Current but Not Former Smoking Is Associated With Higher HbA1c in Adults Without Diabetes

Based on an analysis of NHANES 2015-2018 data, current smoking is significantly associated with higher HbA1c levels in U.S. adults without diabetes, whereas former smoking shows no such association, suggesting the effect diminishes after cessation.

The Gist

The Big Picture: Smoking, Sugar, and the "Sweet Spot"

Imagine your body is a car, and HbA1c is the dashboard gauge that tells you how much sugar (glucose) is floating around in your fuel tank over the last few months. Doctors use this gauge to check if you are driving safely (normal), if you are running a little hot (prediabetes), or if your engine is overheating (diabetes).

For a long time, scientists knew that smoking was bad for this gauge. They knew smokers usually had higher sugar levels. But they didn't quite understand why, or if quitting smoking would actually reset the gauge back to normal.

This study took a huge look at data from nearly 10,000 American adults to answer two questions:

1. Does smoking still raise your sugar gauge even after we account for other bad habits (like being overweight or having high blood pressure)?
2. If you quit smoking, does your sugar gauge go back down to where it was before you started?

The Experiment: Two Ways to Check the Smoke

The researchers didn't just ask people, "Do you smoke?" because people sometimes lie or forget. Instead, they used two methods:

1. The Self-Report: Asking people directly.
2. The Smoke Test: Measuring cotinine in the blood. Think of cotinine as a "smoke fingerprint." It's a chemical left behind by nicotine that stays in your blood for a day or two. It's the objective truth-teller.

The Findings: The "Current" vs. "Former" Smoker Mystery

The study found a fascinating split in the results, depending on whether the person currently had diabetes or not.

1. The "Full Population" (Everyone Mixed Together)

When they looked at everyone (including people who already had diabetes), the link between smoking and high sugar got fuzzy once they adjusted for other factors like waist size and inflammation.

• The Analogy: Imagine trying to hear a whisper in a noisy room. When you account for the noise (other health issues), the whisper (smoking) isn't as loud as we thought. It seemed like smoking's effect was mostly because smokers also tended to have larger bellies or more inflammation, which were the real culprits for the high sugar.

2. The "No-Diabetes" Group (The Healthy Drivers)

This is where the story gets exciting. When they looked only at people who did not have diabetes yet:

• Current Smokers: Their sugar gauge was definitely higher. Even after adjusting for weight and blood pressure, smoking was still pushing the needle up.
• Former Smokers (Quitters): Their sugar gauge was exactly the same as people who had never smoked a day in their lives.

The Metaphor: Think of smoking like a heavy backpack you wear.

• Current Smokers are wearing the backpack. It's heavy, making their engine work harder, and their sugar gauge goes up.
• Former Smokers took the backpack off. Even though they might still be a bit out of shape (higher weight) compared to never-smokers, the moment they dropped the backpack, their sugar levels returned to normal. The damage to the sugar gauge wasn't permanent; it was just the weight of the pack.

The "Threshold" Effect

The study also looked at how much smoke was in the system. They found that you don't need a little bit of smoke to see a change; you need a lot of smoke.

• Passive Smokers: People exposed to second-hand smoke (low levels of the "smoke fingerprint") did not see their sugar levels rise.
• Active Smokers: Only those with high levels of the fingerprint (active smokers) saw the spike.
• The Analogy: It's like a light switch. A little bit of smoke doesn't turn the light on. You have to flip the switch fully (active smoking) to see the effect.

Why Does This Matter?

1. The "Reset Button" is Real
The most hopeful takeaway is that for people without diabetes, quitting smoking seems to "reset" the sugar gauge. The study suggests that the body's ability to manage sugar recovers quickly after you stop smoking, even if you gain a little weight afterward.

2. The Danger of Misdiagnosis
Because smoking pushes the sugar gauge up, a healthy smoker might accidentally get labeled as "prediabetic" just because they smoke, not because they are actually at high risk for diabetes. If a doctor doesn't know the patient smokes, they might worry unnecessarily.

