Associations Between Noise Exposure, Psychological Status, and Fertility Intentions Among Female Healthcare Professionals of Childbearing Age in Shenzhen

Abstract

Objective:

To explore the factors that influence the fertility intention of female medical staff of childbearing age in Shenzhen and analyse the association of noise exposure with their psychological status and their correlation with fertility intention.

Methods:

A questionnaire survey was conducted among 615 female medical staff members. Data on basic demographics, fertility intentions, and the general well-being scale were collected. Individual noise dosimeters were used to measure objective noise exposure during working hours. Logistic regression, chi-square test, and hierarchical regression were used for analysis. A mediation model based on logistic regression was tested using the bootstrap method (5000 resamples).

Results:

Age, occupation, professional title, education level, income, home ownership, overtime, health, marital status, spousal support, family living pattern, working hours, and noise exposure level were significantly associated with fertility intention (P < 0.05). The high noise exposure group (>65 dB[A] had significantly lower pleasure or happy mood scores (10.00 [8.00, 12.00] than the medium (12.00 [10.00, 14.00] and low groups (13.00 [11.00, 15.00], with higher anxiety symptoms (P < 0.05). Living with children and a spouse (OR = 2.170, P < 0.001) and better psychological well-being (OR = 1.071) were promoting factors for high fertility intention (ideal number of children >2). Medium noise exposure (OR = 0.478, P < 0.001), high noise exposure (OR = 0.246, P < 0.001), and working in high-noise departments (OR = 0.412, P < 0.001) were inhibiting factors for high fertility intention. Psychological status partially mediated the noise–fertility association (indirect effect = −0.134, 23.32% of the total effect).

Conclusion:

Female medical staff of childbearing age in Shenzhen were generally less willing to have more than two children. Noise exposure is associated with fertility intention both directly, by increasing psychological burden, and indirectly, by reducing positive emotions. Family support patterns also play an important role.

KEY MESSAGES

• (1)Female childbearing-age healthcare workers in Shenzhen have low fertility intentions (25.20% desiring >2 children, no gender bias).
• (2)Objective noise exposure (LAeq, 8 h) correlates negatively with fertility intentions and psychological state; high-exposure (>65 dB) groups show lower pleasure and higher anxiety.
• (3)Psychological state (especially pleasure) partially mediates this relationship (23.32% of total effect), with low noise exposure, cohabitation with spouse/children, and positive emotions as independent protective factors, while working >8 h daily inhibits fertility intentions.
• (4)Optimising workplace noise control, enhancing mental health interventions, and improving family-friendly policies are recommended to boost fertility intentions.

INTRODUCTION

With the rapid development of China’s social economy and profound changes in the country’s population structure, the birth policy has shifted from ‘one-child’ to ‘universal two-child’ and then to the ‘three-child’ policy. These policy changes are aimed at addressing the dual challenges of population ageing and declining fertility rates.[1,2] However, the relaxation of the policy has not significantly increased the fertility rate, indicating that more complex factors may be affecting people’s fertility decisions. Shenzhen is in the period of the demographic dividend and has not yet entered the ageing stage. To cope with the rapid growth of the population and the increased demand for medical resources due to future population ageing, the government has invested a large amount of medical and health resources in recent years.[3] However, this situation has further aggravated the work pressure and occupational stress of medical staff. In the special working environment of the hospital, long-term exposure to noise such as equipment operation noise, noise from patients and their families, and high-frequency noise in emergency scenes is an occupational stressor that cannot be ignored and is additionally associated with the psychological state and stress levels of medical staff.[4,5] The unique nature of medical workers’ work, coupled with environmental noise, is associated with their limited involvement in family life, which may be correlated with their fertility decisions. Female medical workers, as a crucial part of the healthcare workforce and the primary participants in childbearing, are likely influenced by a combination of factors, including noise exposure, psychological state, and family support.[6,7] However, current research on noise as an occupational environmental factor is insufficient, lacking empirical support on how it is associated with fertility intentions through psychological factors. This study aims to fill this gap through an in-depth analysis of the influencing factors on the fertility intentions of female medical workers of childbearing age in Shenzhen, focusing on revealing the mechanism of noise exposure on psychological state and its correlation with fertility intentions. This work will provide empirical evidence for understanding the fertility behaviour of specific occupational groups and developing targeted fertility support policies and occupational health protection measures.

