Direct Effect Of Smoking Research Articles

DEVELOPING THE TOOLS FOR ENVIRONMENTAL EPIDEMIOLOGY: A REAL-TIME SURVEILLANCE AND INVESTIGATION SYSTEM FOR HAZARDOUS EXPOSURES

The birth weight-specific mortality curves for exposed and unexposed populations (e.g., smokers vs non-smokers) cross so that low birth weight (LBW) babies in exposed populations have a lower neonatal mortality than LBW babies in unexposed populations. There are competing statistical and biologic theories to explain this birth weight “paradox”. The authors used causal diagrams to show that the “paradox” can be conceptualized as selection bias due to stratification on a variable (birth weight) that is affected by the exposure of interest (e.g., smoking), and shares common causes with the outcome (infant mortality). 1991 U.S. Birth data was used to illustrate how crossing of birth-weight specific mortality curves is the expected result of the causal links between risk factors, birth weight and neonatal mortality. For example, the authors estimated the total effect of smoking on neonatal mortality without conditioning on any intermediate variable (relative risk: 1.5). After conditioning on LBW, as is usually done in an attempt to assess the direct effect of smoking not mediated through LBW, and found an inverse association (relative risk: 0.9) between smoking and mortality among LBW infants. This birth weight “paradox” has been extensively studied, and several statistical and biologic theories have been proposed to explain it. A method based on comparing the birth weight Z-scores, rather than the birth weights, between the exposed and the unexposed has been proposed as a strategy to eliminate the “paradox”. We have argued elsewhere that the “paradox” is an example of selection bias due to adjustment for birth weight, a variable affected by the exposure. We show here that the Z-score method can be seen as a sophisticated version of not adjusting for birth weight. This claim is illustrated by applying the Zscore method to real data and to data sets simulated under several states of nature represented by different causal diagrams (directed acyclic graphs). For example, under the null hypothesis (exposure-mortality crude odds ratio equal to 1), the Z-score adjusted odds ratio is also one even though the birth weight-adjusted odds ratio can be arbitrarily far from 1. The need for a clear specification of the causal question that motivates the data analysis is reiterated.

Reply: Smoking and hepatoblastoma: confounding by birth weight?

Sir, The question asked by Spector and Ross is interesting because birth weight is important in understanding parental smoking in relation to hepatoblastoma. Two aspects may be worth clarifying: (a) whether the association between parental smoking and hepatoblastoma is due to the confounding effect of low birth weight of the index child, and (b) whether the association between parental smoking and hepatoblastoma is due to an indirect effect of smoking through low birth weight rather than the direct effect of smoking on fetal liver. Our original analysis of parental prenatal smoking based on self-reported data collected from parents of children with hepatoblastoma and parents of population controls, showed the following: maternal smoking, OR=2.68 (P<0.05); both parents smoking, OR=4.74 (P<0.01); maternal smoking in cases diagnosed at the median age for hepatoblastoma or older, OR=12.02 (P<0.01) (Pang et al, 2003). We have now repeated the analyses adjusting for low birth weight as well as parental age and deprivation. Low birth weight was defined as less than 2000 g (Moore and Persaud, 1993) and treated as 0/1 binary variable. Results were obtained as follows: maternal smoking OR=2.50, both parents smoking OR=4.97, older children OR=12.66, in all cases P<0.05. We may conclude, therefore, that low birth weight does not explain the observed association between parental smoking and hepatoblastoma. We are not convinced that the association reflects an indirect effect through low birth weight. Hepatoblastoma is an embryonal tumour which most likely has a fetal origin and prenatal risk factors probably play a more important role than postnatal risk factors. The case for this may be made as follows: first, in some children, hepatoblastoma is evident at birth or a few months after birth (median age at diagnosis is about 1 year). Second, histopathological examination shows striking similarities in size and configuration between the developing fetal liver cells and the fetal epithelial-type cells of hepatoblastoma. Third, the tumour's ability to synthesise alpha-fetoprotein (AFP) points to a fetal origin. AFP, a major serum protein synthesised by fetal liver cells, is also found in children with hepatoblastoma. Fourth, hepatoblastoma, like Wilms’ tumour, is associated with Beckwith–Wiedemann syndrome and hemihypertrophy, suggesting gestational oncogenic events (DeBaun and Tucker, 1998). Furthermore, previous studies have shown that carcinogenic metabolites of tobacco smoke can cross the placenta, the natural barrier protecting the fetus from foreign substances, and the fetal liver, as a detoxification organ, is immature and vulnerable to carcinogens. A causal association with parental smoking is also supported by the finding of G to T transversions in the TP53 gene in hepatoblastoma (Olivier et al, 2002), a type of mutation that is typical of smoking-induced carcinoma of the lung (Pfeifer et al, 2002). In conclusion, our study suggests that parental smoking may play a role in the development of hepatoblastoma. However, as we stated, the finding may have arisen by chance since we carried out many statistical comparisons. Further studies of parental smoking, birth weight and other potential complications in relation to hepatoblastoma would be helpful. As only about 15 cases of hepatoblastoma occur each year in the UK and 100 cases a year in the US, an international effort would be required.

