This paper reports the concentration of NM in a range of biological samples from prospectively and conveniently enrolled central Australian Aboriginal pregnant women at hospital for the birth of their baby; mothers self-reported their tobacco use as being chewers, smokers, or no-tobacco use. We also considered NM concentrations in relation to the maternal, birthing, placental and neonatal characteristics and outcomes previously described for these families within the overarching research [54, 55].
The population consisted of 58% self-reported tobacco users (SLT 26%, smokers 32%), however in the self-reported no-tobacco group, several high maternal NM concentrations were indicative of tobacco exposure. Re-categorisation of participants based on NM concentrations would have enabled a more robust statistical analysis including correlation and regression, however this re-categorisation was not conducted for the following reasons. Firstly, urinary cotinine levels (i.e., nicotine metabolism) is commonly used as a cut-off value to indicate tobacco use in pregnancy, yet the literature reports a wide range in cut-off values for example 250 ng/mL , 82 ng/mL , and 42 ng/mL . Secondly, nicotine metabolism is highly individual and is influenced by previous exposure, recency of exposure prior to collection, tobacco formulation, diet and body weight. Importantly, evidence shows that following equivalent tobacco exposure, ethnically different populations have varied nicotine intake and cotinine clearance [73, 74] and specific genomic variations have been identified and contribute to the biological variation in NM findings in different populations . As yet, genome research related to the metabolism of nicotine in pregnant Australian Aboriginal populations has not been conducted. Furthermore, there were dissimilarities in study designs and methods between this research and other studies including differences in laboratory techniques and analysis equipment. Given these issues it was considered premature to extensively compare nicotine absorption and clearance concentrations, and maternal and neonatal outcomes in this research cohort with wider findings. Lastly, the Australian maternity data collection is based on self-report, by retaining the self-reported categorisation, it is possible to compare the research findings with the National Perinatal Data Collection.
Maternal nicotine exposure and outcomes
In Australia, the National Perinatal Data Collection reports on mothers’ use of cigarettes twice during pregnancy and there is no opportunity to report SLT or nicotine containing products such as e-cigarettes, nicotine gum, patches, mists, tooth powder or lozenges. Accordingly, the pituri users’ outcomes in this research will have been recorded in the National Perinatal Data Collection as “no-tobacco use”, thus, incorrectly reporting the nicotine exposure outcomes for these Aboriginal and Torres Strait Islander mothers and babies at the local, territory, national and international level. Modifying the pregnancy health assessment and data collection tools to record the broader range and use of nicotine containing products will enable a more inclusive and discriminative assessment of their effects on contemporary Australian pregnancies and outcomes.
The Perinatal data reports the rate of smoking in pregnancy by Indigenous women across Australia is 44% , and the reported rate of smoking in the first 20 weeks of gestation is 52% across the NT. The reported rate of smoking for Indigenous women in the Alice Springs area is 30%  and comparable to the finding in this research of 32%. In this research, which captured SLT use as well as cigarette use, there were indications of higher mean concentrations of NM in the chewer group’s maternal blood, umbilical cord blood, amniotic fluid, neonatal urine Day 1 and breast milk in comparison with the smoker and no-tobacco use groups. Higher concentrations of NM in both tobacco user groups were aligned with several clinically important maternal differences, specifically as follows, in the rate of elevated glucose, hypertension and anaemia.
In general populations, smoking has been identified as a risk factor for the development of diabetes [76, 77], and while the development of impaired glucose regulation is a normal physiological change in pregnancy , both pre-pregnancy and maternal smoking increase the risk of gestational diabetes [26–28]. Globally, the rate of diabetes in pregnancy is higher in Indigenous populations  and there is also a higher use of tobacco in these populations . At the Australian level, the incidence of gestation diabetes has increased from 5.2% in 2000–2001, to 9.3% in 2012–2013, to 15.1% in 2016–2017 . In this research, the high maternal blood NM levels in the tobacco-exposed groups with elevated glucose, and the lower NM levels in the tobacco-exposed groups without elevated glucose suggests the involvement of nicotine in insulin resistance or glucose metabolism as opposed to a mechanism related to only the combustion of tobacco. The presence of elevated glucose impacts the mother during pregnancy as well as her longer-term health, with approximately 50% of affected mothers going on to develop diabetes within 10 years of first diagnosis of gestational diabetes .
