📄 medicine

Temporal Patterns of Cardiometabolic Risk Factors Preceding Incident Cardiovascular Disease: Evidence from the TLGS

This longitudinal study of the Tehran Lipid and Glucose Study (TLGS) cohort demonstrates that longitudinal trends in metabolic risk factors, particularly systolic blood pressure, fasting plasma glucose, and the TyG index, diverge years before clinical cardiovascular disease onset, highlighting their potential as early predictive markers for preventive strategies.

Original authors: maryam mahdavi, Anoshirvan Kazemnejad, Davood Khalili, Ahmadreza Tajari, Abbas Asosheh

Published 2026-06-30

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: maryam mahdavi, Anoshirvan Kazemnejad, Davood Khalili, Ahmadreza Tajari, Abbas Asosheh

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: The "Slow Leak" Before the Flood

Imagine your heart and blood vessels are like a complex plumbing system in a house. For years, you might not see any problems, but slowly, a small leak starts developing. By the time you hear the water dripping or see the flood damage (a heart attack or stroke), the damage has been building up for a long time.

This study, conducted in Tehran, Iran, wanted to find out if we could spot that "slow leak" years before the "flood" happens. The researchers looked at a group of 1,872 adults (aged 40–79) who were healthy at the start. They followed them for 10 years to see who eventually developed cardiovascular disease (CVD).

The Detective Work: Tracking the "Vital Signs"

The researchers didn't just take one snapshot of the participants' health. Instead, they acted like detectives tracking a suspect over time. They checked the participants' health every few years, looking at things like:

Blood Pressure: How hard the water is pushing through the pipes.
Blood Sugar: How much "syrup" is in the blood.
The TyG Index: A special math formula that combines sugar and fat levels to guess how well the body handles insulin (a key hormone for managing sugar).
Cholesterol and Weight: Other factors that clog or strain the system.

What They Found: The "Early Warning Signals"

The study found that the people who eventually got sick didn't just have bad numbers on the day they were diagnosed. Their numbers started going in the wrong direction years before the disease actually hit.

Here are the three main "warning lights" that started flashing early:

Systolic Blood Pressure (SBP): The "Pressure Gauge"
- The Analogy: Think of this like the pressure in a garden hose.
- The Finding: This was the most reliable warning sign. The researchers saw that the blood pressure of the people who later got sick started creeping up 9 years before the event. Even after adjusting for other factors like smoking or medication, this pressure gauge kept separating the "healthy" group from the "future sick" group. It was the most consistent signal.
Fasting Plasma Glucose (FPG) & The TyG Index: The "Sugar and Fat Mix"
- The Analogy: Imagine a car engine that is getting gummed up with sticky syrup.
- The Finding: About 6 years before the disease started, the blood sugar levels and the TyG index (the sugar-fat mix) began to rise faster in the group that got sick compared to the healthy group.
- The Catch: However, when the researchers accounted for lifestyle changes (like people starting to exercise more or taking new medicines), this signal got a bit fuzzier. It suggests that while rising sugar is a warning, it's also heavily influenced by what people do to fix it (like taking pills or changing their diet).

What Didn't Work as a Warning

Interestingly, some things we usually worry about didn't show a clear "early warning" pattern in this specific study.

BMI (Weight): The study found that simply being overweight at the start didn't predict who would get sick years later. It's like saying a heavy backpack doesn't always mean you'll trip; it's how the weight changes over time that matters.
Cholesterol: Unlike blood pressure, cholesterol levels didn't show a clear, steady climb in the sick group compared to the healthy group before the event.

The "One-Time Check" vs. The "Movie"

The paper makes a crucial point about how we usually check health.

The Old Way: Taking a photo (a single blood test) on one day. The paper notes that this photo can sometimes miss people who are actually at high risk because it doesn't show the story of their health.
The New Way (This Study): Watching a movie (looking at the trends over 10 years). The study suggests that watching how these numbers move over time gives a much better prediction than just looking at a single snapshot.

The Limitations: Why We Can't Apply This Everywhere Yet

The authors are careful to say this isn't a magic crystal ball for everyone.

The Location: The study was done in a specific, busy city district in Tehran. The people there have specific diets, lifestyles, and access to doctors that might be different from people in rural Iran or other countries.
The Timing: The health checks happened every 3 years. It's like checking a car's oil only once every three years; you might miss a sudden spike in engine trouble that happens in between.
The Conclusion: The study is "hypothesis-generating." It's like finding a new clue that suggests we should watch these trends more closely, but it doesn't yet prove that changing these trends will definitely stop the disease.

The Bottom Line

If you want to understand your heart health, don't just look at your numbers on the day of your check-up. Look at the trend.

If your blood pressure has been slowly climbing for years, that's a major red flag.
If your blood sugar and fat markers are drifting upward over time, that's another warning sign.

The study concludes that by paying attention to these long-term trends, doctors might be able to spot trouble years before a heart attack or stroke occurs, giving people a chance to fix the "leak" before the "flood" arrives.

