[Show abstract][Hide abstract] ABSTRACT: To study whether passive smoking is a risk factor for aortic arch calcification (AAC) among never smokers.
We have previously reported that active smoking increases the risk of AAC, but the effect of passive smoking has not been reported.
We used baseline data of the Phase 1 Guangzhou Biobank Cohort Study (GBCS). 7702 older Chinese never smokers from the Phase 1 GBCS were included. Information on passive smoking and potential confounders were collected by standardized interviews and laboratory assays. AAC was diagnosed from chest X-ray by two experienced radiologists. Unconditional logistic regression was used to estimate odds ratios of AAC for passive smoking with adjustment for potential confounders.
In women, the risk for aortic arch calcification (AAC) increased significantly with increasing duration of adulthood passive smoking exposure at home, at work and total duration of adulthood home and work exposure [adjusted odds ratio 1.24 (95% confidence interval 1.09-1.41) for high level of total exposure] (P for trend from 0.012 to 0.001). For passive smoking at home, at work and total exposure, significant trends of increasing severity of AAC with increasing duration of exposure were observed in men and women combined (P for trend from 0.05 to 0.002).
Passive smoking is a risk factor for aortic arch calcification. Studies of passive smoking and AAC, especially in developing countries can generate important local evidence to raise awareness and to support public health measures to protect non-smokers from second-hand smoke.
International journal of cardiology 11/2009; 148(2):189-93. DOI:10.1016/j.ijcard.2009.11.001 · 4.04 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Smoking cessation programs are well established in the West, but reports on smoking cessation clinics (SCCs) from China are lacking. On the basis of the Hong Kong experience and with strong support from Guangzhou Health Bureau, we established the first SCC in Guangzhou, China. The objective was to describe the characteristics of smokers, measure quit rates and examine predictors of successful quitting.
During 2006-08, 220 smokers received individual counseling following the five A's and five R's. No medications were used.
At baseline, the mean (SD) age was 40 (14) years. Most (96%) were males, married (73%), currently employed (75%), college educated or above (54%); 77% had previous quitting attempts. By 14 May 2008, 195 reached the 6 months follow-up period. Of them, 79% (151/195) were successfully followed up, and 46 had quit. By intention to treat, the 6-month 7-day point prevalence quit rate was 24% [95% confidence interval (CI) 18-30%]. Smokers with more confidence in quitting or were at action stage were more successful in quitting with adjusted odds ratio of 2.39 (95% CI 1.01-5.30) and 5.50 (95% CI 1.08-28) respectively.
A pilot-model clinic free of charge and with systemic data collection, follow-up and evaluation should be a starting point for smoking cessation program in low-income countries.
Journal of Public Health 11/2009; 32(2):267-76. DOI:10.1093/pubmed/fdp107 · 2.04 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Prior pulmonary TB has been shown to be associated with a higher risk of airflow obstruction, which is the hallmark of COPD, but whether smoking modifies this relationship is unclear. We investigated the relationships between prior TB, smoking, and airflow obstruction in a Chinese population sample.
Participants in the Guangzhou Biobank Cohort Study underwent spirometry, chest radiography, and a structured interview on lifestyle and exposures. Prior TB was defined as the presence of radiologic evidence suggestive of inactive TB. Airflow obstruction was based on spirometric criteria.
The prevalence of prior TB in this sample (N = 8,066, mean age: 61.9 years) was 24.2%. After controlling for sex, age, and smoking exposure, prior TB remained independently associated with an increased risk of airflow obstruction (odds ratio = 1.37; 95% CI, 1.13-1.67). Further adjustment for exposure to passive smoking, biomass fuel, and dust did not alter the relationship. Smoking did not modify the relationship between prior TB and airflow obstruction.
Prior TB is an independent risk factor for airflow obstruction, which may partly explain the higher prevalence of COPD in China. Clinicians should be aware of this long-term risk in individuals with prior TB, irrespective of smoking status, particularly in patients from countries with a high TB burden.
