularly among rural women and smokers, indicates a need for targeted prevention strategies.
• Healthcare Systems: The increased utilization and lower satisfaction among discordant patients highlight a gap in current healthcare delivery. Systems must evolve to support shared decision-making and comprehensive assessment to manage the complexity of multisystem disease.
• Future Research: The study underscores the need for larger, more diverse datasets and unified definitions to better understand the mechanisms driving these complex disease patterns.

Conclusion:
This study establishes that discordant multimorbidity is the dominant form of disease coexistence in China's aging population, carrying a disproportionate burden of poor health and high healthcare costs. The identification of specific disease clusters (centered on cardiometabolic, arthritis, and digestive diseases) provides a critical framework for developing tailored, evidence-based management strategies.