Influence of lifestyle on vitamin bioavailability.

In this review the effects of lifestyle factors, especially alcohol consumption, on vitamin bioavailability are summarized and discussed. Alcohol effects are clearly dose-dependent. Excessive chronic alcohol intake is generally associated with vitamin deficiency (especially folate, thiamine, and vitamin B6) due to malnutrition, malabsorption, and ethanol toxicity. Effects of moderate alcohol use are mainly explained by a lower vitamin intake. In the case of vitamin A and beta-carotene, effects on post-absorptive (lipoprotein) metabolism have been demonstrated. In one diet-controlled crossover study, alcohol consumption resulted in an increase in the plasma vitamin B6 (PLP) content, especially after beer consumption (containing vitamin B6), but also after wine and spirit consumption (not containing vitamin B6). Smoking is also associated with a lower dietary vitamin intake. In the case of vitamin C, B12, folate, and beta-carotene, evidence has been presented for effects on postabsorptive metabolism, due to smoke-induced oxidative stress and/or vitamin inactivation. For vitamin E a direct effect of smoking on absorption has been demonstrated. There is no convincing evidence that low-fat diets negatively affect fat-soluble vitamin absorption, but cholesterol-lowering compounds (diets), or unabsorbable fat substitutes, may do so. Vitamin bioavailability may be compromised from certain vegetables (particularly raw), and/or from high-fiber foods, because of limited digestion and inefficient release of vitamins from the food matrix.

Does a reduced sensitivity to bitter taste increase the risk of becoming nicotine addicted?

Cigarette smoking appears to be on the increase in adolescents. The initiation of regular smoking nearly always begins before adulthood. It is therefore crucial to find ways of identifying those children most vulnerable to nicotine addiction and prioritizing them for preventive measures. We hypothesized that individuals who, in a simple taste test, perceive phenylthiocarbamide (PTC) as bitter may find the taste of cigarettes aversively bitter and could therefore have a reduced vulnerability to nicotine addiction compared to nontasters, who would be the group at greater risk of addiction. We studied 242 Plains American Indians, 136 women and 106 men aged 18-59 years, and found that (allowing for gender differences and the possible direct effects of smoking on taste) the proportion of PTC nontasters to tasters in smokers, even light smokers, was significantly greater than in both nonsmokers and social smokers (chi2= 15.875, 4 df; P=.003), suggesting that nontasters, who are not aversive to the bitter taste of cigarettes, may be more at risk for heavy smoking and therefore more vulnerable to nicotine addiction.

The Relationship Between Continued Smoking During Pregnancy and Gestational Weight Gain in a Culturally Diverse Population