While nicotine is a potent vasoconstrictor , counterintuitively smoking in pregnancy results in a dose-dependent decrease in hypertension [20, 82–84]. Our data for smokers supports this assertion, with fewer cases of elevated hypertension in the smokers than no-tobacco users. There is evidence of hypertension in smoking participants with lower mean maternal blood NM concentrations compared with an absence of hypertension in those with higher NM levels. Chewing tobacco was expected to follow the same trend  (i.e., higher NM and an absence of hypertension) and this was evidenced with only one chewer having a record of hypertension. This one participant had a very high NM (291 ng/mL). In the no-tobacco group, with the removal of the one high NM result (1190 ng/mL) from a participant with hypertension, the mean NM result was 42 ng/mL for those with hypertension, and 53 ng/mL in those without hypertension.
Maternal anaemia occurred less often in participants with higher mean NM concentrations compared with those with lower mean concentrations. These findings are contrary to the literature which shows that smoking and SLT use in pregnancy is associated with maternal anaemia [22, 85–87], however there may be endemic confounders in this population that explain this finding. In the geographical region, Hymenolepis nana (intestinal worms) are endemic, and anaemia is associated with approximately 18% of infections . Nicotine exposure is a treatment for intestinal worms in humans and animals [89–91], and in this population, we hypothesise that exposure to nicotine from pituri chewing and smoking may decrease the worm burden in a dose-dependent manner and thereby decrease the resultant anaemia from this parasite.
Neonatal nicotine exposure and outcomes
Nicotine readily transits from the maternal circulation to the placenta and much has been written around the neonatal outcomes of in-utero exposure during pregnancy, birth and early infancy , however longer-term outcomes are being demonstrated including that in-utero nicotine exposure permanently impacts the foetal pancreas and results in a loss of beta cell mass, leading to a life-long increased risk of impaired glucose and insulin homeostasis, childhood and adult obesegenesis [93–96], and childhood  adult hypertension  and type 2 diabetes .
In this research, evidence of neonatal exposure to nicotine was demonstrated with NM concentrations in arterial and venous cord blood and amniotic fluid in those exposed to tobacco through smoking or chewing and notably, these concentrations were significantly higher than the maternal blood concentrations. The unlabelled cord blood values are comprised of a mixture of venous and arterial samples, so while it may appear that the arterial cord blood NM concentrations may be higher than venous cord blood from chewers (Table 2), these are based on only three participants for whom both labelled samples were available, and the relationship is reversed for the eight smokers for whom there were both cord blood samples.
As observed in this research through NM concentrations measured in urine, the neonate and the mother are able to excrete nicotine via the kidney [18, 100] and Day 1 NM urine concentration is a useful biomarker for nicotine exposure. Following birth and the discontinuation of the supply of nicotine via the placenta, neonatal urine from chewers and smokers declined in NM concentration from Day 1 to Day 3. Whilst this finding provides some reassurance of the neonatal ability to excrete the NM resulting from in-utero exposure, the mean concentrations in both the neonates of chewers and smokers did not equate to that of the neonates from the no-tobacco use group by Day 3. However, we were not able to take account of hydration status with these small volumes, so no inferences can be made.
There is potential for continued neonatal nicotine exposure following birth through breast milk . NM concentration in breast milk from smokers was low and similar to the no-tobacco group (11 mg/mL) while breast milk of chewers was higher (54 ng/mL). Pituri can be used discretely, and continuously throughout the hospital stay which may account for these higher levels. It may also be associated with slower clearance from breast milk in chewers than smokers, as breast milk is clear of nicotine in smokers after four hours of abstinence, but is still present in snus (SLT) users’ after abstaining from tobacco use for 11 hours .
We anticipated higher rates of SCN admission for the smoker group than the no-tobacco group , but in fact the opposite was true, with less smokers compared to no-tobacco neonates treated in SCN. Pituri chewers were most likely to be admitted and this was associated with high mean maternal blood NM, indeed, all mothers whose neonates were admitted to SCN had higher blood NM than mothers of non-admitted neonates.