[Show abstract][Hide abstract] ABSTRACT: The aim of the study was to compare the impact of impaired fasting glucose (IFG) and impaired glucose tolerance (IGT) on vascular function among older Chinese people. A random sample of 671 men and 603 women aged 50 to 85 years without known diabetes from the Guangzhou Biobank Study-CVD was examined in a cross-sectional study. Subjects with no previously confirmed or treated diabetes but with both fasting plasma glucose less than 5.6 mmol/L and 2-hour glucose from 7.8 to less than 11.0 mmol/L were classified as having isolated IGT, and those with no previously confirmed and treated diabetes but with both fasting plasma glucose from 5.6 to less than 7.0 mmol/L and 2-hour glucose less than 7.8 mmol/L were classified as having isolated IFG. A total of 11.0% of the men and 8.6% of the women had isolated IFG, and 17.7% of the men and 18.6% of the women had isolated IGT. The brachial-ankle pulse wave velocity and pulse pressure were increased in both the isolated IFG and isolated IGT subjects compared with the normoglycemia group (both Ps < .001). Compared with subjects with isolated IFG, those with isolated IGT appeared to have a higher age- and sex-adjusted brachial-ankle pulse wave velocity (1543 +/- 22 vs 1566 +/- 17, P = .07) and to be more insulin resistant (2-hour postload insulin: 54.2 +/- 2.13 vs 26.8 +/- 2.99 muU/mL, P < .001), had a worse lipid profile (apolipoprotein [apo] B: 1.07 +/- 0.02 vs 0.97 +/- 0.02 g/L, P < .001; apo B/apo A-1 ratio: 0.80 +/- 0.02 vs 0.69 +/- 0.02, P < .001), but had lower glycosylated hemoglobin levels (6.03% +/- 0.06% vs 5.86% +/- 0.04%, P < .001) (values are mean +/- SE). Subjects with isolated IGT had greater arterial stiffness, probably as a result of being more insulin resistant, with a worse lipid profile than those with isolated IFG. The sole use of fasting glucose level to identify prediabetic people would fail to identify a significant proportion of the at-risk population.
[Show abstract][Hide abstract] ABSTRACT: Smoking has been shown to be associated with carotid atherosclerosis in cross-sectional and prospective studies in Western populations. However, few studies have examined the reversal of risk resulting from quitting smoking, and the results are conflicting.
959 men aged 50-85 years were randomly selected from phase III (2006-2007) of the Guangzhou Biobank Cohort Study into this cross-sectional study. Common carotid artery intima-media thickness (CCA-IMT) was measured by B-mode ultrasonography, and carotid artery plaques were identified. Major cardiovascular risk factors, including fasting triglyceride, low-density and high-density lipoprotein (LDL and HDL) cholesterol and glucose, and systolic and diastolic blood pressure, were assessed.
CCA-IMT and the number of carotid plaque increased from never to former to current smokers (both p≤0.001). Among former smokers compared to current smokers, after adjustment for cigarette pack-years and other potential confounders, the adjusted ORs (95% CI) for quitting for 1-9, 10-19 and 20+ years were 0.77 (0.47 to 1.26), 0.45 (0.26 to 0.79) and 0.37 (0.17 to 0.77) for the presence of CCA atherosclerosis, and 0.69 (0.43 to 1.12), 0.47 (0.27 to 0.82) and 0.45 (0.23 to 0.96) for the presence of carotid plaques, respectively. Longer duration of quitting smoking was also significantly associated with decreasing risk of the severity of CCA atherosclerosis and carotid plaques (all p≤0.001).
Smoking cessation was beneficial in attenuating the risk of carotid atherosclerosis associated with cigarette smoking. The short duration of cessation in earlier studies is a likely explanation for the inconsistent results.
Journal of epidemiology and community health 10/2009; 64(11):1004-9. DOI:10.1136/jech.2009.092718 · 3.50 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: This study investigated the bidirectional relationships of adolescents' and maternal mood, and the moderating effect by gender and perceived family relationships on these relationships. Data were obtained from 626 adolescent-mother dyads and follow-up data were collected one year later from a subset. Adolescents reported their depressive symptoms, and their mothers reported their negative affect. Adolescents described their perception of family relationships. Maternal negative affect and adolescents' depressive symptoms were significantly correlated at baseline. This association was moderated by gender and family relationships. The association was stronger in mother-daughter compared to mother-son dyads. In families where relationships were reported to be poor, adolescent depressive symptoms were uniformly high, regardless of maternal negative affect. However, in families where relationships were good, maternal negative affect was associated with higher adolescents' depressive symptoms. In longitudinal analyses, adolescents' mood at baseline was found to relate to maternal negative affect at follow-up. Family relationships at baseline were also associated with adolescents' depressive symptoms at follow-up. However, there was no prediction from maternal negative affect at baseline to adolescents' depressive symptoms at follow-up. Gender and quality of family relationships did not moderate the longitudinal relationships between adolescents' depressive symptoms and maternal negative affect in either direction.
