Unequal effects of health behaviors in adolescence on adult cardiovascular disease and hypertension by family financial situation in the US: A cohort study

This US cohort study reveals that frequent adolescent alcohol use, particularly when combined with smoking, leads to a significantly higher risk of adult cardiovascular disease and hypertension among individuals from financially disadvantaged families compared to their peers, highlighting the unequal impact of health behaviors across socioeconomic backgrounds.

The Gist

Imagine your health as a long, winding road trip that starts in your teenage years and ends in your 40s. The destination? A healthy heart and normal blood pressure. But along the way, there are potholes (bad habits) and different types of cars (your family's financial situation).

This study is like a massive, 20-year GPS tracker for over 4,000 Americans. It asked a simple but profound question: Does hitting a pothole hurt everyone's car the same way, or does it wreck some cars more than others?

Here is the breakdown of what they found, using simple analogies.

The Setup: The Two Types of Drivers

The researchers looked at teenagers in the 1990s and followed them into adulthood. They split the teenagers into two groups based on their family's wallet:

1. The "Easy Street" Drivers: Families who could easily pay their bills.
2. The "Bumpy Road" Drivers: Families who struggled to pay their bills (financial difficulty).

They then watched to see how four specific "bad habits" (skipping breakfast, smoking, drinking alcohol, and not exercising) affected their heart health 20 years later.

The Big Discovery: The "Double Whammy"

The study found that for most people, bad habits are just bad habits. But for the "Bumpy Road" drivers, some bad habits acted like a double whammy.

Think of it like this:

• For the "Easy Street" drivers: Smoking and drinking are like driving with a flat tire. It's annoying, it slows you down, and it's risky, but you can still get to the destination.
• For the "Bumpy Road" drivers: Smoking and drinking are like driving with a flat tire while your brakes are cut and the road is covered in ice. The same bad habit causes much more damage because the car (the body) is already under more stress from the financial struggle.

The Specific Culprits

1. The "Party Combo" (Smoking + Drinking)
This was the biggest finding. When teenagers from struggling families both smoked and drank alcohol, their risk of heart disease or high blood pressure as adults skyrocketed compared to their wealthy peers who did the same thing.

• The Analogy: Imagine two people jumping off a diving board. The wealthy teen jumps into a deep, warm pool (their body has a safety net). The struggling teen jumps into a shallow, rocky pool. If they both jump, the struggling teen gets hurt much worse. The study found that the combination of smoking and drinking in adolescence created a "rocky pool" effect for those with financial stress.

2. The "Solo Act" (Drinking Alone)
Even just drinking alcohol (without smoking) was more dangerous for the struggling group.

• The Analogy: It's like a small leak in a boat. If the boat is new and sturdy (wealthy background), a small leak is manageable. If the boat is already old and patched up (financial stress), that same small leak can sink the whole ship.

3. The "Maybe" Habit (Skipping Breakfast)
At first, the data looked like skipping breakfast was also a "double whammy" for struggling families. However, when the researchers ran a "lie detector test" (using a negative control outcome—essentially checking if the pattern was just a fluke), it turned out the breakfast result was shaky.

• The Analogy: It looked like skipping breakfast was a crime, but the evidence was actually just a misunderstanding. The researchers aren't sure if skipping breakfast really causes this specific heart damage, so they are putting that one on the "maybe" shelf.

4. The "No-Show" Habits (Not Exercising)
Surprisingly, not exercising didn't show a huge difference between the two groups.

• The Analogy: Whether you were rich or poor, if you didn't exercise, your heart got a little weaker, but it didn't seem to hurt the "Bumpy Road" drivers any more than the "Easy Street" drivers.

Why Does This Happen?

The study suggests that when you grow up with money stress, your body might be in a constant state of "fight or flight" (like a car engine revving too high). When you add bad habits like smoking and drinking to an engine that is already revving too high, the engine blows a gasket much faster than it would for a car that was idling calmly.

The Takeaway for Society

This isn't just about telling teenagers "don't smoke." It's about realizing that the same advice doesn't work equally for everyone.

