Key variables in Fa-Binefa et al. and how to interpret the outcomes

In Fa-Binefa et al., the central exposure variable is smoking-onset age (SOA), meaning the age at which a person began smoking. The study examined whether starting smoking earlier was associated with later risk of cardiovascular disease (CVD) outcomes and all-cause mortality.

Main exposure: smoking-onset age

According to Fa-Binefa et al., SOA was analyzed as an indicator of tobacco exposure in addition to more familiar smoking measures such as smoking status and pack-years. The study evaluated whether SOA had a linear or non-linear relationship with later health outcomes.

• If the relationship appeared linear, SOA was treated as a continuous variable in Cox proportional hazards models.
• If the relationship was non-linear, the researchers used bootstrapping to identify a best cut-point where the slope of the association changed.
• They performed 1000 iterations per outcome and used the median observed cut-point as the final cut-point.

This matters because the study was not only asking whether smokers differed from non-smokers, but whether the age at smoking initiation among smokers predicted later risk.

Main outcomes

Fa-Binefa et al. focused on two outcome areas highlighted in the provided material:

• Cardiovascular disease mortality or CVD-related outcomes
• All-cause mortality

All-cause mortality means death from any cause, not only cardiovascular causes. This is a broad outcome that captures the total mortality burden associated with earlier smoking initiation.

What an adjusted hazard ratio means

The study reported adjusted hazard ratios (HRs) from Cox regression models. In Fa-Binefa et al., “adjusted” means the estimated association between SOA and an outcome was statistically controlled for other variables that could affect risk.

The covariates differed somewhat by outcome, but the tables show adjustment for variables such as:

• Age
• Sex (for all-cause mortality)
• Glucose
• Smoking status
• Pack-years
• Recruitment survey
• Years since quitting smoking
• And, for some models, factors such as cholesterol or hypertension treatment

An HR below 1.0 indicates that a later smoking-onset age is associated with lower risk. Because SOA is measured in years, the HR can be interpreted as the change in hazard for each additional year of delay in starting smoking.

Reported cut-point analysis around age 12

Fa-Binefa et al. found that the association between SOA and cardiovascular mortality was not simply linear across all ages of smoking onset. The bootstrapping analysis identified 12 years as the best cut-point.

The researchers therefore divided SOA into two strata:

• ≤ 12 years
• > 12 years

Within the younger-onset group (≤ 12 years), SOA was inversely associated with CVD mortality. Fa-Binefa et al. reported:

• HR = 0.58
• 95% CI: 0.37–0.90

This means that for those who started smoking at age 12 or younger, each additional year of delay in smoking onset was associated with a 42% lower risk of CVD mortality. The 42% comes from 1.00 - 0.58 = 0.42.

In the older-onset group (> 12 years), the study found no association with CVD risk:

• HR = 1.00
• 95% CI: 0.93–1.08

An HR of 1.00 indicates essentially no change in risk per additional year of later smoking onset in that older-onset stratum.

All-cause mortality finding

For all-cause mortality, Fa-Binefa et al. reported a linear association with smoking-onset age rather than a threshold effect like the one seen for cardiovascular mortality.

• HR = 0.96
• 95% CI: 0.93–0.98

This means that each year of delay in smoking onset was associated with about a 4% lower risk of death from any cause. The estimate is interpreted per one-year increase in SOA.

How to interpret these findings overall

• The key exposure was age at smoking initiation.
• The main outcomes highlighted here were CVD mortality/outcomes and all-cause mortality.
• Adjusted hazard ratios show the association after accounting for other risk-related variables.
• A later smoking-onset age was associated with lower all-cause mortality overall.
• For cardiovascular mortality, the association was strongest among those who started smoking at or before age 12.
• The age-12 cut-point suggests a possible especially vulnerable early-initiation period.

In short, Fa-Binefa et al. suggests that starting smoking very early, especially by age 12 or younger, is linked to worse later cardiovascular mortality outcomes, while each year of delaying smoking initiation is associated with lower all-cause mortality risk.