OBJECTS AND METHODS

Survey Subjects

This study selected female medical staff of childbearing age—aged 20 to 49 years old—at Shenzhen Maternal and Child Health Hospital in Guangdong Province from 30 March 2023 to 30 September 2023 as the research subjects. The written notification included the research purpose, methods, participation time, measurement indicators, data usage, privacy protection measures, voluntary participation rights, and withdrawal mechanism. After those who voluntarily quit and those who did not meet the measurement conditions were excluded, 650 survey subjects were finally identified; questionnaires were distributed, and noise measurements were carried out. The following were the inclusion criteria: (1) on-the-job employees of the hospital; (2) working on the job for 1 year or more; (3) voluntary participation in the survey; (4) age between 20 and 49 years; and (5) good cognitive and communication skills. The exclusion criteria were (1) women under the age of 20 who had not reached the legal marriage age; (2) those who refused to participate in this survey; and (3) those who were unable to complete the noise exposure measurement because they had not worked in their original positions for a full month during the survey period due to leave, job transfer, or other reasons. All research subjects signed informed consent forms. This study was approved by the Research Ethics Committee of Shenzhen Maternal and Child Healthcare Hospital (SFYLS[2023]013) and was conducted in accordance with the ethical principles of the World Medical Association Declaration of Helsinki (2013 revision).[8]

Survey Methods

Questionnaire Design and Content

The questionnaire was divided into three parts. The first part included basic information, including age, occupation type, professional title, overtime situation, and marital relationship. The second part was for fertility intention, involving fertility status, ideal number of children, and ideal child structure. The third part contained the psychological status assessment, using the revised Chinese version of the Global Well-being Scale (GWBS).[9] The original version of the scale consists of 18 items across 6 dimensions: (1) pleasure or happy mood, (2) relaxation or tension, (3) health concerns, (4) energy, (5) satisfaction with life, and (6) optimism about the future. The specific impact of occupational noise exposure on the psychological status of the research subjects was reflected accurately by including two additional noise-related psychological items: ‘frequency of feeling anxious/nervous due to work environment noise in the past month’ and ‘degree of low mood caused by noise’, resulting in a total of 20 items in the revised scale. All items were scored on a four-point Likert scale (1 = never, 4 = always), with total score ranging from 20 to 80. A higher score indicates stronger personal well-being and fewer noise-related negative emotions. The reliability and validity of the revised scale were tested in this study, and the results showed good psychometric properties (Cronbach’s α coefficient > 0.8, Kaiser–Meyer–Olkin [KMO] value > 0.6), confirming the scale’s applicability for measuring the psychological status of the research subjects.

Objective Noise Exposure Measurement

An Aicevoos personal noise dosimeter (Wuhan Zhongce Hongtu Measurement Instrument Co., Ltd., frequency: 1 time/s) was used to record the objective noise exposure of the subjects. The data were professionally calibrated before measurement to ensure accuracy. During the working period (covering different time periods such as day shift, night shift, and shift work), the subjects wore the dosimeter on their chest or shoulder, and it continuously recorded the equivalent sound level (LAeq, 8 h) and peak sound level (Lpeak) for 8 h. The measurement time (work/rest), position, and work task were recorded simultaneously to distinguish the noise source (equipment noise, noise from patients and their families, office noise, etc.). Each subject was continuously measured for three complete work cycles (including different shifts), and the average value was taken as the individual’s long-term noise exposure level. With the ‘Occupational exposure limits for hazardous agents in the workplace Part 2: Physical agents’ (GBZ2.2-2007)[10] and the Sound Environment Quality Standard (GB 3096-2008)[11] used as references, and combined with the actual noise distribution characteristics of hospital clinical settings, the noise exposure was divided into three levels: low exposure (LAeq, 8 h < 50 dB[A], medium exposure (50–65 dB[A], and high exposure (>65 dB[A]. The measured range of LAeq, 8 h was 42.3–78.9 dB(A), and the measured range of Lpeak was 65.8–102.5 dB(A) in this study.

Reliability and Validity Analysis

The reliability and validity of the questionnaire results were evaluated using the reliability coefficient and validity coefficient.[12] (1) Reliability test: The Cronbach’s α coefficients of each dimension of the GWBS scale are all greater than 0.8, indicating that the questionnaire has high reliability. (2) Validity test: The KMO sampling appropriateness scale of the GWBS scale is greater than 0.6, indicating that the questionnaire has high validity. Thus, both reliability and validity passed the test.