3. Motivation to Quit
The study gives a new reason to quit: It's not just about lungs or heart; it's about your blood sugar. If you quit, your body's sugar management system can bounce back to a "never-smoker" state.

The Catch (Limitations)

The researchers admit this study is a snapshot in time (like a photo), not a movie. They can't prove that smoking caused the sugar spike or that quitting caused the drop, because they didn't follow people over many years. However, the biological logic and the consistency of the data make a very strong case.

The Bottom Line

If you don't have diabetes yet, smoking is like carrying a heavy weight that makes your blood sugar look worse than it really is. But the good news? You can drop that weight. Once you quit, your blood sugar levels appear to return to normal, just like someone who never smoked at all. It's a powerful incentive to put down the cigarette and pick up a healthier future.

Technical Summary

1. Problem Statement

Cigarette smoking is a well-established risk factor for type 2 diabetes mellitus (T2DM), with meta-analyses indicating a 30–50% increased risk among current smokers. Previous studies have consistently reported elevated Glycated Hemoglobin (HbA1c) levels in smokers. However, the biological mechanisms underlying this association and its clinical implications remain incompletely understood.

Key gaps addressed by this study include:

• Reversibility: It is unclear whether the effect of smoking on HbA1c is reversible upon cessation, particularly in adults without established diabetes.
• Confounding: Previous studies often failed to simultaneously adjust for central adiposity, systemic inflammation, and lipid profiles, which are known to influence both smoking behavior and HbA1c levels.
• Measurement Bias: Reliance on self-reported smoking status can introduce misclassification due to social desirability bias. The study aims to validate findings using serum cotinine, an objective biomarker.

2. Methodology

Study Design and Data Source

• Design: Cross-sectional analysis.
• Data Source: National Health and Nutrition Examination Survey (NHANES) cycles 2015–2018.
• Population: Adults aged ≥20 years.
• Sample Size:
  • Full population (self-reported): $N=9,214$.
  • Full population (cotinine): $N=9,221$.
  • Subgroup (non-diabetic): $N=7,328$.

Exposure Assessment

• Self-Reported Smoking: Categorized into Never (<100 cigarettes lifetime), Former (≥100 lifetime, not current), and Current (≥100 lifetime, current).
• Serum Cotinine: Measured via isotope-dilution high-performance liquid chromatography. Categorized as:
  • Low/None: <0.05 ng/mL.
  • Environmental exposure: 0.05–<3.0 ng/mL.
  • Active smoking: ≥3.0 ng/mL.
• Sensitivity Analyses: Cotinine was also analyzed as a continuous log-transformed variable, in quintiles, and via winsorization (capped at 99th percentile) to test for non-linearity.

Outcome Assessment

• Primary Outcome: HbA1c (continuous variable), measured via high-performance liquid chromatography.

Covariates and Modeling Strategy

• Adjustment Variables: Age, sex, race/ethnicity, systolic blood pressure, waist circumference (WC), non-HDL cholesterol, and log-transformed high-sensitivity C-reactive protein (hsCRP).
• Model Progression:
  • Model 1: Demographics only.
  • Model 2–5: Sequential addition of blood pressure, BMI, and WC.
  • Model 6 (Primary): Full adjustment including WC, non-HDL, and hsCRP.
  • Model 7: Sensitivity model adding education and poverty-income ratio.
• Statistical Approach: Survey-weighted multivariable linear regression accounting for NHANES complex survey design (strata, PSUs, weights). Interaction terms between smoking and diabetes status were tested.

