Introduction

The “natural” sex ratio at birth (SRB) in most parts of the world is approximately 105 boys per 100 girls and usually falls between 104 and 107 with minor deviations due to infections, parental nutrition, parental age, environmental factors, hormonal changes, and other factors [1–4]. Any deviation from this natural SRB signifies some degree of sex selection, which is often influenced due to certain sex preferences [2].

Skewed SRB has been observed in several countries, particularly in China and India, with a disproportionately high number of “missing” female births, attributed to infanticide and neglect of female children [1,2]. In recent decades, sex-selective abortion has become a potential mechanism for gender-biased sex selection due to son preference, declining fertility, and easily available sex-determination technologies such as ultrasound [1,5]. Gender-biased sex selection, however, also depends on access to abortion after the first trimester, when ultrasound can first be used to detect the sex of a fetus. Previous studies have revealed the impact of ultrasound technology on sex-selective abortion in some Asian and Eastern European countries [1,2,5].

The increasing availability of non-invasive prenatal testing (NIPT), which can determine fetal sex as early as 10 weeks, has raised some ethical concerns regarding its potential misuse for sex-selective abortion. Unlike ultrasound, which detects fetal sex later in pregnancy, NIPT allows for earlier opportunities for sex selection, particularly in settings where abortion laws do not require a reason for termination. While there is no direct evidence linking NIPT to increased sex-selective abortion [6], concerns remain that its accessibility could contribute to such practices, particularly among communities with a strong son preference. However, studies indicate an increase in male-biased SRB among Asian migrants, in high-income countries such as the US [7,8], Canada [9,10], UK [11], and Australia [12,13]. Restricting the disclosure of fetal sex in NIPT results has been proposed as a feasible approach to mitigating sex-selective abortion [14]. Given projections of growing sex imbalances at birth due to pre-natal sex-selection practices [2]. It is crucial to examine how these practices may influence SRB trends, particularly in multicultural settings like Australia, where immigrants from countries with historically skewed SRB may maintain son preference and practice prenatal sex selection in their new countries of settlement.

Australia is a multicultural country with a sizeable immigrant population from Asia and other regions with a strong culture of son preference. It is plausible that such cultural preference combined with accessibility to NIPT services may explain the unbalanced SRB. Notably, in Australia, only two previous studies have examined the SRB according to the mother’s country of birth [12,13], however, neither study specifically investigated sex-selective abortion practices taking into account the sex composition of older siblings and previous abortion. This study aimed to empirically identify the indirect evidence of sex-selective abortion practices for a large Australian cohort.

Materials and methods

Results

Cohort characteristics

There were 2,175,252 births in the analytic cohort during the study period, of which 81.7% (1,778,561) were from Australian-born mothers. The youngest maternal age group (15–24 years) accounted for approximately a quarter of births to mothers from Australia and Lebanon, while advanced maternal age (>35 years) accounted for 28% of births to mothers from China and the Philippines. Parity varied considerably between maternal countries of birth. The lowest proportion of higher order parity (2 or higher) was observed in births to mothers from India (8.6%) and China (8.8%), and highest for births to mothers born in Lebanon (47%) (Table 1).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Maternal characteristics of mothers giving birth in Australia (WA, NSW) by mother’s country of birth, 1994–2015.

https://doi.org/10.1371/journal.pgph.0004672.t001

Overall, parity-specific and maternal country-of-birth-specific sex ratios

The overall sex ratio at birth was 1.058, 95% CI 1.055–1.061. The parity-specific SRB ranged between 1.06, 95% CI 1.06–1.07 for parity 0 births to 1.05, 95% CI 1.04–1.06 for parity 3⁺ births. The SRB for births to Australian, UK, and New Zealand–born mothers was within the range expected for the natural sex ratio (1.05–1.07). However, the SRB exceeded the natural ratio for births to Indian (1.07, 95% CI 1.05–1.09), Chinese (1.08, 95% CI 1.07–1.10) and Vietnamese–born (1.07, 95% CI 1.05–1.09) mothers (Fig 1).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Sex ratio at birth (SRB) in Australia (WA, NSW) by mother’s country of birth stratified by parity, 1994–2015.

