![]() |
||||||
|
||||||
Tobacco smoking prevalence in Pacific Island
countries and territories: a review
Kumanan Rasanathan, Colin F Tukuitonga
The global burden of disease attributable to tobacco use
continues to increase, particularly in low and middle income
countries.1 Unlike other low and middle income
countries in Asia and Africa, most Pacific Island Countries and Territories
(PICT) have long been in the epidemiological transition, with non-communicable
disease comprising the majority of the disease
burden.2,3
Cardiovascular disease is the major cause of mortality in
PICT, as in other low and middle income
countries,4–6 and rates of cancer also
appear to be increasing. Recent estimates of the population attributable
fraction of cardiovascular disease in different PICT due to smoking range, in
male populations, from 11% in Palau to 33% in Kiribati, and in female
populations, from 0.4% in Federated States of Micronesia (FSM) to 32% in
PNG.7 Tobacco use is arguably thus the most
important modifiable risk factor for disease in PICT.
Despite this, there is a lack of robust information about
trends in the prevalence of tobacco smoking in PICT. The availability and
quality of data varies greatly. Whilst estimates for smoking prevalence in PICT
are available in World Health Organization (WHO)
publications,8 these publications do not
necessarily include the most up-to-date sources of information and some
estimates date back to the 1980s. The only review of smoking rates in PICT
evident in the peer-reviewed journal literature appeared in 1986 and this paper
reported on only eight PICT.9
This paper aims to comprehensively review the literature to
report on the current smoking prevalence in both adult and youth populations in
the 22 PICT (shown in Table 1) served by the Secretariat of the Pacific
Community (SPC) and included in the Western Pacific Region of the WHO. This
paper also attempts to consider trends in tobacco use in PICT over the last 30
years and what these trends suggest for future health promotion efforts in the
Pacific region to decrease the burden of disease related to smoking.
It should be noted that while this paper concentrates on
smoking, chewing tobacco and other forms of tobacco use are important in some
PICT, for example in Palau.10
Table 1. Pacific Island countries and
territories*
![]() MethodsThe MEDLINE database was searched for the period
January 1986 to December 2006, using the search terms (‘tobacco’ or
‘smoking’) and (‘Pacific’ or ‘Oceania’ or
the names of the 22 PICT). The WHO’s Infobase (http://www.who.int/ncd_surveillance/infobase/web/InfoBaseCommon/)
and Tobacco-Free Initiative (http://www.who.int/tobacco/en/index.html)
databases were also searched along with the Centres for Disease Control and
Prevention (CDC)’s Youth Risk Behaviours Surveillance System (http://www.cdc.gov/healthyyouth/yrbs/index.htm)
and Global Youth Tobacco Survey (http://www.cdc.gov/Tobacco/global/GYTS.htm).
Unpublished reports were found within the WHO and CDC
databases and also sought by communication with tobacco control personnel in the
region. The aim was to find all estimates of smoking prevalence in PICT that
used random sampling techniques such as to provide nationally representative
estimates that could be used to compare between PICT (for adults, smoking was
defined as daily use, and for youth, smoking was defined as current use).
The most recent such estimate was used to assess
smoking prevalence for each PICT in adult and youth populations, with estimates
prior to 1990 not considered for inclusion as a current estimate. Time trends in
tobacco use in the Pacific were considered by comparing current national
estimates in Cook Islands, Niue, Fiji, Kiribati, Nauru, New Caledonia, and Samoa
with those reported in the aforementioned 1986
paper9 which derived estimates from surveys
conducted between 1975 and 1981. Prevalence estimates were also obtained from
Australia and New Zealand for comparison.
ResultsThe MEDLINE search yielded 10 relevant
papers.9,11–19 Searches of the WHO and
CDC databases and bibliographies identified a further 15 relevant documents and
resources.8,20–33 The surveys identified
used differing methodologies and were undertaken over a range of years. As such,
there are limitations to their comparability between different PICT and between
different timepoints within the same state. However, the surveys utilised all
aimed to produce nationally representative estimates.
Adult smoking prevalencePost-1990 estimates for smoking prevalence were identified
for American Samoa (survey year 2004),33 Cook
Islands (2004),33 FSM (but limited to Kosrae,
1994),14 Fiji
(2002),33 French Polynesia
(1995),8 Guam
(2003),17 Kiribati
(1999),8 Nauru
(2004),33 New Caledonia
(1992),8 Niue
(2002),29 Palau
(1998),22 Papua New Guinea
(1990),8 Samoa
(2004),33 Tokelau
(2005),33 Tonga
(1998),15 Vanuatu
(1998),21 and Wallis and Futuna
(1996).8 These estimates are summarised in
Figure 1 along with comparisons from Australia
(2001)34 and New Zealand
(2006).35 Suitable estimates were not found for
Marshall Islands, Northern Mariana Islands, Pitcairn Island, Solomon Islands,
and Tuvalu.