Journal of Family Psychology 10/2009; 23(5):636-45. DOI:10.1037/a0016379 · 1.89 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Dyslipidaemia commonly coexists with diabetes. We investigated the association of lipid profiles with glucose levels in populations of Asian origin without a prior history of diabetes.
Cross-sectional data of 10,374 men and 12,552 women aged 30-74 years from 14 cohorts, representing seven populations of Asian origin were jointly analysed. Multivariable adjusted linear regression analyses with standardized regression coefficients (beta) were performed to estimate relationships between lipids and plasma glucose.
Within each glucose category, fasting plasma glucose (FPG) levels were correlated with increasing levels of triglycerides (TGs), total cholesterol (TC), TC to high-density lipoprotein (HDL) ratio and non-HDL cholesterol (non-HDL-C) (p < 0.05 in most of the ethnic groups) and inversely associated with HDL-C (p < 0.05 in some, but not all, of the populations). The association of lipids with 2-h plasma glucose (2hPG) followed a similar pattern as that for the FPG, except that an inverse relationship between HDL-C and glucose was more commonly observed for 2hPG than for FPG among different ethnic groups.
Hyperglycaemia is associated with adverse lipid profiles in Asians without a prior history of diabetes. The 2hPG appears to be more closely associated with lipid profiles than does FPG. When assessing the risk of cardiovascular disease, the association of the dyslipidaemia with intermediate hyperglycaemia needs to be considered.
Diabetes/Metabolism Research and Reviews 09/2009; 25(6):549-57. DOI:10.1002/dmrr.994 · 3.55 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: To assess whether there is a statistical interaction between smoking and diabetes that is related to the risk of cardiovascular disease (CVD) in men in the Asia Pacific region.
An individual participant data meta-analysis was conducted on 34 cohort studies, involving 16 492 participants with diabetes (47.4% smokers) and 188 897 without (47.6% smokers). Hazard ratios (HR) and 95% confidence intervals (CI) were calculated for smoking (stratified by study and adjusted for age) for those with and without diabetes.
In men with diabetes, the HR (95% CI) comparing current smokers with non-smokers was 1.42 (1.10-1.83) for coronary heart disease, 1.10 (0.88-1.37) for total stroke and 1.15 (0.98-1.35) for total CVD. The corresponding figures for men without diabetes were 1.47 (1.33-1.61), 1.27 (1.16-1.39) and 1.35 (1.27-1.44), respectively. There was no evidence of a statistical interaction between diabetes and current smoking, the number of cigarettes smoked per day or quitting smoking. Smoking cessation was associated with a 19% reduction in CVD risk, irrespective of diabetes status.
The effects of cigarette smoking and smoking cessation are broadly similar in men with and without diabetes. In Asia, where there are high rates of smoking and a rapidly increasing prevalence of diabetes, strategies that encourage smokers to quit are likely to have huge benefits in terms of reducing the burden of CVD in men with diabetes.
Journal of Diabetes 09/2009; 1(3):173-81. DOI:10.1111/j.1753-0407.2009.00028.x · 1.93 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Smoking and alcohol are major causal factors for upper aerodigestive tract cancer, but reliable quantification of the combined impact of smoking and alcohol on this cancer and its major subtypes has not been performed.
A meta-analysis of studies that had published quantitative estimates of smoking and upper aerodigestive tract cancer by January 2007 was performed. Pooled estimates of relative risks were obtained. Publication bias was investigated through funnel plots and corrected if found to be present.
Overall, 85 studies with information on 53,940 individuals with upper aerodigestive tract cancer were included. The pooled estimate for the association between smoking and the risk of this cancer was 3.47 (95% confidence interval, 3.06-3.92). The risk remained elevated for a decade after smoking cessation but declined thereafter. Individuals who both smoked and consumed alcohol had double the risk of upper aerodigestive tract cancer in comparison with those who only smoked: the relative risk was 6.93 (95% confidence interval, 4.99-9.62) for the former and 2.56 (95% confidence interval, 2.20-2.97) for the latter (P < 0.001).
Public health interventions that simultaneously discourage smoking and heavy drinking would have greater benefits than would be expected from those that target only one of these risk factors.
