NIH Public Access
Author Manuscript
Addiction. Author manuscript; available in PMC 2010 August 3.
Published in final edited form as:
NIH-PA Author Manuscript
Addiction. 2008 May ; 103(5): 841–845. doi:10.1111/j.1360-0443.2008.02190.x.
A comparison of the Fagerström Test for Nicotine Dependence
and smoking prevalence across countries
Karl Fagerström1 and Helena Furberg2
1Smoker's Information Centre, Sweden
2Department of Genetics, University of North Carolina, Chapel Hill, NC, USA
Abstract
Aims—To examine the correlation between the Fagerström Test for Nicotine Dependence
(FTND) score and smoking prevalence across countries.
Design—Cross-sectional study.
NIH-PA Author Manuscript
Setting—Fifteen studies from 13 countries with FTND score data.
Participants—Samples of smokers were identified through systematic literature searches, web
queries and colleagues. Smokers were considered representative of their country's smoking
population if they were drawn from population-based sources, were not seeking smoking cessation
treatment and did not have significant comorbidities. Smoking prevalence data were derived from
the study itself or the country's population rate of daily smoking for the study year.
Measurements—A Pearson correlation coefficient was used to examine the direction and
magnitude of the correlation between FTND score and smoking prevalence across countries.
Findings—FTND scores ranged from 2.8 to 4.6. Smokers in Germany and Norway had the
lowest FTND scores, while smokers in Sweden and the United States had the highest FTND
scores. The prevalence of daily smoking in these countries was very different: 37% and 30% in
Germany and Norway, 19% and 16% in the United States and Sweden, respectively. An inverse
correlation towards higher FTND scores in countries with lower smoking prevalence was found (r
= -0.73, P = 0.001). Current smokers had higher FTND scores than former smokers.
Conclusions—The significant inverse correlation between FTND score and smoking prevalence
NIH-PA Author Manuscript
across countries and higher FTND score among current smokers supports the idea that remaining
smokers may be hardening. Less dependent smokers may quit more easily and remaining
dependent smokers may need more intensive treatment.
Keywords
Correlation; country; FTND score; hardening; nicotine dependence; smoking prevalence
INTRODUCTION
For some time, smokers have been pressured to stop smoking and it has been suggested that
current smokers might be more dependent than former smokers, as less dependent smokers
can quit more easily [1]. Termed the ‘hardening hypothesis’ it is currently being debated,
© 2008 The Authors. Journal compilation © 2008 Society for the Study of Addiction
Correspondence to: Karl Fagerström, Smoker's Information Centre, Berga Alle 1, Helsingborg 254 52, Sweden.
karl.fagerstrom@swipnet.se.
, Fagerström and Furberg Page 2
particularly in the United States, where the smoking prevalence dropped significantly over
the last decades but has plateaued in recent years [2,3]. Empirical and observational
evidence support that today's smoking population contain groups of smokers for whom it is
NIH-PA Author Manuscript
harder to quit, potentially because they have a higher degree of nicotine dependence. Irvin &
Brandon demonstrated that smoking quit rates have declined over time in the United States,
suggesting that today's smokers are more difficult to help [4]. Differences in personal
characteristics between current and former smokers have been reported where socio-
economic status and educational attainment are lower and psychiatric comorbidities are
higher among current smokers [2,5,6]. Consistent with this evidence, the first author (K. F.)
has observed a change in the smoking cessation patient population in Sweden since 1975.
With the continuous decline in cigarette smoking over time among Swedish men [7], he has
noticed that the remaining smokers are more likely to be single, less educated, of lower
income and report more alcohol consumption, chronic pain conditions, depression,
unemployment and sick leave. Finally, Hughes & Brandon [3] proposed that future smokers
might be more dependent and thereby contribute to a hardening of the smoking population
because the same social pressures that urge the less-dependent smokers to quit would
prevent those who find nicotine less reinforcing from starting smoking [8].
Contrary to the ‘hardening hypothesis’, O'Connor reported recently that the number of
cigarettes and nicotine intake had decreased in parallel when they compared two American
smoking populations in 1988–94 and 1999–2002 [9]. These data are inconsistent with the
NIH-PA Author Manuscript
hypothesis that the remaining population of smokers is becoming more dependent on
nicotine over time, as they are smoking fewer cigarettes and nicotine intake also decreased.
If there is such a thing as a ‘hardening target’ of smokers it should be examined by country,
as countries differ dramatically in their antismoking climates. In some countries, the
prevalence of cigarette smoking has just started to decline while it has more than halved in
countries such as the United States and Sweden. Antismoking climates around the world
have been assessed and countries in Europe with strong antismoking climates include
Ireland, the United Kingdom and Sweden, while the Mediterranean and German-speaking
countries generally have less developed antismoking climates [10,11].
In this report we present the available data on degree of dependence as assessed by the
Fagerström Test for Nicotine Dependence (FTND) [12] among samples of smokers from
around the world. The objectives of this report were: (i) to compile existing data on FTND
score and smoking prevalence by country; (ii) to calculate a correlation between FTND
scores and smoking prevalence across countries; and (iii) to examine whether patterns
observed in our data are consistent with the idea of a ‘hardening target’. We hypothesized
that in countries with low smoking prevalence, smokers would have higher FTND scores,
and that current smokers would have higher FTND scores than former smokers.
NIH-PA Author Manuscript
METHODS
We identified 15 samples of smokers with FTND score data from 13 countries through
colleagues, the US National Library of Medicine, the SRNT Listserve and Psych Info. To be
included in the analysis, samples of smokers had to contribute FTND score data and be
representative of their country's general smoking population; studies among specific
subgroups of smokers including psychiatric patients, smoking cessation trial participants or
smokers with significant comorbidities were excluded. Most of the studies we included have
been published formally as independent research papers, while other studies have been
described previously and used in a similar analysis [1].We chose to include data from three
twin registries as they are population-based, and twins exhibit similar means, frequencies
and prevalences to singletons for many traits and adult diseases [13].
Addiction. Author manuscript; available in PMC 2010 August 3.