Survey definitions of adult varied. WHO STEPS data
(discussed further below) was the source of estimates for American Samoa, Cook
Islands, Fiji, Nauru, Samoa, and Tokelau. These surveys defined adult as
25–65 years old. The adult estimates for FSM were derived from 20–85
year olds; for French Polynesia, Niue, Papua New Guinea and Wallis and Futuna,
from those 15 years and over; for Guam, 18 years and over; for Kiribati, 16
years and over; and for New Caledonia, Palau, Tonga and Vanuatu, from those aged
20 years and over. The estimates for smoking prevalence in Australia included
those aged 18 years and over, and for New Zealand, those aged 15 years and
over.
Adult smoking prevalence varied from 0.6% in FSM (Kosrae)
women and 5% in Vanuatu women to 53% in Tongan men, 57% in Kiribati men, and 51%
in Nauru women. Despite this great variation, in general, more men in PICT
smoked than women with the exception of New Caledonia and Nauru, where more
women smoked than men, and Tokelau and French Polynesia, where there was no
gender difference.
Reported estimates of male smoking prevalence in PICT were
all higher than recent estimates from Australia and New Zealand, with the
exception of Palau. However, many PICT reported lower levels of male smoking
prevalence than the age-standardised estimate of 41.3% reported in male Pacific
peoples in New Zealand.35
There was much greater variation in female smoking
prevalence estimates in PICT than in male prevalence estimates. Many PICT had
lower levels of reported smoking in women than in New Zealand and Australia, in
particular, FSM (Kosrae), Vanuatu, Palau, and Tonga.
Figure 1. Adult smoking prevalence in the
Pacific8,14,15,17,21,22,29,33–35
![]() In contrast, PICT such as Nauru, Tokelau, French Polynesia,
New Caledonia, and Kiribati reported much higher levels of smoking for females
than in New Zealand and Australian women. Most PICT reported lower prevalence of
smoking in women than the age-standardised estimate of 33.8% in Pacific women in
New Zealand.35
Age-stratified estimates were available for
Fiji,27 Niue, Palau, Samoa, Tonga (in a
previous study20 differing to the one used for
the most recent country estimate), Nauru (again in a previous
study30), and Vanuatu.
In general, there were not major differences between age
groups, except for lower levels of smoking in elderly age groups (>65 years)
in Nauru, Niue, and Tonga.
Youth smoking prevalenceRecent estimates for youth smoking rates were available for
fewer PICT than for adult smoking levels. Post-1990 estimates were available for
only American Samoa (survey year 1997),31 Cook
Islands (2003),28 Fiji
(1999),23 Guam
(1997),32 Marshall Islands
(2000),18 Niue
(2002),29 Northern Mariana Islands
(2000),24 Palau
(2001),26 and Papua New Guinea
(1990).8 These estimates are summarised in
Figure 2 along with youth smoking rates in New Zealand
(2005)35 and Australia
(2005).36 Estimates were also available for FSM
but these were not stratified by gender but rather undertaken separately for the
islands of Kosrae (41.2% current smoking prevalence in 13–15 year old
students) and Pohnpei (35.1% prevalence in 13-15 year old
students).25
Figure 2. Youth smoking prevalence in the
Pacific8,18,23,24,26,28,29,31,32,35,36
![]() Most of the youth prevalence rates identified were derived
from the Global Youth Tobacco Survey (developed by CDC and WHO) and focused on
adolescents aged 13 to 15 years old or from the Youth Risk Behaviours
Surveillance System (also CDC) for American territories which covered
13–17 year old students.
The Marshall Islands’ estimate covered 11–17
year old students, the Niue estimate was derived from 15–17 year old
adolescents, and the Papua New Guinea estimate included 10–15 year old
young people. The comparison estimates for Australia included 12–17 year
olds students and for New Zealand, 14–15 year old students.
Due to this variability, the youth estimates are less
comparable between PICT than the adult estimates for smoking prevalence
described above.
Palau, Northern Mariana Islands, Guam, Cook Islands, and
American Samoa reported very high rates of youth smoking in both males and
females, significantly higher than estimates for young people in New Zealand and
Australia. These estimates were also higher than those reported for young
Pacific people in New Zealand—10.2% for males and 14.5% for
females.35
All PICT with youth estimates showed similar rates for male
and female young people with the exception of Marshall Islands and Fiji, which
reported much higher rates in young males.
Time trends in smoking prevalenceComparisons of recent national estimates in Cook Islands,
Niue, Fiji, Kiribati, Nauru, New Caledonia, and Samoa with estimates from
1975–1981 reported by Tuomilehto et al9
are presented for adult male populations in Figure 3 and adult female
populations in Figure 4. The older estimates included as adult those aged 20
years and over, except for the Nauru estimate which included those aged 15 years
and over. Older estimates of youth smoking prevalence were not identified.
Figure 3. Adult male smoking prevalence (%)
time trend8,9,29,33
![]() *For Cook Islands, Niue, and Nauru, a single pre-1982
estimate was identified and thus this figure is used for both urban and rural
populations.