Mount Sinai Journal of Medicine A Journal of Translational and Personalized Medicine 08/2009; 76(4):392-403. DOI:10.1002/msj.20125 · 1.62 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: There is some evidence that chronic obstructive pulmonary disease (COPD) and metabolic syndrome may be related, perhaps through systemic inflammation, which is common to both. However, the association between the two conditions has not yet been clearly shown. The present study involved 7,358 adults aged > or =50 yrs from a population-based survey who underwent spirometry, a structured interview and measurement of fasting metabolic marker levels. Airflow obstruction (forced expiratory volume in 1 s/forced vital capacity ratio of less than the lower limit of normal) was present in 6.7%, and the International Diabetes Federation metabolic syndrome criteria were met by 20.0%. The risk of metabolic syndrome was higher in those with airflow obstruction than in those without (odds ratio (OR) 1.47; 95% confidence interval (CI) 1.12-1.92), after controlling for potential confounders. Of the five components of metabolic syndrome, only central obesity was significantly associated with airflow obstruction (OR 1.43; 95% CI 1.09-1.88) after adjusting for body mass index. A similar association was observed in both never and current smokers. In this Chinese sample, airflow obstruction was associated with metabolic syndrome, and, in particular, its central obesity component. This may help explain the increased risk of cardiovascular diseases in COPD, and so could guide future clinical practice.
European Respiratory Journal 08/2009; 35(2):317-23. DOI:10.1183/09031936.00024709 · 7.64 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Observational studies from mainly Western settings suggest breastfeeding may protect against childhood adiposity; however, breastfeeding and adiposity share social patterning potentially generating confounding, making evidence from other settings valuable.
We used multivariable linear regression to examine the prospective adjusted associations of breastfeeding with body mass index (BMI), height and weight z-scores at 7 years of age relative to the 2007 World Health Organization (WHO) growth reference, in a large (n = 8327), population-representative Hong Kong Chinese birth cohort, recruited between April and May 1997 with high follow-up (n = 7026).
Low socio-economic position (SEP) was associated with never breastfeeding and with exclusive breastfeeding for > or =3 months. We did not find any association between breastfeeding and BMI [z-score mean difference 0.07, 95% confidence interval (CI) -0.05 to 0.19], height (0.02, 95% CI -0.07 to 0.11) or weight (0.07, 95% CI -0.05 to 0.18), adjusted for sex, birth weight, gestational age, SEP, second-hand smoke (SHS) exposure, parity, mother's age at birth, mother's place of birth and serious infant morbidity.
In a non-European setting, breastfeeding was not associated with child adiposity, suggesting that observed protective effects may be due to socially patterned confounding by SEP, maternal adiposity and maternal smoking.
International Journal of Epidemiology 08/2009; 39(1):297-305. DOI:10.1093/ije/dyp274 · 9.18 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: The Guangzhou Biobank Cohort Study (GBCS, n=30 519, age >or=50 years) was established to examine the effects of genetic and environmental influences on health problems and chronic disease development. Guangzhou is undergoing massive economic development, but from a baseline that had remained unchanged for millennia. The Cardiovascular Disease Subcohort (GBCS-CVD) consists of 2000 participants who have been intensively phenotyped including a range of surrogate markers of vascular disease, including carotid artery intima-media thickness, cerebral artery stenoses, arterial stiffness, ankle-to-brachial blood pressure index and albuminuria, as well as coagulatory and inflammatory markers. Plasma and leukocytes are stored in liquid nitrogen for future studies. Preliminary demographic data show the female volunteers are younger than the male ones, but present with greater levels of adiposity including central obesity (31 vs 16%). Women had more body fat (33 vs 24%) and associated levels of adipokines. Despite this, body mass index and hip circumferences were similar, which contrasts with Caucasian populations. Men had more physician-diagnosed vascular disease (6.1 vs 2.5%), hypertension (42 vs 34%) and hyperglycaemia (36.6 vs 29.6%) than the women, but were less insulin resistant. In men, smoking (40 vs 2%) and drinking alcohol (67 vs 50%) was more common and they also had lower energy expenditures. The genotype distributions of the 15 typed single nucleotide polymorphisms were all in Hardy-Weinberg equilibrium. This article describes the rationale and methodology for the study. Given the comprehensive characterization of demographic and psychosocial determinants and biochemistry, the study provides a unique platform for multidisciplinary collaboration in a highly dynamic setting.
Journal of human hypertension 07/2009; 24(2):139-50. DOI:10.1038/jhh.2009.52 · 2.70 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: A U-shaped relation between alcohol use and mortality may be due to biological hormesis, differential response (i.e., effect modification) by health status or changes in alcohol use with ill-health and aging (i.e., reverse causality).We examined whether alcohol had the same association with mortality in healthy and unhealthy older people.