Statistical Methods

SPSS 26.0 (IBM Corp., Armonk, NY, USA) was used for statistical analysis. The Shapiro–Wilk test was used to examine the normality of continuous variables. Chi-square and rank-sum tests were used for group comparisons. Spearman correlation analysis was used to explore the correlation between noise exposure and GWBS total score. A hierarchical regression and stepwise confounding control analysis strategy was adopted, and three logistic regression models were constructed in three steps to clearly distinguish the direct effect of noise exposure and the confounding effect of other variables. A two-sided significance level of α = 0.05 was adopted for all tests; P < 0.05 was considered statistically significant. For bootstrap mediation analysis, a 95% CI not containing 0 indicated significance. According to the survey by Hao M et al.,[13] Chinese women’s willingness to have a third child dropped sharply, so the dependent variable was uniformly set as ideal number of children >2 = 1, ≤2 = 0. A mediation analysis model based on logistic regression was employed and was constructed as follows: The independent variable, noise exposure level, was treated as an ordinal categorical variable (low exposure = 0, moderate exposure = 1, and high exposure = 2). Mediation effects were validated using the bootstrap method (5000 resamples). The proportion of the mediation effect was calculated as (mediation effect value/total effect value) × 100%. All figures, including the directed acyclic graph (DAG) and the mediation model diagram, were drawn using GraphPad Prism 9.0 (GraphPad Software Inc., San Diego, CA, USA).

RESULTS

General Information on Female Medical Staff of Childbearing Age

This study distributed 650 questionnaires and collected 615 valid questionnaires, with a valid response rate of 94.62%. The subjects were childbearing-age female medical staff aged 20 to 49, mainly composed of young people aged 20–30 (56.09%). Nurses accounted for the largest proportion of occupations (61.63%). In terms of education, a junior college degree was the educational level achieved by most of the subjects (61.30%); economically, 45.53% had an after-tax annual income of less than 100,000 yuan, and more than half (55.28%) did not own a house and were renting their homes. In terms of marital status, unmarried and first-married accounted for nearly half (48.29% and 49.59%, respectively). Work-wise, 40% had 1–3 years of working experience, more than 60% worked more than 8 h a day, and 30.24% worked overtime most of the time. With regard to their self-assessed health status, the majority were in general or good condition (80.49% in total).

The participants were further divided into high-noise departments (ICU, emergency department, operating room; 32.20%) and medium–low noise departments (67.80%). The average LAeq, 8 h in the high-noise department was 67.54 ± 6.12 dB(A), and that in the medium- and low-noise departments was 55.38 ± 5.74 dB(A) (P < 0.001). Noise exposure status: The average objective measurement of LAeq, 8 h was 58.42 ± 7.35 dB(A), including 166 people (27.00%) in the low-exposure group, 320 people (52.03%) in the medium-exposure group, and 129 people (20.98%) in the high-exposure group. The mean Lpeak of each group was as follows: low-exposure group (72.5 ± 4.2 dB(A)), medium-exposure group (81.3 ± 5.7 dB(A)), and high-exposure group (94.8 ± 6.3 dB(A)), with significant differences among groups (F = 102.35, P < 0.001).

Analysis of the Current Status and Differences in Fertility Intentions Among Female Medical Staff of Childbearing Age

Overall Fertility Situation

Among the respondents, 54.15% had not given birth, 26.50% had given birth to one child, and 19.35% had given birth to two children. With regard to the ideal number of children, 74.80% (460 people) hoped to have ≤2 children, and only 25.20% (155 people) hoped to have >2 children; 63.90% hoped to have one boy and one girl [Table 1].

Table 1.

Current fertility status and intentions among healthcare workers of childbearing age [n (%)]

Fertility intention	Variable	Number of examples (n)	Percentage (%)
Fertility circumstance	No child	333	54.15
	One child	163	26.50
	Two children	119	19.35
Ideal number of children	>2	155	25.20
	≤2	460	74.80
Ideal child structure	0	46	7.48
	1 boy	55	8.94
	1 girl	60	9.76
	1 boy and 1 girl	393	63.90
	2 boys	6	0.98
	2 girls	22	3.58
	3 children and others	33	5.37

Basic Information, Noise Exposure, and Differences in Fertility Intentions

The univariate results showed that among the included variables, except for spouse’s occupation, years of work experience, and body mass index (BMI)—which were not statistically significantly associated with fertility intentions (P > 0.05)—all other included variables demonstrated statistically significant associations with fertility intentions (ideal number of children >2) (P < 0.05), as shown in Table 2.

Table 2.

