European Journal of Clinical Nutrition (2004) 58, 204–211
& 2004 Nature Publishing Group All rights reserved 0954-3007/04 $25.00 www.nature.com/ejcn
ORIGINAL COMMUNICATION Cigarette smoking during pregnancy in rural Nepal. Risk factors and effects of b-carotene and vitamin A supplementation P Christian1*, KP West Jr1, J Katz1, E Kimbrough-Pradhan1, SC LeClerq1, SK Khatry2 and SR Shrestha2 1
Division of Human Nutrition, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD, USA; and 2The Society for Prevention of Blindness, Kathmandu, Nepal
Objective: We examined risk factors of smoking and the association between smoking and pregnancy-related and 6-month infant mortality in rural Nepal, where 30% women reported smoking during pregnancy. Design: Cross-sectional analysis of risk factors associated with smoking status and health consequences of smoking, using prospective data collected as part of a randomized community trial to examine the effect of maternal vitamin A or b-carotene supplementation on maternal mortality. Setting: Rural, southeastern plains of Nepal. Subjects and methods: A total of 17 767 women contributed at least one pregnancy during 3.5 y of the study. Data on cigarette or bidi (rolled tobacco) smoking were collected using a 7-day recall, twice during pregnancy. Associations between smoking status and maternal diet, morbidity profile, household socioeconomic status and serum concentration of retinol, carotenoids and tocopherols were examined. Further, relative risk (RR) and 95% confidence intervals (CI) were calculated to estimate supplement effects on pregnancy-related mortality, stratified by smoking status during pregnancy. Results: Smokers were more likely to be older, illiterate and poor compared to nonsmokers. Fruit and vegetable consumption among smokers and nonsmokers did not vary. However, smokers were more likely to consume meat/fish/eggs and less likely to consume milk than nonsmokers. They were also more likely to report symptoms of vaginal bleeding, edema, severe headache and convulsions during pregnancy relative to nonsmokers. Mortality per 100 000 pregnancies appeared to be higher among smokers than nonsmokers in the placebo group (915 vs 584, RR ¼ 1.57, 95% CI: 0.80–3.08). b-Carotene supplementation reduced pregnancy-related mortality both among smokers (RR ¼ 0.31 95% CI: 0.11–0.89) and nonsmokers (RR ¼ 0.41, 95% CI: 0.19–0.89). Similar results obtained with vitamin A supplementation were not statistically significant. Infant mortality up to 6 months was B30% higher among smokers compared to nonsmokers in the placebo group both before and after adjusting for confounding factors. Neither supplement given to women reduced infant mortality. Conclusions: Cigarette smoking during pregnancy is associated with an increased risk of maternal and infant mortality in rural Nepal. b-Carotene and to some extent vitamin A may reduce the risk of pregnancy-related mortality, but not infant mortality, among both smokers and nonsmokers. European Journal of Clinical Nutrition (2004) 58, 204–211. doi:10.1038/sj.ejcn.1601767 Keywords: smoking; pregnancy; infants; mortality; b-carotene; vitamin A; Nepal
*Correspondence: P Christian, Division of Human Nutrition, Room 2041, Bloomberg School of Public Health, Johns Hopkins University, 615 N. Wolfe St, Baltimore, MD 21205, USA. E-mail:
[email protected] Guarantors: KP West. Contributors: PC analyzed the data and wrote the paper, KPW was the principal investigator of the study, JK helped with study design, implementation and paper editing, EKP managed data entry and
helped with study design and forms development, SLC helped in study design and implementation and editing of the paper, SKK was the country director and had overall responsibility for the management and implementation of the project. SRS was responsible for community advocacy, training, translation of forms and maintaining quality control. Received 2 November 2002; revised 4 March 2003; accepted 5 March 2003
Smoking and pregnancy-related risks in Nepal P Christian et al
205
Introduction Trends in cigarette smoking prevalence in the developing world are on the incline as reflected by per capita consumption increases of 22% in Asia, 24% in Latin America and 42% in Africa (Masironi & Rothwell, 1988). Tobacco use during pregnancy has been well established as a powerful risk factor for adverse pregnancy outcomes, which include spontaneous abortion (Ness et al, 1999), intrauterine growth retardation (Roquer et al, 1995; Ahluwalia et al, 1997; Rasmussen & Adams, 1997; Sadler et al, 1999), preterm delivery (Pollack et al, 2000; Windham et al, 2000), oral clefts (Lieff et al, 1999; Lorente et al, 2000) and fetal death (Kleinman et al, 1988; Wu et al, 1998a; Tuthill et al, 1999). Imbalances in nutritional status, both of macro- and micronutrients, have been reported because of cigarette smoking, which may have a role in the pathogenesis of diseases and adverse outcomes related to smoking (Preston, 1991). For example, circulating and tissue levels of antioxidants such as vitamins E and C and b-carotene are lower among smokers (Dallongeville et al, 1998). It has also been observed that smokers are less likely to consume fruits and vegetables (Preston, 1991) and that smoking in pregnant women is associated with significantly lower intakes of vitamin C and carotenoids (Mathews et al, 2000). Regular multivitamin/mineral use has shown to reduce the risk of fetal death associated with smoking in US women (Wu et al, 1998a, b). The relation between smoking, nutrition and pregnancyrelated outcomes has not been well examined in third-world settings, where rates of cigarette smoking are increasing and nutritional deficiencies, particularly of micronutrients, are endemic. We had the opportunity to examine these relations in a rural region of Nepal, where, in a randomized, placebocontrolled community trial, low dose, weekly supplementation with vitamin A or b-carotene to women of reproductive age reduced pregnancy-related mortality by 44% (West et al, 1999). This rural female population residing in the southeastern plains of Nepal has high rates of malnutrition, with 11% of pregnant women developing night blindness, a symptom of vitamin A deficiency (Christian et al, 1998a), and approximately 75% being anemic and infested with hookworms (Dreyfuss et al, 2000). Maternal mortality was high at 704 per 100 000 in the placebo group (West et al, 1999) and infant mortality up to 6 months of age was B70/ 1000 in the study area (Katz et al, 2000). Furthermore, 30% of women reported smoking tobacco during pregnancy. In this paper, we examine dietary, demographic and other environmental factors associated with cigarette smoking in this population. We also examined the risk of maternal and infant mortality associated with exposure to tobacco smoking during pregnancy and the effects of vitamin A or bcarotene supplementation.
Subjects and methods This paper uses data collected as part of a placebo-controlled, randomized, community-based trial carried out from July
1994 to September of 1997 in 270 wards of 30 village development communities (nine wards in each) of the Sarlahi district located in the southeastern plains of Nepal. The details of the study have been described previously (West et al, 1999; Katz et al, 2000). Briefly, over a period of 3.5 y, B45 000 women of reproductive age (13–45 y) were randomized by ward to receive weekly, one of the following three supplements: placebo, 7000 mg retinol equivalents (RE) of vitamin A as retinyl palmitate and 42 mg of all trans b-carotene (B7000 mg RE, assuming a conversion ratio of 6:1). Local female workers were hired to distribute the supplements to eligible women in their area. The female workers also ascertained pregnancies by questioning the study participants regarding their period and pregnancy status on a weekly basis. Vital status of live born infants was also assessed up to 6 months of life on a weekly basis. Once women reported being pregnant, they were enrolled into a protocol of home-based data collection on a 7-day recall of diet, morbidity, work activities and substance use. These data were collected at the time of pregnancy enrolment (B4 months of gestation) and in the third trimester (B7 months of gestation). Women were asked to recall the number of times they had consumed 38 different types of vitamin A- and carotenoid-rich foods from a list that was developed using formative research on commonly consumed sources of vitamin A and b-carotene in this area of Nepal. Previously, we have shown that vitamin A and b-carotene food intakes, using the 7-day food frequency recall, were significantly associated with functional and biochemical indicators of vitamin A status such as maternal night blindness (Christian et al, 1998b) and low serum retinol (West et al, 1999) and carotenoid concentrations (Yamini et al, 2001). Mid-upper-arm circumference (MUAC), as a measure of nutritional status, was measured at both times. Data on household socioeconomic status and previous pregnancy and reproductive history were also collected. In three (10%) of the 30 village-development communities, which included 27 wards (nine per treatment group), pregnant women underwent a clinic-based assessment that involved a phlebotomy at the time of enrolment. A measure of 5 ml blood was collected and centrifuged to extract serum for retinol, tocopherol and carotenoid analyses that were performed using high-pressure liquid chromatography (Craft, 1996) at the Center for Human Nutrition at the Johns Hopkins University in Baltimore, MD. Hemoglobin (Hb) concentration in the blood was estimated on the spot using a hemoglobinometer (HemoCue Inc., Mission Viejo, CA, USA). Maternal and infant vital status was monitored on a weekly basis throughout the duration of the study. In the event of a death, a home-based verbal autopsy questionnaire was administered to the family members of the deceased. At the end of the study, two physicians reviewed the questionnaires and assigned a proximate cause of death using a consensus process. European Journal of Clinical Nutrition
Smoking and pregnancy-related risks in Nepal P Christian et al
206 Exposure Data on cigarette or bidi (rolled tobacco) smoking were collected using a 7-day recall, twice during pregnancy. Data on number of cigarettes smoked per day were not obtained. However, based on data collected in a subsequent study in the same study area, the average number of cigarettes reported to have been consumed by women per day was 7.5 (Christian et al, unpublished data). In the present study, exposure to cigarette smoking was defined as the consumption of any cigarettes on one or more days per week during histories taken either in the second or third trimesters of pregnancy. However, 97% of smokers reported to have smoked daily in the previous week, leading us to estimate an average of approximately 50 cigarettes or bidis smoked per week in the study population. In about 11% of pregnancies, smoking status was not available and this did not vary by treatment group.
Statistical analysis Demographic, socioeconomic, morbidity and nutritional characteristics of smokers and nonsmokers were compared using the w2-test. In the 10% subsample of women for whom blood was collected, mean serum retinol, carotenoid and tocopherol concentrations were compared between smokers and nonsmokers using the t-test. Relative risks (RR) and 95% confidence intervals (CI) were calculated for pregnancyrelated mortality and infant mortality through 6 months of age by smoking status within each of the treatment arms. Using logistic regression analysis, we examined the association of smoking to maternal and infant mortality, adjusted for age, parity, literacy, MUAC, and socioeconomic status (using land and cattle ownership) and alcohol consumption. Further, we also calculated the RR and 95% CI of pregnancyrelated mortality comparing subjects in either of the two treatment groups to those in the placebo group, stratified by smoking exposure.
Ethical approval The study was reviewed for ethical approval by the Nepal Health Research Council in Kathmandu, Nepal, the Joint Committee on Clinical Investigations at the Johns Hopkins University, Baltimore, MD, and the Teratology Society. Oral
Table 1
informed consent was obtained from the subjects and they were free to refuse supplements or withdraw from the study at any time during its duration.
Results Out of 44 570 women of reproductive age who were enrolled, 17 767 contributed one or more pregnancy to the study yielding a total of 18 794 pregnancies, or approximately 6000 pregnancies in each of the three treatment arms (Table 1). Among these, 0.6–0.8% were lost-to-follow-up for vital status assessment within the first 12 weeks postpartum and about 6–7% of infants were lost-to-follow-up within the first 6 months of life. The prevalence of tobacco smoking was 30.3% in the placebo group, 32.0% in the b-carotene group and 29.0% in the vitamin A group (data not shown), yielding an overall smoking prevalence of 30.3%. Mean alcohol consumption prevalence was lower at 11.2%, but was strongly associated with smoking. A significantly higher proportion of smokers (27.1%) reported having consumed alcohol in the past week compared to those who did not smoke (4.2%, Po0.001). Pregnant women who smoked were more likely to be older, have more children, be illiterate, poorer, and belong to the lower Hindu-castes compared to nonsmoking women (Table 2). Milk or yogurt consumption was considerably lower among women who smoked during pregnancy than among those who did not. In contrast, animal meat and eggs were more likely to be consumed by smoking women than nonsmoking women. Weekly intakes of other foods were comparable. Smokers were more likely to report symptoms of severe illness such as vaginal bleeding, edema and severe headache in the last 3 months of pregnancy relative to nonsmokers in all three treatment groups (Table 3). The proportion of women with low MUAC (o18.5 cm) was higher during pregnancy among smokers in all three groups. Out of all the nutrients measured in the serum, g-tocopherol, and to some extent a-tocopherol concentrations, were significantly lower during pregnancy among smoking women (Table 4). The pregnancy-related mortality rate among smokers was higher (920 per 100 000 pregnancies) than among nonsmokers (590 per 100 000 pregnancies) in the placebo group
Number of women, pregnancies, live births and deaths by treatment allocation: July 1994–September 1997
Women of reproductive age Women ever pregnant Pregnancies Live births Maternal deaths up to 12-week postpartuma Infant deaths up to 6-month postpartumb a
Placebo
b-carotene
Vitamin A
Total
14 783 5778 6084 5147 42 321
14 519 5769 6155 5148 22 322
15 268 6220 6555 5551 30 362
44 570 17 767 18 794 15 846 94 1005
0.8% lost-to-follow-up because of out-migration from the study area. 7.3% lost-to-follow-up because of out-migration from the study area.
b
European Journal of Clinical Nutrition
Smoking and pregnancy-related risks in Nepal P Christian et al
207 (RR ¼ 1.57, 95% CI: 0.80–3.08) (Table 5). This RR was maintained even after adjusting for confounding variables, although the 95% CI included one (data not shown). Among smokers, the pregnancy-related mortality rate in b-carotene recipients was 31% (RR ¼ 0.31, 95% CI: 0.11–0.89) that among women receiving a placebo. Among nonsmokers, this figure was 41% (RR ¼ 0.41, 95% CI: 0.19–0.89). The mortality rate of vitamin A-recipients relative to those in the placebo group was 64% (RR ¼ 0.64, 95% CI: 0.28–1.45) and
Table 2 Demographic and household characteristics and dietary intakes of smoking and nonsmoking pregnant women (all treatment groups combined)
Maximum observations Age >25 y Parity >3 Literate Owned land Owned cattle Hindu-low-caste Dietary intake (any in past 7 days)a Number Meat/fish/eggs Milk/yogurt Tomato Orange Guava Papaya Mango Jackfruit Cabbage Dark-green vegetables
Smoker
Nonsmoker
5012 54.0 40.4 3.6 32.8 57.2 92.8
11 507 26.8 17.5 19.6 53.7 73.0 82.4
5074 72.0 57.1 32.4 5.6 12.2 19.5 9.7 6.0 12.3 74.3
11623 58.8 75.8 27.8 9.2 14.5 19.4 10.7 7.5 15.3 72.6
All Po0.001, comparing smokers vs nonsmokers using the w2-test, except mango, papaya, jackfruit and dark-green vegetables. a Inquired during 1 or 2, 7-day periods during pregnancy. Values are in percentages, unless indicated otherwise
Table 3
74% (RR ¼ 0.74, 95% CI: 0.39–1.39) in smokers and nonsmokers, respectively. Infant mortality up to 6 months was B30% higher among mothers who were smokers compared to nonsmokers in the placebo group, and this elevated risk remained statistically significant after adjusting for maternal age, parity, literacy, nutritional status using MUAC, alcohol consumption and socioeconomic status (data not shown). Maternal receipt of vitamin A or b-carotene did not reduce the excess risk of infant mortality associated with smoking, as reflected by relative risk estimates that were close to one. Cause-specific mortality among women (Table 6) by smoking status showed that eclampsia-related deaths may have been higher among nonsmokers compared to smokers. However, both obstetric causes and infection-related causes appeared to be higher among smokers compared to nonsmokers. Cause of infant mortality (Table 7) revealed slightly higher risks of prematurity, and infections such as diarrhea and acute lower respiratory infection as well as malnutrition among smokers compared to nonsmokers.
Discussion Cigarette smoking during pregnancy is a well-established risk factor for numerous adverse pregnancy outcomes among women of the developed world. This study also showed that cigarette smoking may be associated with an increased risk of pregnancy-related and infant mortality in a developing country (UNICEF, 2000). b-carotene supplementation resulted in a reduction in pregnancy-related mortality of approximately the same magnitude among smokers and nonsmokers. Although vitamin A supplementation produced similar results, the reduction in mortality was smaller and statistically nonsignificant. Unlike maternal mortality, supplementation with vitamin A or b-carotene had no impact on infant mortality irrespective of maternal smoking
Health and nutritional status of smoking and nonsmoking pregnant women by treatment group Placebo Smoker
Nonsmoker
Smoker
Smoker
Nonsmoker
3275 2.1 14.0*** 2.0 16.5*** 9.1
1520 3.4 18.5 1.8 19.5 9.6
3652 1.6*** 14.4*** 1.6 16.7* 8.1
1702 6.4
3599 3.6***
1665 7.6
4022 3.7***
125 47.2
356 34.6*
Illness in 3 months before birth (%) Maximum number 1434 Vaginal bleeding 4.0 Swelling of hands/face 19.2 Convulsions 2.6 Severe headache 22.1 Other severe illness 11.0
3327 2.1*** 14.2*** 1.7* 17.0*** 8.5***
1576 3.1 18.2 2.3 21.6 12.6
Wasting malnutrition Number MUAC p 18.5 cm
3613 2.8***
Anemiaa Number Hb o100 g/l
1601 6.0
73 53.4
280 43.9
Vitamin A
b-carotene Nonsmoker
94 48.9
311 36.3*
a Hb assessed among subsample women. *Po0.05, ***Po0.001.
European Journal of Clinical Nutrition
Smoking and pregnancy-related risks in Nepal P Christian et al
208 Table 4
Serum retinol, carotenoid and tocopherol concentrations among smokers and nonsmokers by treatment group Placebo
Biochemical status (mmol/l) Maximum observations Retinol b-Carotene a-Tocopherol g-Tocopherol Lutein+zeazanthin b-Cryptozanthin Lycopene
Vitamin A
b-carotene
Smoker
Nonsmoker
Smoker
Nonsmoker
Smoker
Nonsmoker
Mean (s.d.) 73 1.00 (0.35) 0.14 (0.13) 15.2 (5.3) 1.13 (0.69) 0.43 (0.32) 0.05 (0.05) 0.01 (0.02)
Mean (s.d.) 278 1.00 (0.37) 0.15 (0.15) 15.4 (4.9) 1.35 (0.82)* 0.47 (0.32) 0.06 (0.06) 0.02 (0.03)
Mean (s.d.) 125 1.12 (0.41) 0.18 (0.16) 15.4 (4.8) 1.14 (0.67) 0.52 (0.35) 0.08 (0.09) 0.02 (0.04)
Mean (s.d.) 356 1.11 (0.40) 0.20 (0.18) 16.2 (5.2) 1.54 (0.93)** 0.52 (0.32) 0.09 (0.08) 0.02 (0.04)
Mean (s.d.) 94 1.24 (0.39) 0.14 (0.14) 14.3 (4.6) 1.22 (0.76) 0.44 (0.26) 0.07 (0.05) 0.01 (0.02)
Mean (s.d.) 307 1.32 (0.35) 0.16 (0.15) 15.4 (5.0) 1.34 (0.88) 0.51 (0.30) 0.07 (0.07) 0.02 (0.04)
*Po0.05, **Po0.001.
Table 5 Effect of b-carotene or vitamin A supplementation on pregnancy-related and infant mortality by cigarette smoking among women during pregnancy Placebo Cigarettes
Yes
Pregnancy-related mortality No. of pregnancies 1639 No. of deaths 15 Death rate (per 100 000) 915 RR (95% CI) 1.57 (0.80, 3.08) RR (95% CI) 1.00 Infant mortality (up to 6 months) No. of livebirths 1352 No. of deaths 116 Death rate (per 1000) 85.8 RR (95% CI) 1.29 (1.01, 1.62) RR (95% CI) 1.00
Vitamin A
b-carotene No
Yes
No
Yes
No
3765 22 584 1.00 1.00
1754 5 285 1.18 (0.40, 3.49) 0.31 (0.11–0.89)
3727 9 241 1.00 0.41 (0.19–0.89)
1706 10 586 1.36 (0.62, 2.99) 0.64 (0.28–1.45)
4174 18 431 1.00 0.74 (0.39–1.39)
3075 205 66.7 1.00 1.00
1467 117 79.8 1.16 (0.92, 1.46) 0.93 (0.71, 1.21)
2985 205 68.7 1.00 1.03 (0.84, 1.26)
1432 121 84.5 1.19 (0.95, 1.49) 0.98 (0.76, 1.28)
3407 241 70.7 1.00 1.06 (0.88, 1.28)
RR=relative risk, CI=confidence interval.
Table 6 Cause-specific pregnancy-related mortality rate by smoking status
Eclampsia Antepartum/postpartum hemorrhage, anemia Obstructed labor/retained placenta/obstetric shock Puerperal sepsis Other infectionsa Injuryb Chronic diseasec No. Information/uncertain
Smoker
n=5099
Nonsmoker
n=11,666
1 4
19.6 78.4
8 7
68.6 60.0
6
117.7
9
77.2
2 9 3 2 3
39.2 176.5 58.8 39.2 58.8
5 12 3 0 5
42.9 102.9 25.7 — 42.9
a Include gastroenteritis (n=6), hepatitis (n=3), sepsis (n=4), tuberculosis (n=4), typhoid (n=1), leishmaniasis (n=1) and other respiratory infection (n=2). b Includes hanging (n=2), snake bite (n=2), burning (n=1) and drowning (n=1). c Includes asthma. Values are number of deaths and mortality rates per 100 000 pregnancies.
status, similar to the lack of an overall effect of supplementation that has been described previously (Katz et al, 2000). While cigarette smoking can result in increased oxidative stress (Traber et al, 2000; Zhou et al, 2000), smoking during European Journal of Clinical Nutrition
Table 7
Cause-specific infant mortality rate by smoking status Smoker n=4251
Prematurity Congenital malformation Birth asphyxia Tetanus Sudden deatha Injury Diarrhea/dysentery Acute lower respiratory infection Sepsis Severe malnutrition Uncertainb Missingc
Nonsmoker n=9467
27 3 5 3 40 0 37 66
7.9 0.7 1.2 0.7 9.4 — 8.7 15.5
53 3 5 2 90 2 31 109
5.6 0.3 0.5 0.2 9.5 0.2 3.2 11.5
21 18 112 21
4.9 4.2 26.3 4.9
38 23 254 41
4.0 2.4 26.8 4.3
a Refers to sudden death in infant when no morbidity was reported and the infant died suddenly without an apparent reason. b Physician review unable to assign a cause of death c No information/data available. Values are number of deaths and mortality rates per 1000 live births.
pregnancy, paradoxically, has been shown to be associated with a lower risk of pre-eclampsia (Conde-Agudelo et al, 1999; Salafia & Shiverick 1999; Conde-Agudelo & Belizan, 2000). Thus, in the present study the effects of b-carotene
Smoking and pregnancy-related risks in Nepal P Christian et al
209 supplementation among smokers in reducing pregnancyrelated mortality can not be explained through an effect on deaths because of pre-eclampsia. Based on verbal autopsy data available on causes of death, smokers in our population appeared to be at a lower risk of death because of eclapmpsia (19.6 per 100 000 pregnancies) compared with nonsmokers (68.6 per 100 000). Antepartum and postpartum hemorrhage, obstructed labor, and retained placenta as causes of death were, on the other hand, approximately 30–50% more frequently observed among smokers than nonsmokers. Similarly, deaths caused by infections were 70% more common among women who smoked than those who did not. On the other hand, the increased risk of infant mortality among smokers appeared to be more nonspecific and not related to any specific cause per se although prematurity, infectious causes and severe malnutrition related causes in this group tended to be higher. We know that cigarette smoking is a strong risk factor for intrauterine growth retardation and prematurity. It is likely that the increased mortality observed among infants of smokers may in part be mediated through these factors. Sudden infant death syndrome (SIDS) has been associated with maternal smoking (Wisborg et al, 2000; Pollock, 2001). In the present study, SIDS was not diagnosed clinically. Using verbal autopsy data, physicians assigned cause of death as ‘sudden’ when no illness was reported prior to death that occurred suddenly. Neither were data on sleeping position collected in the study. Thus, it is likely that this method of cause of death assignment may have lead to substantial misclassification of true SIDS. Further, using this method many deaths were also assigned an ‘uncertain’ cause, which indicated that sufficient information was not available for the physicians to attribute a specific cause to the death. It may be argued that cigarette smoking in this population was simply a marker of other risk factors, since women who smoked tended to be older, thinner, belong to poorer, lowcaste families, and have more inadequate diets and nutritional deficiencies than women who did not smoke. The higher pregnancy-related mortality experienced by smokers could then be attributed to these associated risk factors rather than to the risks posed by cigarette smoking itself. This argument may be flawed for two reasons, the first being that b-carotene supplementation appeared to modulate the excess risk of dying among smokers (RR ¼ 1.57 in the placebo group compared to 1.18 in the b-carotene group). It is unlikely that a nutritional intervention such as b-carotene supplementation could ameliorate the causal effects of the various socioeconomic risk factors associated with smoking cigarettes. Second, in our multivariate analysis, controlling for risk factors significantly associated with pregnancyrelated mortality including MUAC, maternal age, parity, alcohol consumption and socioeconomic status had no effect on the protective effect of b-carotene on mortality of smoking mothers.
Smoking can lower the availability and increase the requirement of various nutrients such as vitamins B12, and C, folate, iron and zinc (IOM, 1990). Increased utilization of antioxidants such as vitamins E and C may contribute to low circulating concentrations of these antioxidants in smokers (Preston, 1991; Cross et al, 1999; Traber et al, 2000). On the other hand, smoking is associated with consumption of a less healthy diet in pregnant women (Haste et al, 1990, 1991; Trygg et al, 1995; Dallongeville et al, 1998) in settings where food availability and access are adequate. One study found that while smokers had lower intakes of most micronutrients, only intakes of antioxidants, especially vitamin C and carotenoids, were found to be significantly depressed after adjustment (Mathews et al, 2000). It was surprising that none of the circulating concentration of antioxidants (except gtocopherol) differed by smoking status. It is possible, however, in this rural population in Nepal, that consumption of fruits and vegetable sources of antioxidant vitamins is universally low, as revealed by extremely low concentrations of many carotenoids, and less likely to vary by smoking status. The poorer nutritional status exhibited by smokers in the present study may be related to, apart from consuming a poor diet, higher infection rates and work load and other factors, all of which may be related to the different lifestyle and socioeconomic background of women who were smokers compared to nonsmokers. Paradoxically, reported consumption of animal protein was higher among smokers relative to nonsmokers. This may be because smokers tend to come from low-caste families among whom meat consumption is more prevalent than among the high-caste Hindus, who are more likely to be vegetarians because of religious reasons. Also, consumption of meat/fish, alcohol and smoking can all occur together within the context of a social gathering among relatives and friends. However, the low frequency at which meat and fish consumption (B1/ week) is reported is unlikely to have a substantial impact on the nutritional value of the diet. Milk, on the other hand, is a high prestige expensive food that was less likely to be consumed by smokers. There are two main limitations of this analysis. One is the lack of adequate quantification of smoking exposure. While we were able to classify women as smokers or nonsmokers based on self-reported histories of smoking during pregnancy, individual-level data on numbers and type of cigarettes consumed were not collected. Further, women’s lifetime risk of smoking was not obtained. The second limitation was the lack of power to demonstrate a statistically significantly higher risk of pregnancy-related mortality, an extremely rare event. We did, however, have adequate power to show an elevated risk of infant mortality because of maternal smoking, which was sustained even after adjusting for numerous confounding factors. While discontinuing smoking during pregnancy is to be recommended strongly to prevent adverse birth outcomes, there is some evidence to suggest that this may not completely prevent deficits in birth length and ponderal European Journal of Clinical Nutrition
Smoking and pregnancy-related risks in Nepal P Christian et al
210 index (Lindley et al, 2000). Data from the 1998 National Maternal and Infant Health Survey showed that while multivitamin/mineral supplement use did not modify the effect of maternal smoking on birth outcomes (Wu et al, 1998b), it was significantly associated with reduced fetal mortality (Wu et al, 1998a). Our study brings forth evidence for a likely beneficial effect of b-carotene supplementation in modifying the risk of pregnancy-related mortality both associated and disassociated with cigarette smoking in undernourished mothers. It is important to note that we used b-carotene in amounts recommended to meet the daily requirements of vitamin A for pregnant or lactating women (42 mg/week), and not in large amounts (20–30 mg/day) used in randomized trials which increased the incidence of and death from lung cancer and cardiovascular disease in older, predominantly male, heavy smokers and workers exposed to asbestos in developed countries (The Alpha-tocopherol Beta-carotene Cancer Prevention Study Group, 1994; Omenn et al, 1996). The major results of the study suggest that (1) cigarette smoking was associated with an increased risk of pregnancyrelated and infant mortality, (2) supplementation with b-carotene, and to some extent vitamin A, decreased pregnancy-related mortality among both smokers and nonsmokers and (3) neither supplement had any effect on the infant mortality among smokers or nonsmokers. The mechanisms of effects, especially that of b-carotene need elucidation, although its antioxidant properties may have some role.
Acknowledgements This study was carried out under cooperative agreement HRN-A-00-97-00015-00 between Office of Nutrition, US Agency for International Development (USAID), Washington, DC, and the Center for Human Nutrition (CHN), Department of International Health, Johns Hopkins University, Bloomberg School of Public Health, Baltimore, MD. It was a joint undertaking between the CHN and the National Society for the Prevention of Blindness, Kathmandu, Nepal. The study was funded by USAID, with financial and technical assistance from Task Force Sight and Life Research Institute (Roche, Basel). Additional support from the Bill and Melinda Gates Foundation is gratefully acknowledged. We acknowledge Dev N Mandal, Tirtha R Shakya, Ramesh K Adhikari, Sanu M Dali, Sadigheh YaminiRoodsari, Lee Shu-Fune Wu, Gwendolyn Clemens, and Alfred Sommer who contributed to the design and implementation of the study.
References Ahluwalia IB, Grummer-Strawn L & Canlon KS (1997): Exposure to environmental tobacco smoke and birth outcome: increased effects on pregnant women aged 30 years or older. Am. J. Epidemiol. 146, 42–47.
European Journal of Clinical Nutrition
Christian P, West Jr KP, Khatry SK, Katz J, LeClerq S, KimbroughPradhan E & Shrestha SR (1998a): Vitamin A or b-carotene supplementation reduces but does not eliminate maternal night blindness in Nepal. J. Nutr. 128, 1458–1463. Christian P, West Jr KP, Khatry SK, Katz J, Shrestha SR, Pradhan EK, LeClerq SC & Pokhrel RP (1998b): Night blindness of pregnancy in rural Nepal—nutritional and health risks. Int. J. Epidemiol. 27, 231–237. Conde-Agudelo A, Althabe F, Belizan JM & Kafury-Goeta AC (1999): Cigarette smoking during pregnancy and risk of preeclampsia: a systematic review. Am. J. Obset. Gynecol. 181, 1026–1035. Conde-Agudelo A & Belizan JM (2000): Risk factors for pre-eclampsia in a large cohort of Latin American and Carribean women. BJOG 107, 75–83. Craft NE (1996): High resolution HPLC method for the simultaneous analysis of carotenoids, retinoids, and tocopherols. FASEB J. 10, A3039. Cross CE, Traber M, Eiserich J & van der Vliet A (1999): Micronutrient antioxidants and smoking. Br. Med. Bull. 55, 691–704. Dallongeville J, Marecaux N, Fruchart JC & Amouyel P (1998): Cigarette smoking is associated with unhealthy patterns of nutrient intake: a meta-analysis. J. Nutr. 128, 1450–1457. Dreyfuss ML, Stoltzfus RJ, Shrestha JB, Pradhan EK, LeClerq SC, Khatry SK, Shrestha SR, Katz J, Albonico M & West Jr. KP (2000): Hookworms, malaria, and vitamin A deficiency contribute to anemia and iron deficiency among pregnant women in the plains of Nepal. J. Nutr. 130, 2527–2536. Haste FM, Brooke OG, Anderson, HR Bland JM, Shaw A, Griffin J & Peacock JL (1990): Nutrient intakes during pregnancy: observations on the influence of smoking and social class. Am. J. Clin. Nutr. 51, 29–36. Haste FM, Brooke OG, Anderson HR & Bland JM (1991): The effect of nutritional intake on outcome of pregnancy in smokers and nonsmokers. Br. J. Nutr. 65, 347–354. Institutes of Medicine (IOM) (1990): Subcommittees on Nutritional Status and Weight Gain during Pregnancy and Dietary Intake and Nutrient Supplements during Pregnancy. Food and Nutrition Board. Nutrition During Pregnancy: weight Gain; Nutrient Supplements. Washington DC: National Academy Press. Katz J, West Jr KP, Khatry SK, Pradhan EK, LeClerq SC, Christian P, Wu LS, Adhikari RK, Shrestha SR, Sommer A & the NNIPS-2 Study Group (2000): Maternal low-dose vitamin A or b-carotene supplementation has no effect on fetal loss and early infant mortality: a randomized cluster trial in Nepal. J. Nutr. 71, 1570–1576. Kleinman JC, Pierre MB, Madans JH, Land GH & Schramm WF (1988): The effects of maternal smoking on fetal and infant mortality. Am. J. Epidemiol. 127, 274–282. Lieff S, Olshan AF, Werler M, Strauss RP, Smith J & Mitchell A (1999): Maternal cigarette smoking during pregnancy and risk of oral clefts in newborns. Am. J. Epidemiol. 150, 683–694. Lindley AA, Becker S, Gray RH & Herman AA (2000): Effect of continuing or stopping smoking during pregnancy on infant birth weight, crown-heel length, head circumference, ponderal index, and brain:body weight ratio. Am. J. Epidemiol. 152, 219–225. Lorente C, Cordier S, Goujard J, Ayme S, Bianchi F, Calzolari E, De Walle HE & Knill-Jones R (2000): Tobacco and alcohol use during pregnancy and risk of oral clefts. Occupational exposure and congenital malformation working group. Am. J. Pub. Health 90, 415–419. Masironi R & Rothwell K. (1988): Trends in and effects of smoking in the world. World Health Stat. Q 41, 228–241. Mathews F, Yudkin P, Smith RF & Neil A. (2000): Nutrient intakes during pregnancy: the influence of smoking status and age. J. Epidemiol. Commun Health 54, 17–23. Ness RB, Grisso JA, Hirschinger N, Markovic N, Shaw LM, Day NL & Kline J (1999): Cocaine and tobacco use and the risk of spontaneous abortion. N. Engl. J. Med. 340, 333–339.
Smoking and pregnancy-related risks in Nepal P Christian et al
211 Omenn GS, Goodman GE, Thornquist MD, Balmes J, Culler MR, Glass A, Keogh JP, Meyskens FL, Valanis B, Williams JH, Barnhart S & Hammar S (1996): Effects of a combination of beta carotene and vitamin A on lung cancer and cardiovascular disease. N. Engl. J. Med. 334, 1150–1155. Pollack H, Lantz PM & Frohna JG (2000): Maternal smoking and adverse birth outcomes among singletons and twins. Am. J. Public Health 90, 395–400. Pollack HA (2001): Sudden infant death syndrome, maternal smoking during pregnancy, and the cost effectiveness of smoking cessation intervention. Am. J. Publ. Health 91, 432–436. Preston AM (1991): Cigarette smoking—nutritional implications. Prog. Food Nutr. Sci. 15, 183–217. Rasmussen KM & Adams B (1997): Annotation: cigarette smoking, nutrition, and birth weight. Am. J. Publ. Health 87, 543–544. Roquer JM, Figueras J, Botet F & Jimenez R (1995): Influence on fetal growth of exposure to tobacco smoke during pregnancy. Acta Paediatr. 84, 118–121. Sadler L, Belanger K, Saftlas A, Leaderer B, Hellenbrand K, McSharry JE & Bracken MB (1999): Environmental tobacco smoke exposure and small-for-gestational age birth. Am. J. Epidemiol. 150, 695–705. Salafia C & Shiverick K (1999): Cigarette smoking and pregnanacy II. Vascular effects. Placenta 20, 273–279. The Alpha-Tocopherol, Beta-carotene Cancer Prevention Study Group (1994): The effect of vitamin E and beta carotene on the incidence of lung cancer and other cancers in male smokers. N. Engl. J. Med. 330, 1029–1035. Traber MG, van der Vliet A, Reznick AZ & Cross CE. (2000): Tobaccorelated diseases. Is there a role for antioxidant micronutrient supplementation? Clin. Chest. Med. 21, 173–187. Trygg K, Lund-Larsen K, Sandstad B, Hoffman JH, Jacobsen G & Bakketeig LS. (1995): Do pregnant smokers eat differently from pregnant non-smokers? Paediatr. Perinat. Epidemiol. 9, 307–318.
Tuthill DP, Stewart JH, Coles EC, Andrews J & Cartlidge PHT (1999): Maternal cigarette smoking and pregnancy outcome. Paediatr. Perinatal. Epidemiol. 13, 245–253. United Nations Children’s Fund (UNICEF) (2000): State of the World’s Children’s Report. New York: UNICEF. West Jr KP, Katz J, Khatry SK, LeClerq SC, Pradhan EK, Shrestha SR, Connor PB, Dali SM, Christian P, Pokhrel RP & Sommer A on behalf of the NNIPS-2 Study Group (1999): Double blind, cluster randomized trial of low dose supplementation with vitamin A or b-carotene on mortality related to pregnancy in Nepal. BMJ 318, 570–575. Windham GC, Hopkins B, Fenster L & Swan SH (2000): Prenatal active or passive tobacco smoke exposure and the risk of preterm delivery or low birth weight. Epidemiology 11, 427–433. Wisborg K, Kesmodel U, Henriksen TB, Olsen SF & Secher NJ (2000): A prospective study of smoking during pregnancy and SIDS. Arch. Dis. Child. 83, 203–206. Wu T, Buck G & Mendola P (1998a): Maternal cigarette smoking, regular use of multivitamin/mineral supplements, and risk of fetal death. The 1998 National Maternal and Infant Health Survey. Am. J. Epidemiol. 148, 215–221. Wu T, Buck G & Mendola P (1998b): Can regular multivitamin/ mineral supplement modify the relationship between maternal smoking and select adverse birth outcomes? Ann. Epidemiol. 8, 175–183. Yamini S, West Jr KP, Wu L, Dreyfuss M, Yang DX & Khatry SK (2001): Circulating levels of retinol, tocopherol and carotenoid in Nepali pregnant and postpartum women following long term b-carotene and vitamin A supplementation. Eur. J. Clin. Nutr. 55, 252–259. Zhou JF, Yan XF, Guo FZ, Sun NY, Qian ZJ & Ding DY (2000): Effects of cigarette smoking and smoking cessation on plasma constituents and enzyme activities related to oxidative stress. Biomed. Environ. Sci. 13, 44–55.
European Journal of Clinical Nutrition