• The Old Way: Tell everyone to stop smoking and drinking.
• The New Way: Realize that for families struggling to pay bills, smoking and drinking are extra dangerous. We need to protect these families even more.

The Solution:
If we can get everyone to drink and smoke less, the people who are already struggling will see the biggest health improvements. It's like fixing the potholes on the road; the people driving the beat-up cars benefit the most, but everyone gets a smoother ride.

In short: Bad habits hurt everyone, but they hurt the poor much harder. To fix heart disease inequality, we need to target the "party combo" of smoking and drinking, especially for kids who are already stressed about money.

Technical Summary

1. Problem Statement and Research Gap

Cardiovascular diseases (CVDs) and hypertension exhibit a well-documented social gradient, where individuals from socioeconomically disadvantaged backgrounds face higher risks. While differential exposure (higher prevalence of risk behaviors in disadvantaged groups) is well-established, the mechanism of differential susceptibility (whether disadvantaged individuals suffer stronger harmful effects from the same behaviors) remains underexplored, particularly during adolescence.

Most existing literature focuses on adults or specific single behaviors. This study addresses the gap by investigating whether family financial situation in adolescence modifies the long-term causal effects of adolescent health risk behaviors on adult CVDs and hypertension. The authors hypothesize that adolescents from financially strained families may experience heightened biological or psychosocial vulnerability, leading to disproportionate health outcomes compared to their advantaged peers engaging in similar behaviors.

2. Methodology

Study Design and Population

• Data Source: The National Longitudinal Study of Adolescent to Adult Health (Add Health), a nationally representative US cohort.
• Cohort: Adolescents aged 12–19 at baseline (Wave I, 1994–1995, $n=4452$ for the causal analysis) followed into early midlife (Wave V, 2016–2018, ages 33–43).
• Exclusion Criteria: Participants older than 19 at baseline, missing health behavior data, missing Wave V biomarker data, or missing parental financial data.

Variables

• Exposures (Adolescent Health Behaviors):
  • Physical Activity: Sports and exercise frequency (categorized as low, moderate, high risk).
  • Substance Use: Cigarette smoking and alcohol use (categorized as never, past, rare, moderate/frequent).
  • Diet: Breakfast skipping (binary: skip vs. not skip).
  • Clustering: Latent Class Analysis (LCA) was used to identify co-occurring behavior patterns, specifically identifying a cluster of concurrent smoking and frequent alcohol use.
• Effect Modifier: Family Financial Situation, defined by parental self-report of difficulty paying bills (Yes/No) during adolescence. A sensitivity analysis used equivalized parental gross income (bottom vs. top quintiles).
• Outcomes (Adult Health):
  • Primary: Composite of CVDs (stroke, heart failure, heart disease, atrial fibrillation) and hypertension.
  • Secondary: Hypertension alone.
  • Measurement: Combination of biomarkers (blood pressure) and self-reported medical history/medication use at Wave V.
• Confounders: A minimally sufficient adjustment set derived from a Directed Acyclic Graph (DAG), including age, sex, race/ethnicity, adverse birth outcomes, chronic disease history, childhood overweight, adolescent mental disorders, adverse social environment, social network position, neighborhood deprivation, parental education, and parental smoking.

Statistical Analysis

• Causal Inference Approach: The study employed Double-Robust Inverse Probability Weighting (IPW).
  • Step 1: Multinomial logistic regression to model exposure probabilities and generate stabilized weights to balance confounders.
  • Step 2: Weighted logistic regression to estimate outcomes, including interaction terms between health behaviors and family financial situation.
• Estimand: The study estimated Conditional Average Causal Effects (cACE) on the additive scale (Risk Difference, RD) to quantify the excess burden of disease.
  • Primary comparison: High-risk behavior vs. Low-risk behavior.
  • Secondary comparison: Moderate-risk vs. Low-risk.
• Uncertainty: 95% Compatibility Intervals (CI) were generated using 1000 bootstrapped sampling designs with multiple imputation for missing covariates.
• Sensitivity Analyses:
  • Alternative effect modifier (income quintiles).
  • Negative Control Outcome: Adult abstention from voting (to detect unmeasured confounding).

