Heaviness of Smoking Index: Scoring, Cutoffs & Interpretation
The HSI estimates cigarette dependence from two questions - how soon after waking someone smokes, and how much they smoke. It predicts cessation outcome about as well as the full six-item Fagerström test, and takes a fraction of the time.
Both questions refer to current smoking. The index is designed for daily cigarette smokers; it is not validated for non-daily or intermittent smokers, for other tobacco products, or for e-cigarettes. If either item is unanswered, no total can be calculated - there is no published imputation rule.
01
At present, how long after waking do you wait before having your first cigarette?
Item score-
02
How many cigarettes per day do you smoke?
Item score-
Scored locally in your browser - nothing is sent until you choose to.
0 of 2 answered
0 / 6
HSI0 / 2
Both questions refer to current smoking. The index is designed for daily cigarette smokers; it is not validated for non-daily or intermittent smokers, for other tobacco products, or for e-cigarettes. If either item is unanswered, no total can be calculated - there is no published imputation rule.
01
At present, how long after waking do you wait before having your first cigarette?
Item score-
02
How many cigarettes per day do you smoke?
Item score-
0 of 2 answered0 / 6
Attribution & licence
Items reproduced from PhenX Toolkit protocol 330201 under CC BY 4.0. Original instrument: Heatherton TF, Kozlowski LT, Frecker RC, Rickert W, Robinson J. Br J Addict. 1989;84(7):791-799.
The HSI measures the heaviness and automaticity of cigarette consumption. Time to the first cigarette indexes how quickly overnight nicotine depletion begins to drive behaviour; cigarettes per day indexes overall exposure. Together they approximate physical dependence well enough to guide a decision about pharmacotherapy. What they do not capture is dependence as DSM or ICD define it - there is no item for craving, loss of control, persistence despite harm, or neglect of other activities. The construct is narrower than the word "dependence" suggests, and is better described as withdrawal-driven smoking.
02 - Scoring & cutoffs
How scoring works.
Add the two item scores. The total runs from 0 to 6. There is no reverse scoring and no weighting.
The bands shown below - 0 to 1 low, 2 to 4 medium, 5 to 6 high - are the ones used by the PhenX Toolkit and NIDA Data Share, and they are the most widely applied. They should be attributed to those bodies rather than to the original authors: the 1989 paper proposed the four-category scoring of each item but contains no total score and no severity bands at all, so the common citation of Heatherton et al. for these bands is a mis-citation. The earliest verifiable use of this trichotomy in the primary literature is Chaiton et al. (2007).
Two other banding conventions circulate, and a clinician comparing papers will meet them. A clinical-handout convention of 0-2 low, 3-4 moderate, 5-6 high appears on several widely distributed forms; no primary publication proposing it has been identified. A third convention of 0-1 low, 2-3 moderate, 4-6 high was used by Cooper et al. (2010). None of the three is criterion-validated.
What is anchored in the primary literature is a single dichotomous threshold: a score of 4 or more. It was benchmarked against the FTND cutoff of 6 or more in three independent samples, with sensitivity 94% and specificity 88% in a US and Spanish sample of 1,462, and sensitivity 79.5% and specificity 96.5% with a kappa of 0.74 in a French sample of 749. If you need one number from this instrument, use 4 or more.
Score
Severity
Interpretation
0–1
Low
Low cigarette dependence. Brief advice and behavioural support may be sufficient; pharmacotherapy is not automatically indicated on the basis of this score alone.
2–4
Medium
Moderate cigarette dependence. Consider pharmacotherapy alongside behavioural support. Note that the validated dichotomous threshold for high dependence - a score of 4 or more - sits inside this band.
5–6
High
High cigarette dependence. Pharmacotherapy is generally indicated, and relapse risk is concentrated in the first two weeks of a quit attempt.
The HSI is not a diagnostic test. It cannot diagnose tobacco use disorder or nicotine dependence on its own. If you have any concerns, discuss them with a healthcare professional.
03 - Validation evidence
How well it performs.
The HSI's validity rests almost entirely on how closely it tracks the full six-item Fagerström test. It does so well. In 1,642 smokers across five US and Spanish samples, the two HSI items at a threshold of 4 or more identified FTND-defined high dependence with 94% sensitivity and 88% specificity (de Leon et al., 2003). A French sample of 749 smokers found the same threshold gave 79.5% sensitivity and 96.5% specificity, with a kappa of 0.74 (Chabrol et al., 2005). A Spanish general-population survey of 1,655 daily smokers reported overall agreement of kappa 0.70 (Pérez-Ríos et al., 2009).
One caveat deserves to be stated plainly, because it is rarely mentioned: the HSI is not independently validated against the FTND. It is made of two of the six FTND items. Every agreement figure above is therefore inflated by criterion contamination, and is better read as a measure of how much information the other four items add than as evidence of validity in its own right (Charkazi et al., 2026).