Analysis of differences in basic information, family circumstances and fertility intentions [n (%)]

Variable	Classification	Ideal number of children		χ2	P
		≤2	>2
Age (years old)	20–30	250 (54.35)	95 (61.29)	32.157	<0.001*
	31–40	157 (34.13)	44 (28.39)
	41–49	53 (11.52)	16 (10.32)
Position	Nurse	272 (59.13)	107 (69.03)	8.941	0.030*
	Doctor	115 (25.00)	27 (17.42)
	Medical Technicians	52 (11.30)	13 (8.39)
	Other	21 (4.57)	8 (5.16)
Professional title	Junior	152 (33.04)	55 (35.48)	9.762	0.021*
	Intermediate	189 (41.09)	64 (41.29)
	Senior	73 (15.87)	16 (10.32)
	None	46 (10.00)	20 (12.90)
Education	Junior college	286 (62.17)	91 (58.71)	7.358	0.025*
	Undergraduate course	148 (32.17)	53 (34.19)
	Master’s degree or above	26 (5.65)	11 (7.10)
Annual income after tax (¥10,000)	<10	210 (45.65)	70 (45.16)	11.763	0.008*
	10–30	135 (29.35)	28 (18.06)
	30–50	85 (18.48)	44 (28.39)
	>50	30 (6.52)	13 (8.39)
Status of owner-occupied property	None (renting a house)	243 (52.83)	97 (62.58)	7.125	0.028*
	1	166 (36.09)	51 (32.90)
	≥2	51 (11.09)	7 (4.52)
Overtime	Most of the time	138 (29.98)	48 (30.97)	10.236	0.006*
	Occasional	215 (46.74)	78 (50.32)
	Hardly	107 (23.28)	29 (18.71)
Self-assessment of health status	Commonly	196 (42.61)	61 (39.35)	8.124	0.044*
	Fine	181 (39.35)	57 (36.77)
	Range	68 (14.78)	24 (15.48)
	Excellence	15 (3.26)	13 (8.39)
Marital status	Unmarried	202 (43.91)	98 (63.23)	14.462	<0.001*
	First marriage	251 (54.57)	54 (34.84)
	Remarriage or divorce	7 (1.52)	3 (1.94)
Spouse’s occupation	Institution	148 (32.17)	49 (31.61)	3.258	0.353
	Enterprise units	137 (29.78)	46 (29.68)
	Liberal professions	94 (20.43)	31 (20.00)
	Other	81 (17.61)	29 (18.71)
Frequency of spousal emotional support	Most of the time	264 (57.39)	24 (15.48)	12.873	0.005*
	Sometimes	122 (26.52)	51 (32.90)
	Occasional	49 (10.65)	63 (40.39)
	Hardly	25 (5.43)	17 (11.03)
Parental relationship with spouse	Harmonious	246 (53.48)	82 (52.90)	9.152	0.010*
	Commonly	177 (38.48)	68 (44.13)
	Not good	37 (8.04)	5 (3.23)
Family residence pattern	Living alone	151 (32.83)	68 (43.87)	20.157	<0.001*
	Living with spouse and children	127 (27.61)	26 (16.77)
	Living with parents	106 (23.04)	30 (19.35)
	Living with one’s spouse’s parents	56 (12.17)	9 (5.81)
	Other	30 (6.52)	12 (7.74)
Working life (years)	1–3	185 (40.22)	61 (39.35)	4.872	0.181
	4–10	205 (44.57)	68 (43.87)
	11–20	59 (12.83)	20 (12.90)
	>20	11 (2.39)	6 (3.87)
Daily working hours (h)	≤8	160 (34.78)	55 (35.48)	12.876	0.005*
	9–10	215 (46.74)	72 (46.45)
	11–12	72 (15.65)	21 (13.55)
	>12	13 (2.83)	7 (4.52)
BMI (kg/m2)	<18.5	28 (6.09)	10 (6.45)	2.154	0.542
	18.5–23.9	323 (70.22)	98 (63.23)
	24.0–27.9	96 (20.87)	41 (26.45)
	≥28.0	13 (2.83)	6 (3.87)
Noise exposure level	Low exposure group	125 (27.17)	41 (26.45)	5.13	0.024*
	Intermediate exposure group	241 (52.39)	79 (50.97)
	High-exposure group	93 (20.22)	36 (23.23)
Department	Low-medium noise department	388 (84.35)	107 (69.03)	23.567	<0.001*
	High-noise department (ICU/ED/OR)	72 (15.65)	48 (30.97)

Note: BMI is body mass index; *P means <0.05.