Literature from smoked tobacco research indicates a reduction in male newborns in the presence of maternal and/or paternal cigarette smoking in a nicotine dose-dependent manner [103–105], and our smokers produced 43% male births, down from the worldwide average of 51.4% . There is emerging evidence that changes to DNA methylation as a result of maternal smoking are greater in male offspring than female offspring . However, while smoking would be expected to be associated with a decrease in birthweight, particularly in male offspring , we found male neonates from smokers were 500 g heavier than those from the no-tobacco group while females were on average 280 g lighter. The tobacco chewer group, despite exhibiting higher mean NM levels in their blood, had 60% male births and less impact on male birthweight.
Maternal smoking and SLT use increases the risk of a small for gestation age (SGA) neonate [108, 109]. Likewise, elevated maternal glucose increases the likelihood of an earlier (spontaneous or induced) birth, and increases the likelihood of an SGA or a large for gestation age (LGA) neonate dependent upon neonatal genome [110, 111]. In Australian Aboriginal pregnancies, the immediate foetal impact of exposure to pre-gestational and gestational diabetes is different to that evidenced in Australian non-Aboriginal pregnancies. In the presence of pre-gestational diabetes, there is a slightly higher incidence of an LGA birth in Aboriginal pregnancies compared with non-Aboriginal pregnancies (32.9% versus 32.7%) and an increase in SGA births (8.2% versus 4.6%). In the presence of gestational diabetes there is an increased incidence of an LGA neonate in Aboriginal pregnancies compared with non-Aboriginal pregnancies (21.1% versus 13.3%) and the reverse for the incidence of an SGA birth in Aboriginal pregnancies compared with non-Aboriginal pregnancies (7.1% versus 8.3%) [110, 111].
In this research, the medical record of participants often recorded both pre-gestational and gestational diabetes, and accordingly, this variable was dealt with as one variable. The research findings show that whilst the mean birthweight for each group across the cohort was similar, lower birthweights were seen in the presence of tobacco exposure when there was an absence of maternal elevated glucose. Tobacco-exposed women with elevated blood glucose had higher maternal blood NM concentrations than their group counterparts without elevated glucose, and had analogous higher birthweight neonates. In the chewers, neonates exposed to elevated glucose and higher maternal NM weighed 1014 g more than the neonates of chewers not exposed to elevated glucose and lower maternal NM. This same trend existed for smokers only if a very small (900 g) pre-term 27-week gestation neonate was excluded from the elevated glucose group for analysis, in which case neonates from smokers exposed to elevated glucose weighed 472 g more than those not exposed.
We grouped participants by their response to the interview question “have you smoked cigarettes or chewed tobacco in this pregnancy?” Surprisingly, participants with visible oral pituri quids answered “no” to this question. It became apparent that chewers did not consider pituri as tobacco. This finding was later confirmed in ethnobotanical interviews with senior Aboriginal women who rejected the notion that pituri was a tobacco plant . The interview question was subsequently changed to “have you chewed pituri or mingkulpa or tobacco in this pregnancy?” Nevertheless, early participant enrolments may have been incorrectly categorised, and this may explain some high NM readings in the no-tobacco use group.
Biological samples were taken at a single time point, so the NM values do not accurately reflect the totality of tobacco and nicotine exposure during the pregnancy, and instead indicate the extent of recent exposure. This is particularly relevant when comparing smokers and chewers, where the continued use of pituri is possible within a hospital, as opposed to the need for a smoker to leave the hospital grounds for a cigarette.
The research conveniently enrolled participants who were considered to be over 28 weeks gestation based on the expected due date. Mothers who experienced early- and mid-pregnancy adverse outcomes or birth or who were transferred to a tertiary health service prior to that time were therefore not invited to participate and this may underestimate the impact of nicotine exposure on pregnancy outcomes.
Finally, the small numbers for each biological sample limits the ability to perform statistical comparisons or draw generalised conclusions. Furthermore, there are a variety of Australian Nicotiana spp. plants which grow across the expanse of Australia with the nicotine content of the individual plants impacted by their location and environmental factors such as soil, rain and temperature and their preparation for use as SLT [6, 9]. Therefore, extrapolation of the findings from this research with other Australian Aboriginal and Torres Strait Islander populations, and other Indigenous worldwide groups and non-Indigenous groups, may not be appropriate.