3. Key Results

Descriptive Findings

• After a 21-year follow-up, 33.2% of participants had developed CVDs or hypertension.
• Adolescents from families with financial difficulty were more likely to skip breakfast and engage in infrequent physical activity compared to advantaged peers, though smoking and alcohol use showed minimal differences in prevalence between groups.

Causal Effects and Effect Modification

The study found significant differential susceptibility for specific behaviors:

1. Concurrent Smoking and Alcohol Use:

   • This combination showed the strongest differential effect.
   • Among disadvantaged adolescents, concurrent smoking and frequent alcohol use resulted in 11.1 percentage points (95% CI: 0.1–22.1) more cases of CVD/hypertension compared to advantaged peers engaging in the same behaviors.
   • The effect was negligible in the advantaged group (0.4 percentage points).

2. Alcohol Use Alone:

   • Frequent alcohol use led to an excess of 9.7 percentage points (95% CI: 0.8–18.5) in cases for disadvantaged adolescents compared to advantaged ones.

3. Breakfast Skipping:

   • Initial analysis suggested a differential effect (5.4 percentage points excess for disadvantaged).
   • However, sensitivity analysis using the negative control outcome (voting abstention) showed a similar large differential effect for breakfast skipping. This suggests the observed association is likely driven by unmeasured confounding rather than a true causal mechanism.

4. Other Behaviors:

   • Smoking alone and infrequent physical activity did not show strong or consistent evidence of differential susceptibility by financial situation.

Sensitivity Analysis

• Using parental income quintiles instead of bill-paying difficulty confirmed the results for alcohol and concurrent smoking/alcohol, with an even larger effect modification for the concurrent behavior (27.3 percentage points).
• The negative control analysis reinforced that the findings for alcohol and concurrent smoking are robust against unmeasured confounding, unlike the breakfast skipping findings.

4. Key Contributions

• Identification of Differential Susceptibility: The study provides robust evidence that socioeconomic disadvantage in adolescence amplifies the long-term cardiovascular harm of specific health behaviors (alcohol and concurrent smoking), independent of the prevalence of those behaviors.
• Adolescent Focus: It shifts the focus from adult behaviors to the critical developmental window of adolescence, showing that susceptibility patterns are established early.
• Behavioral Clustering: By using Latent Class Analysis, the study highlights that concurrent smoking and drinking is a distinct, high-risk lifestyle pattern with disproportionate effects on disadvantaged youth, suggesting interventions should target this specific combination.
• Methodological Rigor: The application of double-robust IPW and negative control outcomes in a nationally representative US cohort strengthens the causal claims and addresses potential biases common in observational epidemiology.

5. Significance and Implications

• Public Health Policy: The findings suggest that "whole-population" interventions (e.g., universal policies reducing alcohol consumption and smoking) could yield greater health equity. Because disadvantaged groups are more susceptible, a uniform relative reduction in these behaviors would result in a larger absolute reduction in disease burden for the disadvantaged, thereby narrowing the socioeconomic health gap.
• Targeted Support: Recognizing heightened susceptibility supports the need for targeted programs that provide additional resources and support to socioeconomically disadvantaged adolescents to mitigate these amplified risks.
• Mechanism Insight: The results support the "biological embedding" hypothesis, where early-life material hardship may sensitize physiological stress response systems, making the body more vulnerable to the toxic effects of alcohol and tobacco.
• Limitations: The study notes potential recall bias for retrospective confounders and the reliance on self-reported behaviors, though the negative control analysis mitigates concerns about unmeasured confounding for the primary findings.

In conclusion, this study demonstrates that the pathway from adolescent risk behaviors to adult cardiovascular disease is not uniform; it is significantly steeper for those facing financial hardship, particularly regarding alcohol use and the co-occurrence of smoking and drinking.