LEARNING OUTCOME: Participants will learn about the relationship between smoking and adequacy of weight gain during pregnancy. A direct effect of smoking on birth weight has been established, however, less research has examined the relationship between continued smoking during pregnancy and gestational weight gain; the latter having a known independent effect on size at birth. The purpose of this analysis is to assess the impact of continued smoking during pregnancy on gestational weight gain adequacy using the Massachusetts Pregnancy Nutrition Surveillance System (PNSS), and to examine effects within three race/ethnic groups (total n=21,758): Non-Hispanic White (NHW), Non-Hispanic Black (NHB) and Hispanic. The WIC Program, which provides the data for this analysis and is the principal source of data for PNSS, serves over 32,000 pregnant/postpartum women per year from diverse racial and ethnic backgrounds. In addition to being a health and nutrition program, WIC promotes smoking cessation or reduction during pregnancy. As part of this effort, data on smoking status is collected at 2 points in time during pregnancy. Results, using a Chi-square, indicate that a significantly higher percentage of smokers (31.6%) exhibit inadequate gestational weight gain (adjusted for pre-pregnancy weight status) than non-smokers (24.8%, p < .0001). This higher prevalence (p < .001) of inadequate weight gain among smokers in comparison to non-smokers was consistently noted for NHW (30.3% vs. 21.1%), NHB (37.3% vs. 26.9%) and Hispanic women (36.1% vs. 29.3%). Moreover, as compared to smokers, a significantly lower percentage of those who quit smoking during pregnancy exhibited inadequate weight gain (18.2% vs. 31.6%, p < .0001). This positive association between smoking cessation and weight gain was also observed within NHW, NHB, and Hispanic groups (p < .0001). These results suggest that continued smoking during pregnancy is associated with inadequate weight gain during pregnancy. This may be one of the mechanisms for the adverse effects of smoking on birth weight.

Down's syndrome and maternal smoking in early pregnancy

Several previous studies have suggested that maternal smoking is associated with a decreased incidence of trisomy 21. By using the Swedish health registries, 1,321 infants with Down's syndrome (DS) were selected among 1,117,021 infants born in 1983-1993 with known smoking exposure in early pregnancy. No association between maternal smoking and all cases of DS was found [age-adjusted odds ratio (OR) for maternal smoking: 0.98; 95% confidence interval (CI): 0.86-1.11], but heterogeneity over strata existed. A slightly decreased OR (0.91; 95% CI: 0.72-1.15) for any maternal smoking was indicated among primiparas, but among multiparas, no effect of smoking on the incidence of DS could be detected (OR: 1.01; 95% CI: 0.87-1.17). The difference between the OR for smoking > or = 10 cigarettes per day among primiparas (OR: 0.59; 95% CI: 0.38-0.90) and multiparas (OR: 1.06; 95% CI: 0.86-1.31) was statistically significant. If not due to statistical fluctuation, the findings indicate that no direct effect of smoking on DS risk exists but the association observed in primiparas is due to covarying factors.

Distinguishing withdrawal relief and direct effects of smoking

Distinguishing withdrawal relief and direct effects of smoking

Direct effects of smoking on the heart: Silent ischemic disturbances of coronary flow

Cigarette smoking is strongly associated with ischemic heart disease and acute coronary events. The effect of smoking a single cigarette on regional myocardial perfusion was studied in 13 chronic smokers with typical stable angina pectoris using positron emission tomography and rubidium-82 ( 82Rb). Findings were compared with the effects of physical exercise. After exercise, 8 patients (61%) had angina, ST depression and abnormal regional myocardial perfusion. Uptake of 82Rb increased from 49 ± 8 to 60 ± 7 in remote myocardium, but decreased from 46 ± 3 to 37 ± 5 in an ischemic area. The remaining 5 patients (39%) had homogeneous increases in 82Rb uptake without angina or ST depression. After smoking, 6 of the 8 patients with positive exercise test responses had a decrease in 82Rb uptake, from 47 ± 3 to 35 ± 6 in the same segment of myocardium affected during exercise. However, in contrast to exercise, the events during smoking were largely silent. The absolute decreases in regional 82Rb uptake after smoking occurred at significantly lower levels of myocardial oxygen demand than after exercise. This suggests that an impairment of coronary blood supply is responsible. Thus, in smokers with coronary artery disease, each cigarette can cause profound silent disturbances of regional myocardial perfusion that are likely to occur frequently during daily life. Such repeated insults may represent an important mechanism linking smoking with coronary events.

Smoking habits and the leukocyte count.

In 86,488 multiphasic examinations, mean leukocyte counts were highest in cigarette smokers, intermediate in ex-cigarette and cigar or pipe smokers, and lowest in nonsmokers. Among the races, whites had the highest, yellows next, and blacks the lowest leukocyte counts. The leukocyte count was related to quantity smoked, inhalation, and smoking duration. Most groups who changed smoking habits showed corresponding changes in leukocyte counts. Higher leukocyte counts in smokers appeared largely to be a direct effect of smoking, although a small part of the increase seemed attributable to chronic bronchitis. A contribution of genetic or constitutional differences between smokers and nonsmokers was not ruled out. “Normal” leukocyte count values should take into account age, sex, race, and smoking status.