We used Cox regression analysis to examine the association of alcohol with mortality by health status in a population-based cohort of 56,167 people (65+ years), enrolled during July 1998 to December 2000 in Hong Kong.
After a mean follow-up of 4.1 years, there were 3,819 deaths in 54,087 subjects. Adjusted for age, socioeconomic position and lifestyle, both occasional and moderate types of alcohol use were associated with lower mortality compared to never-drinkers, relative risk (RR) 0.72 (95% confidence interval [CI] 0.63-0.83) and 0.73 (95% CI 0.59-0.90) in men and 0.77 (95% CI 0.64-0.94) and 0.54 (95% CI 0.29-1.01) in women. However, these associations were not maintained in those with good health status: RR 1.02 (95% CI 0.74-1.39) and 1.09 (95% CI 0.71-1.68) in men and 0.63 (95% CI 0.36-1.12) and 1.27 (95% CI 0.40-4.01) in women.
Moderate alcohol use may only be beneficial for older adults in poor health.
Annals of epidemiology 07/2009; 19(6):396-403. DOI:10.1016/j.annepidem.2009.01.011 · 2.00 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Dyslipidaemia is a major risk factor for cardiovascular disease and is only detectable through blood testing, which may not be feasible in resource-poor settings. As dyslipidaemia is commonly associated with excess weight, it may be possible to identify individuals with adverse lipid profiles using simple anthropometric measures. A total of 222 975 individuals from 18 studies were included as part of the Obesity in Asia Collaboration. Linear and logistic regression models were used to assess the association between measures of body size and dyslipidaemia. Body mass index, waist circumference, waist : hip ratio (WHR) and waist : height ratio were continuously associated with the lipid variables studied, but the relationships were consistently stronger for triglycerides and high-density lipoprotein cholesterol. The associations were similar between Asians and non-Asians, and no single anthropometric measure was superior at discriminating those individuals at increased risk of dyslipidaemia. WHR cut-points of 0.8 in women and 0.9 in men were applicable across both Asians and non-Asians for the discrimination of individuals with any form of dyslipidaemia. Measurement of central obesity may help to identify those individuals at increased risk of dyslipidaemia. WHR cut-points of 0.8 for women and 0.9 for men are optimal for discriminating those individuals likely to have adverse lipid profiles and in need of further clinical assessment.
[Show abstract][Hide abstract] ABSTRACT: Chronic kidney disease is a major worldwide public health problem that causes substantial morbidity and mortality. Studies from the Asia-Pacific region have reported some of the highest chronic kidney disease prevalence rates in the world, but access to dialysis is limited in many countries, making it imperative to identify high-risk individuals. We performed a participant-level data overview of prospective studies conducted in the Asia-Pacific region to quantify the magnitude and direction of the associations between putative risk factors and renal death. Age- and sex-adjusted Cox proportional hazards models were applied to pooled data from 35 studies to calculate hazard ratios (95% CIs) for renal death associated with a standardized change in risk factors. Among 560 352 participants followed for a median of 6.8 years, a total of 420 renal deaths were observed. Continuous and positive associations among systolic blood pressure, diastolic blood pressure, fasting blood glucose, and total cholesterol levels with renal death were observed, as well as a continuous but inverse association with high-density lipoprotein cholesterol. Systolic blood pressure was the strongest risk factor for renal death with each SD increase in systolic blood pressure (19 mm Hg) associated with >80% higher risk (hazard ratio: 1.84; 95% CI: 1.60 to 2.12). Neither cigarette smoking nor excess weight was related to the risk of renal death (P>0.10). The results were similar for cohorts in Asia and Australia. These results suggest that primary prevention strategies for renal disease should focus on individuals with elevated blood pressure, diabetes mellitus, and dyslipidemia.