Technical Summary: Temporal Patterns of Cardiometabolic Risk Factors Preceding Incident Cardiovascular Disease

Problem Statement
Cardiovascular disease (CVD) remains the leading cause of mortality globally and in Iran, where it accounts for approximately 46% of all deaths. While established risk scores (e.g., ACC/AHA Pooled Cohort Equations) are effective for baseline risk stratification, they rely on single-time-point measurements. Previous analyses of the Tehran Lipid and Glucose Study (TLGS) cohort indicated that some individuals who later experienced severe cardiovascular events had baseline risk estimates below standard treatment thresholds. This suggests that static baseline assessments may fail to capture the long-term evolution of cardiometabolic risk. The study addresses the need to understand whether longitudinal trends of metabolic risk factors—specifically their trajectories over time—precede clinical CVD onset and can serve as early predictive markers beyond single measurements.

Methodology

Study Design & Population: This longitudinal study utilized data from the Tehran Lipid and Glucose Study (TLGS). The analysis included 1,872 adults aged 40–79 years who were free of CVD at the baseline examination (the third cycle, 1999–2018). Participants were followed for 10 years until March 2018.
Data Collection: Standardized protocols were used to collect demographic, lifestyle, and metabolic data, including anthropometrics, blood pressure, and fasting/2-hour plasma glucose and lipid profiles.
Outcome Definition: Incident CVD was defined as the occurrence of coronary heart disease (myocardial infarction, unstable angina, or angiography-confirmed CAD) or stroke.
Statistical Analysis:
- Participants were categorized into those who developed incident CVD and those who did not.
- Generalized Estimating Equations (GEE) were employed to model the longitudinal trends of metabolic indicators. This approach allowed for the inclusion of all available repeated measurements and handled missing data robustly.
- The models assessed the interaction between time (follow-up intervals) and CVD status, adjusting first for age and sex (Model 1), and subsequently for medication use, smoking, physical activity, and family history (Model 2).
- Trends were examined at 3-year intervals retrospectively from the time of CVD diagnosis (or final visit).

Key Results

Incidence: Over the 10-year follow-up, 117 participants (6.3%) developed incident CVD.
Baseline Differences: At baseline, the CVD group was significantly older, had higher body weight, blood pressure (SBP/DBP), fasting plasma glucose (FPG), 2-hour plasma glucose, total cholesterol, LDL, non-HDL, and TyG index compared to the non-CVD group.
Longitudinal Trends (Trajectory Divergence):
- Systolic Blood Pressure (SBP): SBP showed a consistent upward trend in the CVD group, with significant divergence from the non-CVD group evident nine years prior to the CVD event ( $P_{interaction} = 0.013$ ). This separation remained significant even after full adjustment for covariates.
- Fasting Plasma Glucose (FPG) & TyG Index: Both FPG and the Triglyceride-Glucose (TyG) index demonstrated significant divergence between groups starting six years before diagnosis in age- and sex-adjusted models ( $P_{interaction} = 0.031$ and $0.032$, respectively). However, these interactions were attenuated and became non-significant after adjusting for medication, lifestyle, and family history, suggesting that clinical management and lifestyle changes partially explain these trajectory differences.
- CVD Risk Score: The predicted 10-year CVD risk score (AHA/ACC) showed a significant rise in the six years leading up to diagnosis ( $P_{interaction} < 0.001$ ).
- Other Factors: No significant group-by-time interactions were observed for BMI, triglycerides, total cholesterol, LDL, or HDL, indicating similar time trends between converters and non-converters for these specific markers.

Significance and Claims
The authors position these findings as hypothesis-generating rather than a formal external validation or refutation of existing risk scores. The study claims that:

Longitudinal Monitoring Adds Value: Longitudinal trends of specific metabolic factors, particularly SBP, FPG, and the TyG index, are associated with incident CVD years before clinical onset.
SBP as a Robust Marker: SBP demonstrated the most consistent separation between groups, maintaining significance even after adjusting for confounders, suggesting it is a robust longitudinal predictor.
Complementary Role: These trajectory-based findings should be viewed as complementary to established baseline risk scores (like the ACC/AHA equations) rather than replacements. They may help identify high-risk individuals whose risk evolves beyond what is captured by a single baseline assessment.
Clinical Implication: The results emphasize the importance of repeated risk assessments and early detection of metabolic risk factor trends to inform preventive strategies.

Limitations Acknowledged by the Authors

Generalizability: Data were collected from an urban district in Tehran; findings may not extrapolate to rural Iranian populations or other international cohorts with different genetic, lifestyle, and healthcare contexts.
Temporal Resolution: Measurements were taken every three years, which limits the ability to capture short-term fluctuations or precise inflection points. The reported divergence times (3, 6, 9 years) are approximate intervals.
Residual Confounding: While adjustments were made for several factors, incomplete control for time-varying diet and socioeconomic changes may bias estimates.
Causality: The study identifies metabolic trajectories as risk markers, not independent causal determinants. The observed associations likely reflect the joint impact of metabolic status, lifestyle dynamics, and clinical management.