The more persuasive evidence is predictive and biochemical. In the International Tobacco Control Four Country cohort, both HSI items independently predicted whether a quit attempt was maintained for at least a month, across three separate wave-to-wave replications and after adjustment for demographics; daily consumption also predicted whether an attempt was made at all, whereas time to first cigarette did not do so consistently (Borland et al., 2010). HSI scores also correspond to measured nicotine exposure, explaining 20% of the variance in blood cotinine, 11% in salivary cotinine and 13% in exhaled carbon monoxide in a sample of pregnant smokers (Kwok et al., 2014).
κ = 0.70
Agreement with the full FTND
94% / 88%
Detecting high dependence
r = .70
Stability over time
R² = 0.20
Correspondence with biochemistry
04 - How it compares
How it compares to the alternatives.
Instrument
Items
Time
When to reach for it
HSI
2
<1 min
The briefest usable index of cigarette dependence. Reach for it when you need a fast estimate of withdrawal-driven smoking to inform a pharmacotherapy decision, or when survey length is the binding constraint.
The parent instrument, renamed the Fagerström Test for Cigarette Dependence in 2012 — the same six items, unchanged. Prefer it for women and lighter smokers, where the HSI's sensitivity falls to around 62% and the extra four items carry real information.
Cigarette Dependence Scale (CDS-5 / CDS-12)
5 or 12
1-2 min
Reach for it when you want dependence as DSM or ICD frame it — craving, compulsion, loss of control — rather than consumption intensity. Note that it did not outperform the HSI for predicting cessation.
DSM-5 tobacco use disorder criteria
11 criteria
5-10 min, clinician
Use when a diagnosis and a severity specifier are needed for the record. The HSI is a severity index and cannot supply either.
No standard error of measurement, minimal detectable change or reliable change index has ever been published for the HSI, and it would be difficult to derive one. Classical precision statistics assume an internal-consistency reliability that a two-item index cannot meaningfully supply. What exists instead is test-retest evidence: scores correlated at r = .70 across a three-year gap, with each item alone at .63 or better (Borland et al., 2010).
Treat a one-point change with caution. Two structural features make it hard to interpret. The bands are unevenly wide — 0 to 1, 2 to 4, 5 to 6 — so the same one-point move may or may not cross a boundary depending on where it starts. And the index appears to compress at the top of the range, underestimating severity in the most dependent smokers (Charkazi et al., 2026), so points are not equal-interval. The HSI is a reasonable way to decide whether to offer pharmacotherapy. It is not a way to measure progress.
06 - Limitations
What it cannot tell you.
Floor effects in light smokers. In a population of relatively light smokers, 63% scored 0 or 1, and the authors concluded the index measured "little more than the number of cigarettes per day".
Cronbach's alpha is sometimes quoted for the HSI, but no primary source supports a figure, and internal consistency is close to meaningless for a two-item index in any case. The reliability evidence that does exist is test-retest.
Performance may be worse in women. A Spanish general-population survey found sensitivity of 83.1% in men but only 62.3% in women, missing nearly 40% of highly dependent female smokers (Pérez-Ríos et al., 2009). A French study of 749 smokers reported the opposite, finding the index performed as well in women as in men (Chabrol et al., 2005). The disagreement is unresolved and may reflect differences between workplace and general-population samples.
Cigarettes only. It is not validated for cigars, pipes, waterpipe, smokeless tobacco or e-cigarettes.
Weak prediction of cessation overall. Across studies it accounts for roughly 1% of variance in quit outcome, with an AUC around 0.56 in one large cancer-patient sample. Its predictive signal is concentrated in the first one to two weeks of a quit attempt and is not clearly present beyond a month.
The relationship between score and quit success is not monotonic. In a Canadian cohort of 2,938 smokers, both low and high scorers were more likely to have stopped at follow-up than those scoring in the middle (Chaiton et al., 2007), which complicates the intuition that a higher score simply means a worse prognosis.
Only seven possible scores, which makes it poor for tracking change within a patient.
Studies comparing the HSI with the FTND are subject to incorporation bias, since the HSI is two of the FTND's six items. High agreement between them is partly an artefact of shared content.
Not validated in adolescents, non-daily smokers, or poly-tobacco users. One Indian study found its utility limited where most smokers consume fewer than ten cigarettes a day.
No administration time has been published for this or any comparable instrument, so none is stated here.
See how Aisel removes friction where it costs most. A 20-minute walkthrough tailored to your clinic.
We value your privacy
We use cookies to analyse site usage and improve your experience. Analytics and embedded media (e.g. YouTube) only load if you accept. Read our cookie policy.