[Show abstract][Hide abstract] ABSTRACT: Although the J-shaped association between alcohol consumption and blood pressure (BP) is well known, the effect of alcohol sensitivity on this relationship is less clear. We studied the association of alcohol sensitivity and alcohol use with BP and hypertension. This cross-sectional analysis included 19 335 older participants from the Guangzhou Biobank Cohort Study recruited from 2003 to 2006, using clinically measured BP and self-reported alcohol use and alcohol sensitivity. Alcohol use was rare in women, in whom light-to-moderate drinkers (<140 g ethanol per week) without alcohol sensitivity had lower systolic and diastolic BPs (mean difference 5.3 (95% CI 3.8-6.9) mm Hg and 1.9 (1.1-2.7) mm Hg, respectively) and a reduced risk of hypertension (0.62 (0.53-0.72)) relative to never drinkers. Similarly, excessive drinkers (>or=140 g ethanol per week) without alcohol sensitivity had a significantly higher systolic and diastolic BP and risk of hypertension than did nondrinkers (mean difference 5.1 (2.8-7.4) mm Hg, 2.7 (1.5-4.0) mm Hg and 34% (8-66%), respectively, for men). These differences were even greater for men with alcohol sensitivity (mean differences 12.0 (8.9-15.2) mm Hg, 6.2 (4.5-7.9) mm Hg and 95% CI (46-159%), respectively). Alcohol sensitivity and alcohol use were both associated with elevated BP and risk of hypertension in an older Chinese population. Alcohol sensitivity may aggravate the effect of drinking on BP. Limiting alcohol use to two drinks per day for men and one drink a day for women may be suitable for East Asians. Reduction of alcohol consumption should be an important public health target.
Hypertension Research 06/2009; 32(9):741-7. DOI:10.1038/hr.2009.92 · 2.66 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: It has been hypothesized that the emerging epidemic of diabetes in economically transitioning or recently transitioned populations is due to mismatch between developmental and mature environments. We took advantage of migration within an ethnically homogenous population to investigate this hypothesis, and the potentially modifying role of postnatal growth conditions, proxied by greater height. We used multivariable logistic regression in a population-based cross-sectional study from 1994 to 1996 of 2,341 long-term Hong Kong residents aged 25-74 years, either born in contemporaneously developed Hong Kong or migrants from economically undeveloped Guangdong. Migrant status was not associated with clinically diagnosed diabetes, odds ratio 1.05 (95% confidence interval 0.69-1.58) in adult migrants compared to Hong Kong-born natives and 1.22 (0.83-1.80) in preadult migrants, adjusted for age, sex, socio-economic position, and lifestyle. However, the association of diabetes with migrant status varied with height, suggesting a potentially complex relationship between indicators of prenatal and postnatal nutritional exposures. Compared to tall Hong Kong-born natives, the odds ratio of diabetes was 2.36 (1.20-4.61) in tall migrants, 1.94 (1.07-3.53) in short Hong Kong-born natives, but 1.04 (0.48-2.23) in short adult migrants. Additionally adjusting for body mass index and waist-hip ratio had little effect, apart from attenuating the association between short height and diabetes prevalence in Hong Kong-born natives. Whether the current epidemic of diabetes is a long-standing effect of such mismatch or a "first-generation through effect" generated by rapid economic development causing disproportionate growth remains to be determined.
American Journal of Human Biology 05/2009; 21(3):346-53. DOI:10.1002/ajhb.20874 · 1.70 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Objectives: To examine the relation of perceived peer smoking prevalence with smoking behavior and intention among primary students.
Methods: In 2008, primary 4 to 6 students in 14 randomly selected schools completed an anonymous questionnaire. Smoking status was categorized as current (at least <1 cig/week), former (smoked any numbers of cigarettes but not any more) and ever (current and former smoker) versus never. Intention to smoke when grown up (yes versus no) was recorded. Perceived peer smoking prevalence among all primary school students was classified as over-estimated (half) and grossly over-estimated (most/all) versus under/accurately estimated (none/some). Data were weighted by census school grade distribution. Logistic regression was used to calculate the odds ratios (ORs) for smoking behavior and intension adjusting for socio-demographic characteristics and taking into account clustering effect of schools.
Results: Among 3995 never and 257 ever-smoker, 73.3% under-estimated or accurately estimated, 16.4% over-estimated and 10.3% grossly over-estimated the prevalence of peer smoking. The adjusted current smoking ORs (95% CI) for over-estimation and gross overestimation were 1.30(0.42-4.08) and 5.56(1.88-16.4) (P for trend <0.001). The corresponding ORs for former and ever smokers were 1.87(1.22-2.84), 5.81(4.41-7.65) and 1.80(1.26-2.58), 5.75(4.19-7.88), respectively, (all p for trend <0.01). Over-estimation and gross over-estimation were positively associated with smoking intention (P<0.05).
Conclusions: Over-estimating peer smoking prevalence was positively associated with smoking and intention to smoke in primary students. Prospective studies are needed to examine the temporality of the association and trials are needed to investigate the effects of correcting perceived prevalence of smoking on smoking behaviours and intention.
Funding: The Hong Kong Council on Smoking and Health
12th World Congress on Public Health World Health Organization; 04/2009