3. Key Contributions

1. Differentiation by Diabetes Status: The study explicitly stratifies analysis by diabetes status, revealing that the association between smoking and HbA1c is driven primarily by the non-diabetic population, a nuance often missed in aggregate analyses.
2. Reversibility Evidence: It provides evidence that HbA1c levels in former smokers without diabetes return to levels comparable to never smokers, suggesting the metabolic impact of smoking on glycemia is largely reversible.
3. Mechanistic Insight: Through sequential modeling, the study demonstrates that the association between smoking and HbA1c is largely mediated by central adiposity (waist circumference) and inflammation (hsCRP), rather than being a direct independent effect.
4. Validation of Biomarkers: The study confirms that objective cotinine measurements align with self-reported data regarding the association with HbA1c, validating the use of both metrics while highlighting that environmental tobacco smoke exposure (low cotinine) does not significantly elevate HbA1c.

4. Results

Demographics

• Current smokers were younger (mean age 44.3) and had lower educational attainment compared to former (53.7) and never smokers (46.8).
• Former smokers exhibited higher adiposity (WC and BMI) than never smokers.

Primary Findings (Full Population)

• In minimally adjusted models, both former and current smoking were associated with higher HbA1c.
• After full adjustment (Model 6): The association became non-significant for both groups.
  • Former: $\beta=0.029\%$, $p=0.30$.
  • Current: $\beta=0.053\%$, $p=0.13$.
• This suggests that in the general population, the observed HbA1c elevation is confounded by cardiometabolic risk factors (adiposity, lipids, inflammation).

Subgroup Findings (Adults Without Diabetes)

• Current Smoking: Remained significantly associated with higher HbA1c even after full adjustment.
  • $\beta=0.067\%$ (95% CI: 0.039 to 0.094), $p<0.001$.
• Former Smoking: Showed no significant association with HbA1c compared to never smokers.
  • $\beta=-0.001\%$, $p=0.923$.
• Cotinine Analysis: Among non-diabetics, active smoking (cotinine $\ge3.0$ ng/mL) was associated with higher HbA1c ($p<0.001$), whereas environmental exposure was not. No monotonic dose-response relationship was observed.

Mediation Analysis

• Adjustment for waist circumference accounted for approximately 63% of the attenuation in the association for former smokers and ~33% for current smokers.
• hsCRP showed the strongest independent association with HbA1c among all covariates in the fully adjusted model ($\beta=0.141$, $p<0.001$).

5. Significance and Implications

Clinical Implications

• Screening Accuracy: The modest increase in HbA1c among current non-diabetic smokers ($\approx0.07\%$) could push individuals across diagnostic thresholds (e.g., from normal to prediabetes), potentially leading to misclassification if smoking status is not considered.
• Cessation Counseling: The finding that former smokers have HbA1c levels comparable to never smokers (despite having higher adiposity) supports the message that metabolic recovery is achievable. This provides a strong, specific motivator for cessation counseling: "Your blood sugar levels can return to normal."
• Early Intervention: The absence of residual elevation in former smokers suggests that the window for reversing smoking-induced glycemic effects is open, particularly before the onset of established diabetes.

Public Health Impact

• The study reinforces the need for early smoking cessation interventions to prevent the development of prediabetes and diabetes.
• It clarifies that the "smoking-HbA1c" link is not a fixed biological constant but is mediated by modifiable factors like central adiposity and inflammation.

Limitations

• Cross-Sectional Design: Cannot establish causality or determine the exact timeline for HbA1c recovery post-cessation.
• Residual Confounding: Potential unmeasured confounders (diet, physical activity, alcohol) were not included.
• Interaction Test: The statistical interaction between smoking and diabetes status did not reach significance, suggesting the subgroup findings should be interpreted with caution and require replication in longitudinal studies.

Conclusion
The study concludes that while current smoking is associated with elevated HbA1c in adults without diabetes, this effect is mediated largely by central adiposity and inflammation. Crucially, this elevation appears to be reversible upon cessation, as former smokers exhibit HbA1c levels indistinguishable from never smokers. These findings support the integration of smoking cessation into diabetes prevention strategies and suggest that HbA1c screening should account for active smoking status to avoid misclassification.