Impact of Noise Exposure on Risk of Developing Stress-Related Obstetric Health Effects: A Systematic Review and Meta-Analysis

Abstract

Background:

Exposure to noise can increase biological stress reactions and that could increase the risk of stress-related prenatal effects, including adverse obstetric outcomes; however, the association between exposure to noise and adverse obstetric outcomes has not been extensively explored. The objective of this review was to evaluate the evidence between noise exposures and adverse obstetric outcomes, specifically preeclampsia, gestational diabetes, and gestational hypertension.

Materials and Methods:

A systematic review of English language, comparative studies available in PubMed, Cochrane Central, EMBASE, and CINAHL databases between January 1, 1980 and December 29, 2021 was performed. Risk of bias for individual studies was assessed using the Risk of Bias Instrument for Nonrandomized Studies of Exposures, and the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach was used to assess the certainty of the body of evidence for each outcome.

Results:

Six studies reporting on preeclampsia, gestational diabetes, and gestational hypertension were identified. Although some studies suggested there may be signals of increased responses to increased noise exposure for preeclampsia and gestational hypertension, the certainty in the evidence of an effect of increased noise on all the outcomes was very low due to concerns with risk of bias, inconsistency across studies, and imprecision in the effect estimates.

Conclusions:

While the certainty of the evidence for noise exposure and adverse obstetric outcomes was very low, the findings from this review may be useful for directing further research in this area, as there is currently limited evidence available. These findings may also be useful for informing guidelines and policies involving noise exposure situations or environments.

INTRODUCTION

Exposure to noise has been linked to several adverse health effects, including adverse birth outcomes. In particular, studies have explored outcomes including preterm birth, low birth weight, being small for gestational age, and congenital malformations.[1,2,3] It is recognized that exposure to noise can increase biological reactions related to stress[4] and that could increase the risk of stress-related prenatal effects, including adverse obstetric outcomes.

It has been proposed that noise exposure from a variety of sources, such as occupational and traffic sources, during pregnancy could act as a general stressor on the mother.[5] As a stressor, noise has been hypothesized to activate the sympathetic nervous system and cause several physiological and psychological changes that may affect pregnancy, such as increases in stress hormones and triggering of inflammatory pathways.[2,6] This could lead to changes that promote adverse obstetric outcomes, such as gestational hypertension or diabetes, or preeclampsia.[6]

Previous reviews have examined the relationship between exposure to noise and birth outcomes. Hohmann et al.[1] explored exposure to chronic noise during pregnancy for childhood and birth outcomes, including birth weight and preterm birth, as well as pediatric outcomes, including blood pressure and stress hormones, and reported that there were no associations between chronic noise and pregnancy outcomes, while noting a high variation in study design, outcome, exposure, and confounder assessments. Similarly, Ristovska et al.[2] conducted a systematic review on the associations between reproductive outcomes, including birth weight and preterm birth, and occupational and environmental noise exposure.[2] This review reported that epidemiological studies related to environmental or occupational noise exposure have shown that very high noise exposure could be associated with low birth weight, but not with the other reproductive outcomes.[2] The authors recognized exposure assessment was a major limitation and recommended future research with objective exposure assessment and adjustment for confounding. Further, Dzhambov et al.[3] conducted meta-analyses exploring the effects of exposure to noise on pregnancy (including gestational hypertension), birth outcomes, and fetal development, and reported that women exposed to high noise levels during pregnancy were at an increased risk for gestational hypertension, but no significant association was found for preeclampsia. These reviews, however, focused mainly on adverse birth outcomes and the health effects on the children, rather than adverse obstetric outcomes of the mothers themselves. Furthermore, these reviews recognized that the current evidence had limitations due to concerns with risk of bias and variation in study designs and recommended further research studies.

The objective of the current systematic review was to examine the association between noise exposure and the risk of stress-related obstetric effects, specifically preeclampsia, gestational diabetes, and gestational hypertension. This included examining the strength of the association using certainty of evidence, as well as updating the available evidence. This review is one in a series of systematic reviews reporting on the associations between noise levels and cardiovascular, metabolic, and obstetric health outcomes.[7,8]

METHODS

Details of the methods have previously been published elsewhere.[8] In brief, a systematic review and meta-analysis of exposure to incremental increases in noise on adverse obstetric outcomes was performed using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist for the development of this review [Table S1].[9] The protocol has been registered in PROSPERO (CRD42020209353).