The vertical dashed line at 1.06 indicates the reference SRB, and the shaded area around it represents the 95% CI for overall SRB for Australia based on our data.

https://doi.org/10.1371/journal.pgph.0004672.g001

Maternal country of birth specific sex ratios by sex of the previous birth(s)

The SRB at second birth to mothers from China was 1.09, 95% CI 1.06–1.13. This ratio was elevated to 1.11, 95% CI 1.06–1.16 when the previous birth was female as compared to male (1.08, 95% CI 1.03–1.12). The SRB was markedly higher at the third birth when all previous births were females (1.34, 95% CI 1.14–1.57) than the SRB for all third births (1.11, 95% CI 1.02–1.23). The SRB at second birth to mothers from India was 1.10, 95% CI 1.06–1.15, which was elevated when the previous birth was a female (1.14, 95% CI 1.08–1.21). At the third birth, the odds of male birth were 31% higher than female birth (SRB 1.31, 95% CI 1.05–1.63) for mothers from India when all previous births were female. Similarly, the SRB at second birth when the first birth was female ranged from 1.16, 95% CI 1.08–1.24 for births to mothers from Lebanon to 1.11, 95% CI 1.04–1.20 for births to mothers from the Philippines. In contrast to mothers from China, India and Lebanon, SRBs at second and third birth to mothers from Australia, New Zealand, and the UK did not vary by sex of the previous birth (Figs 2 and 3, S1 Table).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Sex ratio at birth (SRB) of first order and second-order births in Australia (WA, NSW), conditional on the sex of the previous sibling for mothers with their first two consecutive births by mother’s country of birth, 1994–2015.

The vertical dashed line at 1.06 indicates the reference SRB, and the shaded area around it represents the 95% CI for overall SRB for Australia based on our data.

https://doi.org/10.1371/journal.pgph.0004672.g002

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Sex ratio at birth (SRB) of third-order births in Australia (WA, NSW), conditional on the sex of the previous sibling for mothers with their first three consecutive births by mother’s country of birth, 1994–2015.

The vertical dashed line at 1.06 indicates the reference SRB, and the shaded area around it represents the 95% CI for overall SRB for Australia based on our data.

https://doi.org/10.1371/journal.pgph.0004672.g003

Discussion

Our results indicate that the SRB for Australian births to women of Asian origin, particularly from China, and India is higher than the natural ratio, especially at higher parities. We revealed male–biased SRB particularly in second and third–order births from Indian and Chinese mothers with an earlier daughter or two earlier daughters that was not explained by the stopping rule. Additionally, we observed substantially more induced abortions among Indian and Chinese-born mothers, particularly at earlier gestations, which coincided with the introduction of NIPT. This study provides the most compelling observational evidence to date of male-biased SRB among overseas-born mothers, which appears to be attributed to prenatal sex determination followed by selective abortion of females.

The study findings indicate an imbalanced SRB consistent with previous research from western countries where mothers were born overseas [7,10,24–27], including two from Australia [12,13], particularly at higher–order births.. While few studies have examined the SRB modification by previous sex [24–26], only two studies accounted for both sex of previous births and previous induced abortions [9,10]. Our findings of an increased SRB when the previous births were females, as compared to male or mixed sex, align with findings of previous studies among women of Indian and Chinese origin [10,25,26,28]. Although preference for mixed–sex composition — should such a preference exist — is typical in western societies such as Australia, data on sex preference are not always unequivocal with some studies suggesting a preference for daughters [29]. Our sensitivity analyses support that the SRB is highly influenced by previous sex, particularly for births from Indian, Chinese, Vietnam and Lebanon–born mothers. Our findings of an increased SRB at parity two and male: female PSR for births from China, Lebanon, and Philippines–born mothers, suggest a preference for male children while limiting family size. Furthermore, a higher SRB for first births to Chinese and Vietnam–born mothers also indicates possible sex selection from the first birth. Although our findings suggest a consistent trend of male-biased SRB, we note that wider confidence intervals at higher order births (parity 3+) for some countries (e.g., China, India, and Vietnam) reflect smaller sample sizes and greater variability in fertility behaviors such as stopping rules. Despite this, the overall trend of male-biased SRB remains consistent and aligns with existing literature on sex ratio imbalances among immigrant groups, particularly from countries with a strong son preference. These findings should be interpreted with caution, particularly for higher birth orders where statistical uncertainty is greater. However, most estimates remain outside the expected natural SRB range, reinforcing the validity of our results.

Prenatal sex selection in Asian and Eastern European countries can be attributed to three major factors: son preference (demand factor); the effect of decreased fertility on the risk that families may end up not having a boy thereby encouraging them to resort to abortions (the squeeze factor); and access to prenatal sex selection technology and abortion services (supply factor) [5]. The demand factor posits that certain families deliberately want or avoid having daughters or sons, which can mainly be explained in terms of current societal norms [10]. In high–income countries such as Australia with strict first–trimester abortion laws and high access to contraception [30], sex selection can occur through a differential stopping rule or sex selection via sex–selective termination of pregnancy. Our data suggest that the male–biased SRB for births to Chinese–born mothers can be partially attributed to the stopping rule, unlike for births from Indian, and other Southeast Asian–born mothers. A recent report also shows that, 97% of induced abortions in WA in the year 2009–2015 performed at <13 weeks of gestation were for reasons other than fetal abnormalities potentially including sex selection [31].