There are significant difficulties in comparing estimates in
this way due to the differing methodologies employed in these surveys. However,
smoking rates appeared to have decreased in most male populations. Time trends
for female populations were variable with increasing rates noted in Cook Island
and New Caledonian women.
Figure 4. Adult female smoking prevalence (%)
time trend8,9,29,33
*For Cook Islands, Niue, and Nauru,
a single pre-1982 estimate was identified and thus this figure is used for both
urban and rural populations.DiscussionPICT continue to show very high levels of tobacco smoking,
with the exception of a few female PICT populations. Whilst there is some
evidence that adult male smoking rates in PICT have decreased in the last 30
years, smoking rates remain high and there are concerning levels of smoking in
youth populations.
Overall, rates of smoking in PICT are higher than those seen
in neighbouring high income countries such as Australia and New Zealand,
although similar to prevalence for Pacific peoples in New Zealand.
These findings are concerning given the high burden of
cardiovascular disease in PICT and increasing rates of cancer. In the context of
struggling health systems, action to combat tobacco use is thus a major priority
to improve the health of populations in PICT.
There is a lack of comprehensiveness and rigour to some of
the surveys included above from which estimates are derived. As such, some
estimates may not be representative for all parts of the nations they derive
from. In some PICT, it is very difficult to carry out genuinely national surveys
given the multiple islands that these states contain. Furthermore, there are
vast differences between populations within many PICT and thus these national
figures may mask similar difference in tobacco use. However, this criticism
potentially applies to national figures for all countries.
A more pressing concern is the age of some of the estimates,
and the hitherto lack of regular monitoring of smoking prevalence using
consistent methodology. As such, it is difficult to be sure of the time trends
in smoking in PICT and the apparent decrease seen in adult male smoking
prevalence should not be cause for complacency given the high rates of youth
smoking and the latency of much smoking-related harm.
Many of these concerns about the availability and quality of
smoking prevalence data in PICT will be addressed by the implementation and
publication of WHO’s STEPS survey for non-communicable disease risk
factors, which includes tobacco use.37
STEPS surveys have already been carried out in many PICT.
STEPS data for tobacco use in adults is available for Fiji, American Samoa,
Samoa, Tokelau, Nauru, and Cook Islands, and these estimates are used above.
However, currently Fiji is the only PICT to have published a final report based
on the STEPS survey results.
Data has also been collected in Marshall Islands, FSM,
Vanuatu, Tonga, Solomon Islands, and Kiribati33
though it is unclear when results will be available. As such, the availability
and comparability of smoking prevalence data in PICT should improve greatly in
the medium term. It is thus imperative that all PICT publish and disseminate
results of STEPS surveys and continue to undertake smoking surveys using STEPS
methodology at regular intervals to enable robust monitoring of smoking rates to
inform tobacco control efforts.
Despite the reservations about the estimates reviewed in
this paper, all of the reported surveys employed randomised sampling aimed at
producing a representative estimate and/or have been recognised by WHO as
providing a national estimate. Whilst there should be some caution exercised in
interpreting differences between PICT, due to the differing methodologies used,
the estimates provide the best indication available of tobacco smoking
prevalence in these states.
Most importantly, given the large magnitude of smoking
prevalence in many PICT, concerns about methodology should not obscure
recognition of the scope of the continuing threat that tobacco poses to public
health in these countries and territories and of the need for further action.
All PICT eligible to do so have become parties to the
Framework Convention on Tobacco Control
(FCTC)38 and have pledged to make progress on
legislation and policy consistent with the FCTC at the meetings of Pacific
Health Ministers in Tonga in 2003 and Samoa in
2005.39,40 This is encouraging but PICT need to
make urgent progress on action to reduce tobacco use in their populations.
In particular, tobacco taxation remains an underutilised
tool with rates of tax often much lower than in countries such as Australia and
New Zealand. Indeed, tobacco taxation has been partially responsible for the
decline in tobacco use in many high income countries such as New
Zealand41 and it has been suggested that
tobacco taxation is even more effective in low and middle income countries,
judging from Papua New Guinea data.42
The findings reviewed here thus suggest that PICT should
proceed with measures such as taxation and other tobacco control legislation and
policy, as detailed in the FCTC, at the same time as they carry out improved
monitoring of tobacco smoking rates in their states.
The importance of implementing STEPS monitoring of tobacco
use will lie not in establishing the problem of tobacco use but rather in
evaluating the effectiveness of tobacco control measures. Without such immediate
action, the burden of disease from cardiovascular disease and cancer in PICT is
likely to continue to increase at great cost to the peoples of the
Pacific.
Competing interests: None.
Author information: Kumanan Rasanathan,
Honorary Senior Lecturer;
Colin F Tukuitonga, Associate Professor; Section of Pacific and International Health, School of Population Health, University of Auckland, Auckland Correspondence: Dr Kumanan Rasanathan,
Section of Pacific and International Health, School of Population Health,
University of Auckland, Private Bag 92019, Auckland, New Zealand. Fax: +64 (0)9
373 7624; email: k.rasanathan@auckland.ac.nz
References:
|
||||||
| Current
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads) Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals |