Nicotine & Tobacco Research Advance Access published December 22, 2014 Nicotine & Tobacco Research, 2014, 1–9 doi:10.1093/ntr/ntu255 Review

Review

Maternal Smokeless Tobacco Use in Pregnancy and Adverse Health Outcomes in Newborns: A Systematic Review Downloaded from http://ntr.oxfordjournals.org/ at North Dakota State University on June 1, 2015

Aadil S. Inamdar MSc1, Raymond E. Croucher PhD2, Mrunalini K. Chokhandre MSc3, Mohammed H. Mashyakhy MSc4, Valeria C. C. Marinho PhD5 Department of Preventive Dental Sciences, College of Dentistry, Jazan University, Jazan, Kingdom of Saudi Arabia; 2Barts and The London, School of Medicine and Dentistry, Queen Mary University of London, London, UK; 3Department of Clinical Pharmacology, Faculty of Medicine, Jazan University, Jazan, Kingdom of Saudi Arabia; 4Department of Restorative Dental Sciences, College of Dentistry, Jazan University, Jazan, Kingdom of Saudi Arabia; 5Centre for Clinical and Diagnostic Oral Sciences, Barts and The London, School of Medicine and Dentistry, Queen Mary University of London, London, UK 1

Corresponding Author: Aadil S. Inamdar, MSc, Department of Preventive Dental Sciences, College of Dentistry, Jazan University, PO Box 114, Jazan 45142, Kingdom of Saudi Arabia. Telephone: 966-535368035; Fax: 966-173310438; E-mail: [email protected]

Abstract Background: Perinatal morbidity and mortality are important indicators of maternal health and the future health of the child. Smokeless tobacco (ST) use during pregnancy is associated with low birth weight (LBW), preterm births, stillbirths, and small for gestation age (SGA). This study systematically reviews and summarizes evidence on the association of maternal ST use with these adverse health outcomes in newborns. Methods: Electronic databases (Medline, Embase, Lilacs) were searched in July 2013 using appropriate keywords complemented with reference list searching. Observational studies of maternal ST use and these outcomes were considered; LBW, preterm, stillbirth, SGA. A  comprehensive assessment of quality and risk of bias in all included studies was performed. RevMan software was used for data analysis. Results are expressed as crude odds ratios with 95% confidence intervals. Chi-square and I2 tests checked for heterogeneity and quantified inconsistency between results. Results: There were 9 studies (16 reports) included (7 cohort-studies, 1 case-control study, and 1 cross-sectional study). They were clinically and methodologically diverse. Significant associations with ST use were seen in 5/7 studies for LBW, in 3/6 studies for preterm, in all 4 studies for stillbirth and in 1/2 studies assessing SGA. Heterogeneity between results was moderate for LBW (I2 = 44%) and stillbirth (I2 = 52%), and high for preterm (I2 = 87%) and SGA (I2 = 65%). Meta-analysis was considered inappropriate due to risk of bias and confounding. Conclusions: Although most studies show an association between ST use in pregnancy and adverse health effects in newborns, these results may be limited by confounding and bias. Quality observational studies are needed to strengthen this evidence base.

Introduction Perinatal morbidity and mortality are important indicators of maternal health and the future health of the child.1 Adverse pregnancy outcomes such as low birth weight (LBW) accounts for 15% of infants

born in this world. The United Nations Children’s Fund (UNICEF) reports 45% of newborns being LBW in South-east Asia and Eastern Mediterranean.2 Moreover, nearly eight million perinatal deaths occur each year globally.3,4 Ninety eight percent of these deaths occur in the

© The Author 2014. Published by Oxford University Press on behalf of the Society for Research on Nicotine and Tobacco. All rights reserved. For permissions, please e-mail: [email protected].

1

2

Methods Selection Criteria All published and unpublished cohort, case-control and cross-sectional studies were considered which addressed maternal exposure

to ST during pregnancy and adverse outcomes in the newborns (LBW, preterm birth, stillbirth, and SGA). Females at any stage of pregnancy between conception and birth in any health care setting were considered. ST of at least one form such as snuff, plug, quid, snus and gutka, with or without additives, used in any amount or frequency, during pregnancy, or at some time during pregnancy, was considered. Concurrent use of cigarette smoking or other forms of tobacco was only considered for inclusion when adjusted for in the analysis (but not set as a prerequisite for inclusion). Studies of oral forms of khat and betel quid that do not contain tobacco were excluded. At least one of these adverse pregnancy outcomes in newborns was considered (although closely related, these outcomes have been defined separately by the World Health Organization [WHO] International Classification of Diseases [ICD])36: “LBW” (ICD P07.1 birth weight < 2,500 g [1,000–2,499 g]), “preterm” (ICD O60.1 [infant born sooner than 37 weeks calculated from the beginning of the last menstrual period]), “stillbirth” (ICD P95 [an infant delivered without life after 24 weeks of gestation]) and “SGA” (ICD P05.1 [weight and length below 10th percentile for gestational age]) infants. Adverse pregnancy outcomes related to the mothers were not considered.

Search Methods Electronic databases (Medline, Embase, Lilacs) were comprehensively searched, using appropriate keywords and controlled vocabulary terms in July 2013, without date or language restrictions (Supplementary Data 1). This was supplemented by searching the reference lists of relevant studies and by contacting selected authors for unpublished studies.

Data Collection and Analysis Selection of Studies Two reviewers (ASI and MKC) examined titles, keywords, and abstracts of studies and identified reports potentially eligible for full text assessment. Any conflicts were clarified by a third reviewer (MHM). The full texts of all potentially relevant reports were obtained and assessed using predetermined inclusion and exclusion criteria. Data Extraction and Quality (Risk of Bias) Assessment A data extraction sheet was prepared based on a previous systematic review37 and conforms to the guidelines from the Cochrane Handbook for systematic reviews.38 The form included three separate sections for cohort studies, case-control studies, and cross-sectional studies. Data extraction was carried out independently by two reviewers (ASI and MKC) and a fourth reviewer resolved doubts (VCCM). Data was collected from each included study on the external and internal validity, control of/adjustment for confounding, and characteristics/definitions of the exposure and outcome measures. Crude and adjusted odds ratios (OR) were extracted from all reports where available. Where included studies were described in more than one publication (report), we used the primary study reference to inform our quality assessment. In addition, for each outcome, data from the most recent of these reports, with bigger pool of patients was used for analysis and graphical representations in the forest plot.

Data Synthesis RevMan software (V5.2) was used to analyze the data for heterogeneity and produce a graphical display of results. Results were expressed as crude OR and their 95% confidence intervals (CI).

Downloaded from http://ntr.oxfordjournals.org/ at North Dakota State University on June 1, 2015

developing countries.5 Stillbirths accounts for 60% of these perinatal deaths.6 Preterm births are associated with neonatal mortality and pulmonary, ophthalmologic and neurologic morbidity.7 Small for gestational age (SGA) is associated with an increased risk of stillbirth, neonatal death, hypoglycemia, hypocalcaemia, and polycythemia and over the long term permanent deficits in growth and neurocognitive development.8 LBW and SGA infants are linked with diminished cognitive growth and development and chronic diseases later in life.9 Although cigarette smoking is the most common form of tobacco used worldwide, smokeless tobacco (ST) is estimated to be used by 300 million people globally and is gaining popularity in many parts of the world.10–12 ST users are not exposed to the toxins which are produced as a result of the burning of tobacco and paper but are still exposed to high levels of nicotine, which not only cause dependence, but may also lead to a variety of other health hazards. Absorption of nicotine from different tobacco products depends on the quantity of tobacco by weight and its total nicotine content rather than the product form.13 Nicotine readily crosses the placental barrier, with slightly higher concentrations in the fetal serum and amniotic fluid,14,15 causing fetotoxic effects induced by the oxidative and endoplasmic reticulum stress responses in the fetal trophoblastic cells.16 Detailed mechanisms of these toxicological effects of nicotine on human embryonic development are beginning to emerge, proposing the role of nicotine in increased inflammatory processes, negative regulation of gene expression, cell proliferation and cell differentiation as well as differentiation of bone tissue.17 ST use is prevalent in females of reproductive age groups in Asia, with a higher trend in teenagers.10,18,19 In several European countries (Sweden, Norway) and parts of Africa and America, ST is either used orally (chewed) or used nasally (snuff),20–23 while it is consumed with other ingredients (areca nut [betel], betel leaf, and lime) in the South Asian diaspora.24 Maternal smoking has already been linked to adverse health effects in the newborns25 and the use of ST in pregnancy could be as detrimental as cigarette smoking to the health of the newborn child.26 Reducing maternal and child morbidity and mortality are a challenge and pose a threat to achieving the Millennium Development Goals.27,28 Infants with health problems need extra special care, increasing avoidable burden to prevent and treat their conditions and results in substantial costs to the health sector.29 ST use during pregnancy has indeed been associated with perinatal morbidity and mortality,30 but it is important to appreciate the differences between associations and effects.31 Findings from observational studies appear to support this association but the evidence appears limited and unclear. There is currently no review done systematically that addresses this association. Reviews on ST use have only a minor section addressing adverse pregnancy outcomes,32,33 or they report on maternal and neonatal health effects in general.34,35 It is necessary to address this important gap and assess systematically, the available evidence on the association of prenatal ST exposure on the newborns. We report a systematic review where the aim was to identify studies investigating the association between maternal ST use during pregnancy and adverse health effects in newborns and to assess and summarize the main features, methodological quality and results of the included studies.

Nicotine & Tobacco Research, 2014, Vol. 00, No. 00

Nicotine & Tobacco Research, 2014, Vol. 00, No. 00 An OR of more than one indicates increased risk of the adverse effect (association) with ST use. Statistical significance was shown by a p value less than .05. For testing the heterogeneity between results of studies and quantifying inconsistency, chi-square (χ2) and I2 tests were performed.39 Studies are presented according to the outcome measure reported and grouped/stratified according to the study design as cohorts, casecontrols, and cross-sectional studies. Data was to be pooled according to each relevant outcome (LBW, preterm, stillbirth, SGA).

Results

mainly from self-reports through questionnaires and personal interviews, from public and hospital records and sociodemographic data. There was a high risk of recall bias and performance bias (misclassification of exposure) in all studies where information on ST use was obtained mainly from self-reports (none of the studies reported ascertaining the levels of nicotine/cotinine by any biochemical measurements). Diagnosis of outcomes was obtained mainly through public hospital records (five studies), birth registers (one study) and questionnaires (seven studies). Two studies used standardized (WHO) birth weight measures.40,42 One study reported using crude methods of measuring weight.44 Only two studies reported recording of birth weight by trained personnel.46,54 Gestational age was calculated as days since the last menstrual period to the date of delivery together with ultrasound scans, in the seven reports41,46,49–53 of the two studies for SGA. There was a high chance of detection bias (misclassification of outcome), as the studies were unclear in reporting blinded outcome assessment (none of the cohort or case control studies mention abstractors being blinded to the “ST user” status of the mother). Statistical adjustments for measured confounding factors were not made in three44,45,48 of the nine included studies. Although factors such as maternal age, education of the mother, BMI and SES were generally considered, the adjusted confounders differed across studies in number and type, (that is, no study measured the same set of confounding variables and a few did not consider any). In addition, studies failed to adjust for important confounders such as smoking, alcohol, drug and substance abuse, diet and nutrition, and underlying maternal medical conditions.

Effects of Exposure Studies included in the review were both clinically and methodologically diverse. Crude ORs were not presented in four40,44–46 of the seven cohort studies, adjusted ORs were also not presented in four43–46 of the seven cohort studies and in the cross-sectional48 study. Heterogeneity between study results was moderate for LBW and stillbirth and high for preterm and SGA. Pooling of data in meta-analyses was not considered appropriate due to risk of bias and confounding in the studies. Nevertheless, most studies show an association between ST use and LBW, preterm, and stillbirths. The forest plots with the crude results (OR) from the studies included for each relevant outcome are presented as a useful visual aid to interpreting findings (Figures 1 and 2 and Supplementary Figures 2 and 3). Synthesis focused on describing the magnitude, precision, direction and consistency of effects. Not all studies provided adjusted ORs, but they are reported in this review where available. The ORs greater than one suggest that there is a greater risk of having a LBW, preterm, SGA, or a stillbirth with ST use in pregnancy.

LBW There was an association between LBW and maternal ST use in five40,42,43,45,48 of the seven studies assessing this relationship (Figure 1). Crude results of four of the five cohort studies indicate a significant higher risk of LBW with ST use, with point estimates of crude/unadjusted ORs varying from 1.61 (95% CI = 1.11–2.35) to 2.49 (95% CI = 1.34–4.61). Adjusted ORs of 3.14 (95% CI = 2.08–4.88)40 and 1.6 (95% CI  =  1.1–2.4)42 reported in two studies also showed an increased risk of LBW with ST use. The only cross-sectional study for LBW used data from a survey and also showed a significant result: an OR of 3.21 (95% CI = 1.70–6.06), which indicated that the newborns of ST users were thrice more likely to be born with LBW.48 Estimates

Downloaded from http://ntr.oxfordjournals.org/ at North Dakota State University on June 1, 2015

The search yielded a total of 584 papers, (385 papers from Medline, 179 from Embase, 12 from Lilacs, and eight from reference list search). After the duplicates were removed and reports examined by title, keywords and abstracts, 23 reports were considered potentially eligible for inclusion and assessed as full-texts based on the predetermined inclusion and exclusion criteria. There were eight reports excluded from this review (of eight studies), and 16 reports (of nine studies) included (Supplementary Figure 1). The Supplementary Table 1 shows the excluded studies with reasons for exclusion. The nine studies included are listed in Table  1 (only primary study references cited). There were seven cohort studies,40–46 one case-control study,47 and one cross-sectional study.48 Three of the included studies had more than one report about each. For the Swedish study primary report,41 five additional reports49–53 were identified. For the Mumbai-India study primary report,42 another report54 was identified, while for the other ParelMumbai-India study primary report48 another report55 was also identified. There were six studies from Asia (five from India and one from Bangladesh), one study from Sweden, one from the United States and one from South Africa. The largest number of subjects involved for an outcome (stillbirth) was from a report of the Swedish study with 676,499 participants.51 The predominant form of ST used in Asian studies was the commercially available chew (zarda, ala-pata, biripatti, leaf-powder, gul, mishri, gutka, paan-masala). Snuff (snus) was used by the Swedish and South African subjects, while Iqmik and the commercial chew, by subjects in the U.S. study. Most studies did not report the measure of exposure. ST users were categorized as light or heavy according to the frequency of intake in a day in the two reports42,54 of the Mumbai-India study. While another study grouped users according to the amount of ST intake per day,44 others reported use of ST, at least once per day. There were seven studies reporting on LBW,40,42–46,48 six reporting on preterm birth,42,44–47,49 four studies were reporting on stillbirth,44,45,51,54 and two studies reporting on SGA.46,50 Among the common measured confounding factors in the included studies were: maternal age (six studies), education of mother (four studies), sex of child (three studies), maternal height (three studies), socioeconomic status (SES) (four studies), antenatal care visits (two studies), body mass index (BMI) (three studies), parity (five studies), anemia (one study), chronic hypertension and pre-gestational diabetes (one study), and smoking (three studies). However, three studies did not report on any confounding factors.44,45,48 Included studies show high variation in the methodological quality, as summarized in Table  2 and Supplementary Tables 2 and 3. Studies done in India and Bangladesh were small in size with low external validity. Information on the ascertainment of ST use was

3

40

CT; India

Study type; Country

CT; Bangladesh

CT; India

CS; India

CT; India CT; South Africa

(Hosain, 2006)43

(Krishna, 1978)44

(Mehta, 1990)b48,55

(Pratinidhi, 2010)45 (Steyn, 2006)46

Commercial chew

Type of ST NR

Measure and frequency of ST use

Snuff (snus) Daily use Preterm 8,321(615,751), stillbirth 9,198(676,499), SGA 9,129(672,778) 359(532) Iqmik, commercial chew Daily use Light: 1×/day/ LBW 168(974), preterm Commercial chew past 6 months, and stillbirth 202(1,110) (mishri, Gutka, paan Heavy:5×/day/past masala) 6 months 187(350) Commercial chew, NR (Zarda, ala-pata, gul) Stillbirth 220(1,388), Commercial chew 50–100 mg/day preterm 209(1,148) 65(178) Commercial chew Daily use (mishri) 218(705) Mishri Daily use 120(1,496) Snuff (snus) Daily use

78(201)

n(N)

Q Q, HR

Q

Q, HR

Q, HR

HR Q, HR

BR

Q

Not done or NR Not done or NR MA, S, Sx, ME, alcohol, SES

LBW LBW, Preterm, stillbirth LBW, preterm, SGA

Not done or NR

LBW, preterm, stillbirth

MA, MH, parity, BMI MA, ME, MW, parity, SES, ANC, S

Preterm LBW, preterm, stillbirth

MA, Sx, ME, parity, SES, ANC

MA, MH, ME, BMI, S, Sx, parity, CH, PGD

Preterm, stillbirth, SGA

LBW

MA, MH, BMI, anemia, parity, SES

LBW

Data source Outcome measures reported Adjustments for confounders

Downloaded from http://ntr.oxfordjournals.org/ at North Dakota State University on June 1, 2015

Note. ANC = ante natal care; BR = birth register; BMI = body mass index; CC = case control; CH = chronic hypertension; CS = cross-sectional; CT = cohort; HR = hospital records; LBW = low birth weight; MA = maternal age; ME = maternal education; MH = maternal height; MW = maternal weight; n = number of ST users; N = total number of subjects (ST + non-ST users) in the study; NR = not reported; PGD = pre-gestational diabetes; Q = interview questionnaire; S = smoking; SES = socioeconomic-status; SGA = small for gestation age; ST= smokeless tobacco; Sx = sex of infant. a All six reports of the study used. b Both reports of the study used.

CC; United States CT; India

(England, 2013)47 (Gupta, 2004)b42,54

(England, 2003)a41,49–53 CT; Sweden

(Deshmukh, 1998)

Study IDref

Table 1. Included Studies and Their Characteristics

4 Nicotine & Tobacco Research, 2014, Vol. 00, No. 00

Performance

Internal validity

Detection

Attrition

Selection bias and control of confoundinge

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

* *

* *

*

* *

*

*

*

*

*

*

*

*

*

* *

*

*

*

*

*

*

*

Assessors BioAttrition rate Groups same for both from same Participation ST user status chemical Diagnosis of Confirmation blinded to ST Equal of outcome? user status? follow-upc groups? Completeness d community? MA BMI S SES D SA O ascertained? measures? outcome? Representativenessa rateb

External validity

COHORTS

Downloaded from http://ntr.oxfordjournals.org/ at North Dakota State University on June 1, 2015

Note. A (*) indicates the measure was adequately addressed in the study. The absence of a (*) indicates that either the factor was not measured, or not reported, or if it was, it did not meet the (*) requirement. BMI = body mass index; D = diet and nutrition; MA = maternal age; O = other including maternal weight, working status, number of antenatal care visits, place of delivery, previous stillbirths and caesareans, underlying maternal medical conditions, etc; S = smoking; SA = substance abuse, drug, alcohol; SES = socioeconomic status; ST= smokeless tobacco. a Studies received a (*) if the sample included all eligible women from a defined area, or a random sample of those women. b Studies received a (*) if the percentage of participation was 80% or more. c Studies received a (*) if groups were followed within 10% of each other or for the same time. d Studies received a (*) if a full description of those lost-to-follow-up was not biased or the participants in the final analysis was 80% or more. e For control of confounding a (*) indicates that the group was adjusted for shown factors in analysis. f All six reports of the study. g Both reports of the study.

(Deshmukh 1998)40 (England, 2003)f41,49–53 (Gupta, 2004)g42,54 (Hosain, 2006)43 (Krishna, 1978)44 (Pratinidhi, 2010)45 (Steyn, 2006)46

Study IDref

Table 2. Quality Assessment of the Included Studies

Nicotine & Tobacco Research, 2014, Vol. 00, No. 00 5

Nicotine & Tobacco Research, 2014, Vol. 00, No. 00

6 of effects for all six studies providing data on LBW varied from ORs of 0.90 (95% CI = 0.44–1.82) to 3.21(95% CI = 1.70–6.06) (complete data was not available for a study44 to be included in the forest plot). The test for heterogeneity among the cohort studies was not significant (χ2 = 7.11, degree of freedom [df] = 4 [p = .13]), but there was moderate variability observed between the studies (I2  =  44%), that is, 44% of the variability between the studies was attributable to between-study differences and not random variation.

(Supplementary Figure 2). Crude results consistently show a higher risk of stillbirth with ST use, with point estimates of unadjusted ORs varying between 1.60 (95% CI  =  1.15–2.21) to 4.57 (95% CI  =  1.13–18.43) in the four reports. Only one study reported adjusted OR of 2.6 (95% CI = 1.4–4.8) which was significant.54 The test for heterogeneity was not significant (χ2 = 6.22, df = 3 [p < .10]). There was moderate heterogeneity between studies (I2 = 52%).

SGA Preterm

Stillbirth There were four studies (cohorts) addressing stillbirth44,45,51,54 and all showed a significant association of ST use and stillbirth

Figure 1. A forest plot of maternal smokeless tobacco use and low birth weight.

Figure 2. A forest plot of maternal smokeless tobacco use and preterm.

Discussion This is the first systematic review addressing maternal ST use and LBW, preterm, SGA, and stillbirths in the newborns. A  PubMed search for systematic reviews on ST use in pregnancy retrieved 18 reports. None of them focuses on the effects of ST use during pregnancy in the newborns. Four of these reviews reported findings on ST use during pregnancy, one of which reported on general health56 while the others had only a small section on adverse pregnancy outcomes.32–34 Their conclusions are consistent with those reported in this review, pointing towards a positive association between maternal ST use and adverse outcomes in the newborns. This review has its strength in its comprehensive coverage, without limiting the search for studies by publication status, language or geographical area. The methodology, conduct and reporting of this systematic

Downloaded from http://ntr.oxfordjournals.org/ at North Dakota State University on June 1, 2015

There were statistically significant associations of preterm birth with maternal ST use in three 42,44,49 of the six studies looking at this association (Figure  2). Estimates of effects were varying from ORs of 0.69 (95% CI = 0.16–2.91) to 5.38 (95% CI = 3.24–8.92) among all six studies. Of the five cohort studies, three showed a higher risk of preterm birth with ST use, with point estimates of unadjusted ORs varying between 1.32 (95% CI = 1.21–1.45) to 5.38 (95% CI = 3.24–8.92) (two studies45,46 showed nonsignificant results, one of them in the other direction46). Three studies reported adjusted OR’s, where two studies indicated a higher risk of preterm birth with ST use with adjusted ORs of 1.4 (95% CI = 1.0–2.1)42 and 1.29 (95% CI = 1.17–1.43)49. The only case-control study reporting on this outcome, showed a nonstatistically significant association of maternal ST use and preterm deliveries, with an unadjusted OR of 1.44 (95% CI = 0.99–2.11) and adjusted OR of 1.23 (95% CI = 0.78–1.93).47 The test for heterogeneity for cohort studies reporting preterm was statistically significant (χ2 = 30.21, df =4 [p < .0001]), and the studies showed high variability (I2 = 87%).

Two studies (cohorts) addressed ST use and SGA46,50 (Supplementary Figure  3), where one report indicated a significant higher risk of SGA with ST use, with point estimates of crude OR of 1.29 (95% CI = 1.13–1.49).50 The other report showed a nonsignificant result in the other direction.46 Although the test for heterogeneity was not significant (χ2  =  2.87, df  =  1 [p  =  .09]), there was high variability (I2 = 65%) observed between the two studies’ results.

Nicotine & Tobacco Research, 2014, Vol. 00, No. 00

studies will present the best evidence available to reflect on any association.59 The clinical strength of these observational studies would be in identifying the specific type of ST used, with importance to the dose and frequency of intake. In addition, relevant confounders including maternal smoking and drug abuse should be identified early and excluded or adjusted for in the analysis. Furthermore, the current systematic review addressed outcomes in the newborn that are evaluated immediately after birth (LBW, preterm, SGA, and stillbirth). However, studies of other potential adverse pregnancy outcomes in newborns like neonatal deaths, gross congenital anomalies, metabolic, endocrine, immune disorder, or any high-risk conditions can be potentially considered. Observational studies in this review indicate an increased risk of LBW, preterm, and stillbirth with maternal ST use. However, such results may be limited by confounding and bias, and should be viewed with caution. Further studies can address at least three areas of research with ST. Firstly, a clearer differentiation between the types of ST used, as they vary widely in content. Secondly, studies assessing health effects in relation to the amount of ST used.60 Lastly, the effects of ST may not be limited to these four adverse newborn outcomes (LBW, preterm, stillbirth, SGA), and further associations can be explored. Quality observational studies are needed to strengthen the current evidence base.

Supplementary Material Supplementary Data 1, Tables 1–3, and Figures 1–3 can be found online at http://www.ntr.oxfordjournals.org

Funding None declared.

Declaration of Interests None declared.

References 1. Filippi V, Ronsmans C, Campbell OM, et  al. Maternal health in poor countries: the broader context and a call for action. Lancet. 2006;368:1535–1541. 2. The United Nations Children’s Fund (UNICEF). Children in an Urban World: The State of the World’s Children 2012. 2012. http://www.unicef. org/sowc/files/SOWC_2012-Main_Report_EN_21Dec2011.pdf. Accessed September 8, 2013. 3. The United Nations Children’s Fund (UNICEF), World Health Organization (WHO). Low Birthweight: Country, regional and global estimates. UNICEF N.  Y. 2004. http://www.unicef.org/publications/files/ low_birthweight_from_EY.pdf. Accessed September 8, 2013. 4. WHO. World Health Statistics 2012—World Health Organization. 2012. http://apps.who.int/iris/bitstream/10665/44844/1/9789241564441_eng. pdf. Accessed September 8, 2013. 5. World Health Organization (WHO). WHO | The world health report 2004—changing history. Geneva, Switzerland: WHO; 2004. http://www. who.int/whr/2004/en/. Accessed November 1, 2014. 6. Stanton C, Lawn JE, Rahman H, Wilczynska-Ketende K, Hill K. Stillbirth rates: delivering estimates in 190 countries. Lancet. 2006;367:1487–1494. 7. Beck S, Wojdyla D, Say L, et al. The worldwide incidence of preterm birth: a systematic review of maternal mortality and morbidity. Bull World Health Organ. 2010;88:31–38.

Downloaded from http://ntr.oxfordjournals.org/ at North Dakota State University on June 1, 2015

review conforms to the PRISMA statement.57 Also, our assessment of risk of biases and methodological quality was extensive, using an appropriate assessment tool which let us make a more informed judgment regarding the inappropriateness of pooling the results in a meta-analysis. There are currently seven cohort studies, one case control study and one cross-sectional study assessing the association of maternal ST use with adverse outcomes in the newborn (LBW, preterm, stillbirth, SGA). Although most of the studies included reported on LBW and preterm deliveries, few reported on SGA and stillbirths. Out of the nine studies included in this systematic review seven studies point towards an association of maternal ST use with LBW, preterm, and stillbirth. Two to three fold increases in the risk of LBW and preterm was shown with ST use in pregnancy and similarly large associations were consistently indicated in the studies assessing stillbirth, measured by crude ORs. However, poor control of confounding factors and other methodological issues do undermine the studies included in this review. A high chance of performance and detection biases was present, as none of the studies confirmed the “ST user” status of the participants directly by any bio-chemical measures, and the assessors of outcomes were not blinded to the “ST user” status of the mother. The blinded status of assessors may not be of vital importance for SGA or stillbirth, as there is a highly unlikely chance of introducing bias, but these have been still reported in this review. Studies included in this review varied greatly in location—they were from Asia, Sweden, South Africa, and Alaska (United States). The type and use of ST also varied amongst participants of these observational studies, the difference most likely to be cultural or demographic. ST, in the Asian subcontinent is used in many forms and ingredients (zarda, ala-pata, gul, mishri, gutka, paan-masala, etc.), whereas snus/snuff use is prevalent in parts of Europe, Africa, and America. These two major types of ST differ in tobacco content, doses, and frequencies of use and have different patho-physiological effects, which can be potential confounders. Confounding factors varied across all studies included in this review and may be attributed to the study location and type of exposure. Observational studies, unlike randomized controlled trials, can only adjust for known confounders and only those that are measured without error. Some studies either did not measure or did not report any confounders. Among studies that did report confounders, they were variable and statistical adjustment was only attempted in a few. But important confounders such as smoking, alcohol, substance and drug abuse and underlying maternal health condition were not generally measured across studies. The effect of unknown confounders may well be operating in either direction within and across all of the included studies. Furthermore, misclassification of confounders can greatly hinder the effectiveness of any statistical adjustment procedure.58 The following study limitations should be acknowledged. Firstly, although the search strategy was comprehensive and the main databases have been covered, there are chances of studies not indexed, being missed out. The authors were also contacted, but were unable to provide us with unpublished or missing data. Secondly, there are many perinatal morbidity and mortality outcomes in the newborn, some associated with the prenatal ST use by mother, but only four major conditions (LBW, preterm, SGA, and stillbirth) have been explored in this review. Considering the nature of exposure (ST use in pregnant females), randomized interventional studies would not be practically possible. In the absence of controlled trials, high quality observational

7

8

31. Petitti DB. Associations are not effects. Am J Epidemiol. 1991;133:101–102. 32. Critchley JA, Unal B. Health effects associated with smokeless tobacco: a systematic review. Thorax. 2003;58:435–443. 33. England LJ, Kim SY, Tomar SL, et  al. Non-cigarette tobacco use among women and adverse pregnancy outcomes. Acta Obstet Gynecol Scand. 2010;89:454–464. 34. Broadstock M. Systematic Review of the Health Effects of Modified Smokeless Tobacco Products. Christchurch, New Zealand: New Zealand Health Technology Assessment, Department of Public Health and General Practice, Christchurch School of Medicine and Health Sciences; 2007. http://nzhta.chmeds.ac.nz/publications/smokeless_tobacco.pdf. Accessed September 4, 2013. 35. Ratsch A, Bogossian F. Smokeless tobacco use in pregnancy: an integrative review of the literature. Int J Public Health. 2014;59:1–10. doi:10.1007/ s00038-014-0558-6. 36. WHO. ICD-10: International statistical classification of diseases and related health problems - World Health Organization. 2004. http://apps. who.int/classifications/icd10/browse/2010/en. Accessed September 8, 2013. 37. Siegfried N. Does male circumcision prevent HIV infection? PLoS Med. 2005;2:e393. 38. Higgins JP, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions. Version. 5.1.0 [updated March  2011]. The Cochrane Collaboration. 2011. http://handbook.cochrane.org/. Accessed September 8, 2013. 39. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21:1539–1558. 40. Deshmukh JS, Motghare DD, Zodpey SP, Wadhva SK. Low birth weight and associated maternal factors in an urban area. Indian Pediatr. 1998;35:33–36. 41. England LJ, Levine RJ, Mills JL, Klebanoff MA, Yu KF, Cnattingius S. Adverse pregnancy outcomes in snuff users. Am J Obstet Gynecol. 2003;189:939–943. 42. Gupta PC, Sreevidya S. Smokeless tobacco use, birth weight, and gestational age: population based, prospective cohort study of 1217 women in Mumbai, India. BMJ. 2004;328:1538. 43. Hosain GM, Chatterjee N, Begum A, Saha SC. Factors associated with low birthweight in rural Bangladesh. J Trop Pediatr. 2006;52:87–91. 44. Krishna K. Tobacco chewing in pregnancy. Br J Obstet Gynaecol. 1978;85:726–728. 45. Pratinidhi A, Gandham S, Shrotri A, Patil A, Pardeshi S. Use of ‘Mishri’ a smokeless form of tobacco during pregnancy and its perinatal outcome. Indian J Community Med. 2010;35:14–18. 46. Steyn K, de Wet T, Saloojee Y, Nel H, Yach D. The influence of maternal cigarette smoking, snuff use and passive smoking on pregnancy outcomes: the birth to ten study. Paediatr Perinat Epidemiol. 2006;20:90–99. 47. England LJ, Kim SY, Shapiro-Mendoza CK, et al. Effects of maternal smokeless tobacco use on selected pregnancy outcomes in Alaska Native women: a case-control study. Acta Obstet Gynecol Scand. 2013;92:648–655. 48. Mehta AC, Shukla S. Tobacco and pregnancy. J. Obstet. Gynaecol. India 1990;40:156–160. 49. Baba S, Wikström A-K, Stephansson O, Cnattingius S. Influence of smoking and snuff cessation on risk of preterm birth. Eur J Epidemiol. 2012;27:297–304. 50. Baba S, Wikström A-K, Stephansson O, Cnattingius S. Changes in snuff and smoking habits in Swedish pregnant women and risk for small for gestational age births. BJOG. 2013;120:456–462. 51. Baba S, Wikström A-K, Stephansson O, Cnattingius S. Influence of snuff and smoking habits in early pregnancy on risks for stillbirth and early neonatal mortality. Nicotine Tob Res. 2014;16:78–83. 52. Wikström A-K, Cnattingius S, Stephansson O. Maternal use of Swedish snuff (snus) and risk of stillbirth. Epidemiology. 2010;21:772–778. 53. Wikström A-K, Cnattingius S, Galanti MR, Kieler H, Stephansson O. Effect of Swedish snuff (snus) on preterm birth. BJOG. 2010;117:1005–1010. 54. Gupta PC, Sreevidya S. Smokeless tobacco use and risk of stillbirth: a cohort study in Mumbai, India. Epidemiology. 2006;17:47–51.

Downloaded from http://ntr.oxfordjournals.org/ at North Dakota State University on June 1, 2015

8. Kramer MS. The epidemiology of adverse pregnancy outcomes: an overview. J Nutr. 2003;133:1592S–1596S. 9. de Onis M, Onyango AW, Borghi E, Garza C, Yang H. Comparison of the World Health Organization (WHO) Child Growth Standards and the National Center for Health Statistics/WHO international growth reference: implications for child health programmes. Public Health Nutr. 2006;9:942–947. 10. Gupta B, Kumar N. A cross-country comparison of knowledge, attitudes and practices about tobacco use: findings from the global adult tobacco survey. Asian Pac J Cancer Prev. 2014;15:5035–5042. 11. Jha P, Ranson MK, Nguyen SN, Yach D. Estimates of global and regional smoking prevalence in 1995, by age and sex. Am J Public Health. 2002;92:1002–1006. 12. Lopez AD, Collishaw NE, Piha T. A descriptive model of the cigarette epidemic in developed countries. Tob Control. 1994;3:242. 13. Digard H, Proctor C, Kulasekaran A, Malmqvist U, Richter A. Determination of nicotine absorption from multiple tobacco products and nicotine gum. Nicotine Tob Res. 2013;15:255–261. 14. Pastrakuljic A, Schwartz R, Simone C, Derewlany LO, Knie B, Koren G. Transplacental transfer and biotransformation studies of nicotine in the human placental cotyledon perfused in vitro. Life Sci. 1998;63:2333–2342. 15. Hukkanen J, Jacob P III, Benowitz NL. Metabolism and disposition kinetics of nicotine. Pharmacol Rev. 2005;57:79–115. 16. Repo JK, Pesonen M, Mannelli C, Vähäkangas K, Loikkanen J. Exposure to ethanol and nicotine induces stress responses in human placental BeWo cells. Toxicol Lett. 2014;224:264–271. 17. Feltes BC, Poloni J de F, Notari DL, Bonatto D. Toxicological effects of the different substances in tobacco smoke on human embryonic development by a systems chemo-biology approach. PLoS ONE 2013;8:1–19. doi:10.1371/journal.pone.0061743. 18. Rani M, Bonu S, Jha P, Nguyen SN, Jamjoum L. Tobacco use in India: prevalence and predictors of smoking and chewing in a national cross sectional household survey. Tob Control. 2003;12:e4. 19. Warren CW, Riley L, Asma S, et al. Tobacco use by youth: a surveillance report from the Global Youth Tobacco Survey project. Bull World Health Organ. 2000;78:868–876. 20. Furberg H, Lichtenstein P, Pedersen NL, Bulik C, Sullivan PF. Cigarettes and oral snuff use in Sweden: prevalence and transitions. Addiction. 2006;101:1509–1515. 21. Mazurek JM, Syamlal G, King BA, Castellan RM. Smokeless tobacco use among working adults—United States, 2005 and 2010. MMWR Morb Mortal Wkly Rep. 2014;63:477–482. 22. Peltzer K. Smokeless tobacco and cigarette use among black secondary school students in South Africa. Subst Use Misuse. 2003;38:1003–1016. 23. World Health Organization. WHO Report on the Global Tobacco Epidemic, 2009  : implementing smoke-free environments. Rapport de l’OMS sur l’épidémie mondiale de tabagisme, 2009 : mise en place d’espaces non-fumeurs 2009. 2009. http://apps.who.int//iris/handle/10665/44229. Accessed November 3, 2014. 24. Gupta PC, Warnakulasuriya S. Global epidemiology of areca nut usage. Addict Biol. 2002;7:77–83. 25. Rogers JM. Tobacco and pregnancy. Reprod Toxicol. 2009;28:152–160. 26. Gunnerbeck A, Wikström AK, Bonamy AKE, Wickström R, Cnattingius S. Relationship of maternal snuff use and cigarette smoking with neonatal apnea. Pediatrics. 2011;128:503–509. 27. Hogan MC, Foreman KJ, Naghavi M, et  al. Maternal mortality for 181 countries, 1980–2008: a systematic analysis of progress towards Millennium Development Goal 5. Lancet. 2010;375:1609–1623. 28. Rajaratnam JK, Marcus JR, Flaxman AD, et al. Neonatal, postneonatal, childhood, and under-5 mortality for 187 countries, 1970-2010: a systematic analysis of progress towards Millennium Development Goal 4. Lancet. 2010;375:1988–2008. 29. Gray N, Henningfield JE, Benowitz NL, et  al. Toward a comprehensive long term nicotine policy. Tob Control. 2005;14:161–165. 30. Rogers JM. Tobacco and pregnancy: overview of exposures and effects. Birth Defects Res C Embryo Today. 2008;84:1–15.

Nicotine & Tobacco Research, 2014, Vol. 00, No. 00

Nicotine & Tobacco Research, 2014, Vol. 00, No. 00 55. Krishnamurthy S, Joshi S. Gender differences and low birth weight with maternal smokeless tobacco use in pregnancy. J Trop Pediatr. 1993;39:253–254. 56. Bartal M. Health effects of tobacco use and exposure. Monaldi Arch Chest Dis. 2001;56:545–554. 57. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med. 2009;151:264–269.

9 58. Deeks JJ, Dinnes J, D’Amico R, et al. Evaluating non-randomised intervention studies. Health Technol Assess. 2003;7:1–179. 59. OCEBM Levels of Evidence Working Group. The Oxford 2011 levels of evidence. 2011. http://www.cebm.net/index.aspx?o=5653. Accessed September 8, 2013. 60. Stepanov I, Jensen J, Hatsukami D, Hecht SS. New and traditional smokeless tobacco: comparison of toxicant and carcinogen levels. Nicotine Tob Res. 2008;10:1773–1782.

Downloaded from http://ntr.oxfordjournals.org/ at North Dakota State University on June 1, 2015