Literature search

Peer-reviewed primary studies and systematic reviews were searched for in PubMed, MEDLINE, EMBASE, Cochrane CENTRAL, and CINAHL by an informational specialist for records between January 1, 1980 and December 29, 2021 [see Table S2 for search strategies]. Additional references not identified by the search strategy were searched using the reference lists of initially included systematic reviews. Results of the literature search and screening are presented in a PRISMA flow diagram.

Study screening and eligibility

As previously reported,[8] studies were eligible if they were published in English, conducted in humans, measured noise exposure using A-weighted noise metrics: dB(A), Lden, Lnight, LAeq16, LAeq8h, Ldn, Lmax, sound exposure level, and provided at least one comparison of noise levels in relation to preeclampsia (a serious blood pressure condition that occurs during pregnancy), gestational diabetes (a type of diabetes that develops during pregnancy), and gestational hypertension (high blood pressure during pregnancy) [Table 1].

Table 1.

Study eligibility

Population	Exposure	Comparator	Outcomes
General population, including persons participating in laboratory studies of noise exposure	Incremental increase in noise exposure as measured in A-weighted noise metrics	Noise exposure (i.e., sound pressure level) as measured in dBA. Noise exposure can come from aircraft, road traffic, rail, wind turbine, or industrial/occupational noise exposure	Preeclampsia, gestational diabetes, and gestational hypertension

Two reviewers independently reviewed titles and abstracts of eligible studies from the searches, as well as full texts of studies deemed eligible. Study screening was conducted using the screening software program Covidence (Covidence systematic review software, Veritas Health Innovation, Melbourne, Australia. Available at: www.covidence.org). Discrepancies that arose during screening were resolved by consensus or consultation with a third reviewer.

Data collection

A standardized and pilot-tested data extraction form [Table S3] was used to independently extract study data by two reviewers in Microsoft Excel. The details collected from the studies included: publication information, study design, study population, source and ascertainment of exposure, ascertainment of outcome, statistical analysis, study results for relevant outcomes, and funding information. All three outcomes were measured and reported in studies using relative risks (RRs) and odds ratios (ORs). Like screening, discrepancies were resolved by consensus or consultation with a third reviewer.

Data analysis

Data were pooled in a random-effects meta-analysis, where appropriate. Separate random-effects meta-analyses were conducted for each individual outcome. Studies were grouped based on the continuous measures of noise exposure reported (per 10 dBA Lden increases in noise). When study results were too heterogeneous to be pooled, findings were described narratively. Random-effects meta-analyses were performed using the metafor package in R (version 4.0.3). Between-study heterogeneity was assessed in several ways, including visual inspection of forest plots, using chi-square tests (using P < 0.1 as a threshold for clinical significance), and using I² statistics. Since no outcomes met the minimum of 10 studies, funnel plots could not be used to assess publication bias.

Risk of bias

Independently for each study, two reviewers assessed risk of bias using a preliminary version of the Risk of Bias Instrument for Nonrandomized Studies of Exposures (ROBINS-E) for nonrandomized (i.e., observational) studies [Table S4].[10] If conflicts arose between assessments, they were resolved either through consensus or by consulting a third reviewer.

For all studies where risk of bias was assessed using the ROBINS-E, the following confounders were considered as critical for adjustment: age and smoking status.

Grading of Recommendations Assessment, Development, and Evaluation evidence assessment

For each type of noise-exposure source and health outcome, the overall certainty of the evidence (CoE) was assessed. The CoE was assessed using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach by considering the following five domains for rating down: risk of bias, inconsistency, indirectness, imprecision, publication bias, and three domains for rating up: large or very large magnitude of effect, dose-response gradient, opposing residual confounding.[11] Within the GRADE approach, the body of evidence from nonrandomized studies was started at low CoE.

RESULTS

Literature Search

The search identified 11,482 records, of which six primary studies reporting on obstetric outcomes were included [Figure 1]. The effects of noise exposure on the following outcomes are presented: preeclampsia, gestational diabetes, and gestational hypertension [Table 2]. Estimates extracted from the studies (with transformations) are presented in Tables S5 to S7.

Figure 1.

PRISMA flow diagram. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Table 2.

Characteristics of included studies

	Studies % (n)
Total studies	6
Population
Neonates	0
Children (1–18 years)	0
Adults (18–64 years)	100 (6)
Seniors (≥65 years)	0
Noise source*
Aircraft	1
Road traffic	66.7 (4)
Rail	1
Industrial/occupational	33.3 (2)
Mixed	0
Laboratory-simulated	0
Prenatal outcomes*
Preeclampsia	33.3 (2)
Gestational diabetes	50.0 (3)
Gestational hypertension	66.7 (4)

^*May add up to >100% because some studies reported on more than one noise source/outcome.

Obstetric outcomes

Six studies reported on the following outcomes: preeclampsia, gestational diabetes, or gestational hypertension [Table 3].[12,13,14,15,16,17] Concerns with risk of bias due to exposure assessment, missing data, and measurement of outcomes were identified [Table 4]. As shown in Table 5, there was very low CoE for the effects of increased noise exposure on obstetric outcomes.

Table 3.

Characteristics of obstetric studies

Study	Country	Design	Population (N) − dataset/cohort (if applicable)	Sex (% female) and age (years)	Noise source	Setting	Exposure range and characterization	Outcome and method of ascertainment	Funding
Bendokiene et al., 2011[12]	Lithuania	Cross-sectional	Pregnant women(3121) − European Commission FP6 HiWATE project	100; NR	RO	H	(1) Low noise (≤50 dBA); n = 1782(2) Moderate noise (51–60 dBA); n = 1184(3) High noise (≥61 dBA); n = 155	Gestational hypertension: two or more of the physician’s blood pressure measurements met the criteria: a systolic blood pressure ≥ 140 mm Hg or a diastolic blood pressure ≥ 90 mm Hg, regardless of antihypertensive medication or controlling of blood pressure	NR
Lissåker et al., 2021[13]	Sweden	Cohort	Mother–child pairs from Swedish Medical Birth Register(1,109,516)	100; NR	O	W	(1) <70 dBA; n = 699,783(2) 70–74 dBA; n = 211,500(3) 75–79 dBA; n = 59,993(4) 80–85 dBA; n = 128,586(5) ≥85 dBA; n = 9654	Preeclampsia, gestational diabetes, and gestational hypertension: ICD-9 and ICD-10 codes reported in the Medical Birth Register	AFA Försäkring
Nurminen and Kurppa, 1989[14]	Finland	Cross-sectional	Mothers of infants with structural defects(1190) − Finnish Register of Congenital Malformations	100; NR	O	W	(1) <80 dBA; n = 95(2) ≥80 dBA; n = 945	Gestational hypertension: from both the antenatal record of the maternity health care center and the Register questionnaire	Finnish Work Environment Fund
Pedersen et al., 2017a[15]	Denmark	Cohort	Pregnant women(72,745) − Danish National Birth Cohort	100; NR	RO	H	(1) Low (<57.5 dB); n = 35,534(2) High (≥57.5 dB); n = 35,331	Preeclampsia and gestational hypertension per 10 dB increase: obtained diagnoses (ICD-10 codes) and dates of diagnosis from the Danish National Patient Register	The Danish Council for Independent Research
Pedersen et al., 2017b[16]	Denmark	Cohort	Pregnant women(72,745) − Danish National Birth Cohort	100; mean (SD): 29.4 (22.8–37.0)	RO	H	(1) 54.8 dBA (48.4–65.1); n = 18,187(2) 55.5 dBA (49.0–65.9); n = 18,186(3) 56.8 dBA (49.3–67.6); n = 18,188(4) 64.0 dBA (52.4–73.2); n = 18,184	Gestational diabetes per 10 dB increase: presence of GDM was determined by risk factors and fasting blood glucose levels collected by the DNBC	The Danish Council for Independent Research
Thacher et al., 2021[17]	Denmark	Cohort	Pregnant women(629,254)	100; median: 30.9	RO, RA, A	H	(1) Road traffic, Lden max 5 year; n = 617,828(2) Railway Lden max 5 year; n = 617,828(3) Aircraft, Lden: <45 dB; n = 15,288(4) Aircraft, Lden: 45–50 dB; n = 252(5) Aircraft, Lden: ≥50 dB; n = 433	ICD code	Independent Research Fund Denmark and the Health Effects Institute

A, air; AM, ambient; G, gun; GDM, gestational diabetes mellitus; H, home; L, laboratory; LS, laboratory-simulated; NR, not reported; O, occupational; RA, rail; RO, road; S, school; W, workplace; WT, wind turbine.

Table 4.

Risk of bias for obstetric studies

Study	Bias due to confounding	Bias due to exposure assessment	Bias due to selection of participants	Risk of bias due to post-exposure interventions	Risk of bias due to missing data	Risk of bias in measurement of outcomes	Risk of bias in selection of the reported result
Bendokiene et al., 2011[12]	Moderate	Moderate	Low	Low	Moderate	Moderate	Low
Lissåker et al., 2021[13]	Low	Moderate	Low	Low	Moderate	Moderate	Low
Nurminen and Kurppa, 1989[14]	Low	Moderate	Low	Low	Moderate	Moderate	Low
Pedersen et al., 2017a[15]	Low	Moderate	Low	Low	Serious	Serious	Low
Pedersen et al., 2017b[16]	Low	Moderate	Low	Low	Moderate	Serious	Low
Thacher et al., 2021[17]	Moderate	Low	Low	Low	Moderate	Low	Low

Table 5.

Summary of findings for noise exposure and obstetric outcomes

References	Outcomes	Relative or Absolute effects (95% CI)	No. of participants (studies)	Certainty of the evidence (GRADE)
[15]	Preeclampsia: cohort assessed with: exposure to road traffic noise	OR*: 1.10 (1.02–1.18)	72,745(1 observational study)	⊕◯◯◯Very low†,‡
[13]	Preeclampsia: cohort assessed with: exposure to occupational noise	Adjusted RR of preeclampsia in adjusted cases exposed occupational noise per 10 dBA:<70 dBA cases: 20,921 referent 70–74 dBA cases: 6478; RR: 1.02; 95% CI: 0.98–1.05 75–80 dBA cases: 1820; RR: 1.00; 95% CI: 0.94–1.06 80–85 dBA cases: 4140; RR: 1.11; 95% CI: 1.07–1.16 ≥85 dBA cases: 332; RR: 1.06; 95% CI: 0.94–1.19	1,087,944(1 observational study)	⊕◯◯◯Very low†,§
[16,17]	Gestational diabetes: cohort assessed with: exposure to road traffic noise	OR*: 0.91(0.89–0.93)	690,573(2 observational studies)	⊕◯◯◯Very low||
[13]	Gestational diabetes: cohort assessed with: exposure to occupational noise	Adjusted RR of gestational diabetes in adjusted cases exposed occupational noise per 10 dBA:<70 dBA cases: 6092 referent 70–74 dBA cases: 1959; RR: 0.99; 95% CI: 0.93–1.05 75–80 dBA cases: 543; RR: 0.93; 95% CI: 0.84–1.04 80–85 dBA cases: 1178; RR: 0.98; 95% CI: 0.90–1.06 ≥85 dBA cases: 98; RR: 0.93; 95% CI: 0.75–1.16	1,109,516(1 observational study)	⊕◯◯◯Very low†,$
[12,15]	Gestational hypertension: cohort/cross-sectional assessed with: exposure to road traffic noise	Cohort:OR*: 1.08; 95% CI: 1.02–1.15 Cross-sectional:≤50 dBA cases: 1782 referent 51–60 dBA cases: 1184; OR: 1.00; 95% CI: 0.80–1.25 ≥61 dBA cases: 155; OR: 1.36; 95% CI: 0.86–2.15	75,866(2 observational study)	⊕◯◯◯Very low‡,$,¶
[13,14]	Gestational hypertension: cohort/cross-sectional assessed with: exposure to occupational noise	Cohort:<70 dBA cases: 27,714 referent 70–74 dBA cases: 8402; RR: 1.02; 95% CI: 0.99–1.05 75–79 dBA cases: 2328; RR: 1; 95% CI: 0.95–1.05 80–85 dBA cases: 5313; RR: 1.1; 95% CI: 1.09–1.14 ≥85 dBA cases: 410; RR: 1.03; 95% CI: 0.93–1.14 Cross-sectional: RR: 1.8; 95% CI: 1.0–3.0	1,110,706(2 observational studies)	⊕◯◯◯Very low†,$

CI, confidence interval; OR, odds ratio; RR, relative risk. GRADE Working Group grades of evidence. High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate that the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited that the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate that the true effect is likely to be substantially different from the estimate of effect.^*Change per 10 dBA increase. ^†Moderate concerns with measurement of exposure, missing data, and measurement of the outcome. ^‡Pedersen et al.^[14] adjusted for maternal age, parity, prepregnancy BMI, height, disposable income, education, and season of conception. ^$Concerns with imprecision because the 95% CI cannot exclude the potential for harm. ^||Moderate concerns with confounding, measurement of exposure, and missing data. Serious concern with measurement of outcome. ^¶Moderate concerns with confounding. Serious concerns with missing data and measurement of outcomes. Supplementary Tables Impact of Noise Exposure on Risk of Developing Stress-Related Obstetric Health Effects: A Systematic Review and Meta-Analysis

Preeclampsia

Two studies reported on noise exposure (road traffic and occupational noise exposure) for the outcome of preeclampsia.[13,15] Pedersen et al.[15] reported a slight increase in the odds of preeclampsia per 10 dBA increase in road traffic noise exposure (OR: 1.1; 95% confidence interval, CI: 1.02–1.18). Lissåker et al.[13] reported on the RR of preeclampsia per 10 dBA increase in noise. When compared with <70 dBA levels, there may be an increase in the RR of preeclampsia per 10 dBA incremental increase; however, the evidence is very uncertain. The RR estimates did not suggest a dose-response relationship between noise exposure levels and preeclampsia.

The overall CoE for an association between exposure to road traffic noise and preeclampsia is very low due to concerns with risk of bias [Table 5]. The overall CoE for an association between exposure to occupational noise and preeclampsia is very low due to concerns with risk of bias and imprecision [Table 5].

Gestational diabetes

Three studies reported on noise exposure for the outcome of gestational diabetes.[13,16,17] One study measured noise from road traffic, one study measured noise from occupational exposure, and one study measured noise from road, railway, and aircraft traffic. From the two studies that measured road traffic noise, it was seen that a 10 dBA increase in road traffic noise may be associated with a decreased risk of gestational hypertension (OR: 0.91; 95% CI: 0.89–0.93) [Figure 2]; however, the certainty in the evidence is very low [Table 5].[16,17] Lissåker et al.[13] categorically reported on the RR of gestational diabetes with occupational noise exposure <70 dBA acting as the reference group. The results did not report an increased risk of gestational diabetes in population groups with higher occupational noise exposure.

Figure 2.

Forest plot of road traffic noise and risk of gestational diabetes (per 10 dBA increase).

Based on the quality of evidence, the association between exposure to road traffic noise and gestational diabetes is very uncertain due to concerns with risk of bias [Table 5]. The overall CoE for an association between exposure to occupational noise and gestational diabetes is very low due to concerns with risk of bias and imprecision [Table 5].

Gestational hypertension

Four studies reported on noise exposure for the outcome of gestational hypertension.[12,13,14,15] Two studies measured noise from road traffic and two studies measured noise from occupational exposure. One study reported on a prevalent outcome, and one reported on an incident outcome. Pedersen et al.[15] reported a slight increase in the odds of gestational hypertension per 10 dBA increase in road traffic noise exposure (OR: 1.08; 95% CI: 1.02–1.15). Similarly, Bendokiene et al.[12] reported an increase in the odds of gestational hypertension when pregnant women were exposed to road traffic noise ≥61 dBA (OR: 1.36; 95% CI: 0.86–2.15). Lissåker et al.[13] categorically reported on the risk of gestational hypertension with occupational noise exposure <70 dBA acting as the reference group. The study reported an increased risk of gestational hypertension associated with higher noise level exposure groups, with the greatest increase in RR noted in those exposed to 80 to 85 dBA levels (RR: 1.1; 95% CI: 1.06–1.14). Nurminen and Kurppa[14] reported an increased risk of gestational hypertension in populations exposed to >80 dBA levels of occupational noise exposure compared to those exposed to <80 dBA (RR: 1.8; 95% CI: 1.0–3.0). The RR estimates did not suggest a dose response relationship between noise exposure levels and gestational hypertension.

The overall CoE for an association between exposure to road traffic noise and gestational hypertension is very low due to concerns with risk of bias and imprecision [Table 5]. Similarly, the overall CoE for an association between exposure to occupational noise and gestational hypertension is also very low due to concerns with risk of bias and imprecision [Table 5].

DISCUSSION

Statement of the principal findings

Overall, this review identified a paucity of evidence reporting on the effects of noise exposure on obstetric outcomes. As indicated in some studies, higher levels of noise may be associated with an increase in the risk of preeclampsia and gestational hypertension; however, the CoE for these outcomes was very low due to concerns with risk of bias, inconsistency, and imprecision in the estimates of effects. This review portrays the need to explore further evidence on the strength of association between exposure to noise and the development of adverse obstetric effects. In addition, only studies that measured exposure to noise from road traffic, occupational sources, railway, and aircraft traffic were found. This means that the impact or certainty of evidence of other sources of exposure (e.g., mixed) on obstetric outcomes could not be assessed. This review, therefore, illustrates the need for further research that examines how noise exposure, including how different types of source exposure, may impact pregnancy-related outcomes in the mother.

This systematic review is one in a series reporting on the associations between the exposure to noise and cardiovascular, metabolic, and obstetric health outcomes.[7,8] In the cardiovascular review, it was found that higher noise exposure may be associated with an increased response in cardiac output, vascular resistance, hypertension, blood pressure, and heart rate; however, the certainty in the evidence was very low. Similarly, in the metabolic review, it was found that there may be an increase in waist circumference and adrenaline with increased noise exposure, but the CoE was also very low.

Strengths and limitations of the study

The strengths of this review include the use of a comprehensive literature search as well as transparent systematic review methods. A risk of bias tool developed for assessing studies of exposures was used for individual studies and the certainty of the body of evidence for each outcome by noise source was assessed using the GRADE approach. However, limitations included very few eligible studies and heterogeneity across studies, which prevented the pooling of results across studies using meta-analytic techniques. The study results were therefore presented narratively in the evidence profiles. Similarly, subgroup analyses, as stated a priori in the protocol, could not be performed due to the insufficient number of studies.

Meaning of the study: possible explanations and implications for stakeholders

Findings from this review may be used to direct further research in this area as there is currently limited evidence on the effects of noise exposures on the obstetric outcomes of preeclampsia, gestational diabetes, and gestational hypertension. However, the current state of evidence does not provide a robust evidence base to inform decision makers tasked with developing guidelines and policies for allowable noise exposure or noise mitigation strategies. Future research in this area will undoubtedly benefit from standardization in methodology, on both the assessment of noise exposure and the evaluation of relevant biological measures.

Unanswered questions and future research

Due to the limited evidence in this area, additional research is needed to understand the effect of noise from various sources, such as railway and aircraft traffic, on adverse obstetric outcomes. Future studies should consider the importance of utilizing high-quality exposure assessments, adjustments for potential confounders (e.g., air pollution in environmental studies), measuring noise exposure at multiple points throughout the pregnancy, and standardized outcome assessments. This will aid in the determination of whether noise exposure has differential impacts on obstetric outcomes of the mother, dependent on the trimester, as well as aid in increasing the CoE of the studies. Studies should also be conducted in a variety of settings, including examining the impact of lower levels of noise to examine the possibilities for noise threshold effects.

Financial support and sponsorship

This review was funded by Health Canada under contract no. 4500414567 with Rebecca Morgan.

Conflicts of interest

There are no conflicts of interest.