The use of sex–determining technologies, such as ultrasound has led to an increase in sex–selective abortion in Asian countries [5], and it is feasible that the ease, timing, and accuracy of NIPT could further facilitate this practice [6]. Unlike ultrasound, NIPT for determining sex is available much earlier in pregnancy, at 10 weeks of gestation, giving families ample time to seek termination services (considering the legal gestational age for termination of pregnancy in each jurisdiction). In Australia, where abortion laws do not require parents to disclose their reason for termination, restricting the disclosure of fetal sex through NIPT has been suggested as one of the most effective ways to prevent sex-selective abortion [14].

Indeed, our study shows an increase in SRB in the period of 2010–2015 and increased induced abortions at earlier gestational ages for Indian and Chinese–born mothers, around the same period that NIPT become widely available [32]. It would be beneficial to have direct evidence of sex determination through NIPT (which became available in Australia in 2012) on the incidence of induced abortion. However, we are not aware of any data or published reports on the sex ratio of aborted fetuses in a large representative cohort in Australia. Our findings suggest that male–biased SRB is more pronounced for Indian and Chinese–born mothers with prior induced abortion than their Australian counterparts indicating the impact of prenatal sex determination, including NIPT on sex–selective abortion.

Other than sex–selective abortion, neglect and infanticide of females [28], and differential under-reporting of female births are common reasons for the imbalanced SRB in Asian countries [11,28]. These factors are implausible in the context of Australia. The imbalanced SRB may be partially explained by factors including infections, maternal nutrition, and hormonal influences [2–4]. However, it is unlikely to explain the discrepancies indicated in our conditional sex ratios (conditioned on parity or previous sex). In Australia, the most likely explanation for the imbalanced SRB observed among overseas–born mothers, in our study – particularly those from India, and China is sex–selective abortion.

Abortion laws in Australia are generally decriminalized with varying restrictions on gestation and consent requirements (S2 Fig). For example, in WA, abortion was permitted up to 20 weeks, requiring approval from two doctors beyond this limit. As of March 2024, the new WA abortion reform reduced this requirement to one doctor’s approval and gestational limit of 23 weeks. However, abortion after 20 weeks cannot explain the male–biased SRB in our study, as these cases would have been recorded in the MNS data with their sex information and included in our SRB estimates. The lack of reliable data on how abortion laws are implemented across jurisdictions makes it unclear whether sex-selective procedures occur within the state, interstate, or overseas. However, if sex-selective abortion does occur within the state of residence, it is more likely to happen earlier in pregnancy when such procedures are more accessible and less restrictive legal environments. While our study examines population-level SRB, individual instances of sex-selection may still occur without distorting overall result. Pre-pregnancy selection via assisted reproduction, though possible, the overall influence on population-level SRB is relatively small.

NIPT is widely available for non–medical sex determination in many countries [6,33]. But in countries such as the UK, providers are discouraged from offering it for non–medical reasons [34]. In Australia, the availability of NIPT for sex determination is widely supported [35], but guidelines proscribed sex selection for all except medical reasons, although future changes to this guideline are possible. With NIPT becoming more accessible and affordable, there may be an increased pressure to relax current guidelines, potentially leading to more sex–selective abortions and “missing” female births. Consequential demographic imbalances can lead to marital squeeze [11], reinforce gender essentialism and stereotypes, and contribute to broader societal harms, including increased violence, restrictions on women's reproductive autonomy, and negative mental health impacts [1]. These imbalances also undermine gender equality by reinforcing discriminatory norms and limiting women's rights. This underscores the importance of ongoing SRB monitoring, enforcement of existing guidelines, and public awareness campaigns on the consequences of sex selection.

Conclusions

Our study has implications for clinical and public health practice in Australia and possibly other high–income countries. Our findings suggest patterns consistent with male biased SRB among certain immigrant groups in Australia, particularly from countries with a strong son preference, such as India and China. While this may be indicative of prenatal sex-selection, our study does not establish causality. The availability of advanced sex–determination technologies such as NIPT has the potential to facilitate sex–selective termination of pregnancy and gives a context from which SRB monitoring needs to be further considered. To address this issue, there needs to be strict enforcement of existing legislation, culturally appropriate discussions about reproductive options, and public awareness campaigns about the consequences of sex imbalance. Furthermore, researchers are urged to try to understand how reproductive decision–making occurs in certain cultures, particularly in families from South Asian backgrounds, in a culturally acceptable manner. Finally, we strongly recommend that sex–determination technologies like NIPT should not be used to reveal non–medical traits of the fetus, including sex as restricting this information is one of the most feasible approaches to preventing sex-selective abortion and reducing gender-biased reproductive practices.

Supporting information

Acknowledgments

The authors would like to thank the Linkage, Data Outputs and Research Data Services Teams at the Western Australian Data Linkage Branch, and the Centre for Health Research Linkage (CHeReL), NSW. We also thank the Data custodians for the Midwives Notification System, the Western Australian Registry of Births, Deaths and Marriages, Hospital Morbidity Data Collection, and NSW Perinatal Data Collection for providing data for this project.