Overall GWBS Scores

The overall psychological state level of female healthcare workers of childbearing age (GWBS total score) was 76.00 (69.00, 79.00), indicating a moderate level of overall well-being. The ‘pleasure or happy mood’ dimension had the highest median score of 12.00 (10.00, 14.00), and statistically significant differences existed among different fertility intention groups and noise exposure groups (P < 0.05). The score of this dimension in the low-noise-exposure group (13.00 [11.00, 15.00] was significantly higher than that in the moderate-exposure group (12.00 [10.00, 14.00] and the high-exposure group (10.00 [8.00, 12.00]. The scores for anxiety and irritability-related specific items in the high exposure group (3.00 [2.00, 4.00] were significantly higher than those in the moderate and low exposure groups (2.00 [1.00, 3.00], 1.00 [1.00, 2.00]. Differences among noise exposure groups were also observed in the other four dimensions (relaxation/tension, energy, life satisfaction, and future optimism; all P < 0.05), whereas no significant difference was found in the health concerns dimension (P > 0.05). Notably, the magnitudes of these differences were smaller than those in the ‘pleasant or pleasant mood’ dimension, indicating that noise exposure had a more significant impact on positive emotions. The results are shown in Table 3.

Table 3.

Comparison of GWBS scores among different noise exposure groups [M (P25, P75)]

GWBS dimension/item	Total (n = 615)	Low noise	Medium noise	High noise
Total score	76.00 (69.00, 79.00)	79.00 (73.00, 80.00)	77.00 (70.00, 79.00)	71.00 (64.00, 77.00)*
1. Pleasure or happy mood	12.00 (10.00, 14.00)	13.00 (11.00, 15.00)	12.00 (10.00, 14.00)	10.00 (8.00, 12.00)*
2. Relaxation or tension	10.00 (8.00, 12.00)	11.00 (9.00, 13.00)	10.00 (8.00, 12.00)	8.00 (6.00, 10.00)*
3. Health concerns	9.00 (7.00, 11.00)	9.00 (7.00, 11.00)	9.00 (7.00, 11.00)	8.00 (6.00, 10.00)
4. Energy	10.00 (8.00, 12.00)	11.00 (9.00, 13.00)	10.00 (8.00, 12.00)	8.00 (6.00, 10.00)*
5. Satisfaction with life	9.00 (7.00, 11.00)	10.00 (8.00, 12.00)	9.00 (7.00, 11.00)	7.00 (5.00, 9.00)*
6. Optimism about the future	8.00 (6.00, 10.00)	9.00 (7.00, 11.00)	8.00 (6.00, 10.00)	6.00 (4.00, 8.00)*
Noise-specific item 1: Frequency of anxiety/tension	2.00 (1.00, 3.00)	1.00 (1.00, 2.00)	2.00 (1.00, 3.00)	3.00 (2.00, 4.00)*
Noise-specific item 2: Degree of low mood	2.00 (1.00, 3.00)	1.00 (1.00, 2.00)	2.00 (1.00, 3.00)	3.00 (2.00, 4.00)*

Note: GWBS is General Well-being Scale. The revised version of the GWBS with additional noise-related items was adopted in this table. All 20 items were scored on a four-point scale, with the total score ranging from 20 to 80. *P < 0.05 versus low-noise-exposure group.

Analysis of Factors Influencing the Fertility Intentions of Medical Staff of Childbearing Age in Shenzhen

Model design

Model 1 (baseline mixed model): Only demographic and sociological characteristics such as age, marital status, education level, income, and housing status were included to control for baseline confounding. Potential clustering effects were ruled out by calculating intracluster correlation coefficients, which were <0.05, indicating negligibility. Therefore, standard logistic regression was ultimately used to achieve parsimony and interpretability.

Model 2 (core variable model): With Model 1 used as a basis, the direct effect of noise exposure was analysed after controlling for baseline confounding factors.

Model 3 (full-variable model): This model supplements Model 2 with variables such as psychological state (pleasure dimension), family living pattern, years of work, and daily working hours to clarify the independent effects of each key variable.

Some variables that were significant in univariate analysis (e.g., occupation, professional title, self‑assessed health, spousal support, and parental–spousal relationship) were not included in the final multivariable model due to multicollinearity (e.g., VIF > 10) or weak relevance to the core mechanism. Only variables with P < 0.05 after adjustment were retained to ensure model robustness and parsimony.

Results of stratified regression analysis

Spearman correlation analysis showed that objective noise exposure was negatively correlated with GWBS total score (rs = −0.386, P < 0.001). The likelihood ratio tests for all three models were statistically significant (P < 0.001), with progressively optimised fit: the McFadden pseudo-R² increased from 0.186 in Model 1 to 0.243 in Model 2 and further to 0.327 in Model 3, indicating that the model’s explanatory power for fertility intention was significantly enhanced after incorporating noise exposure and psychological status.

The results of Model 1 indicated that age 31–49 (OR = 1.523, P = 0.041) and first marriage status (OR = 2.847, P < 0.001) were promoting factors for fertility intention, while economic factors such as income and property ownership had no significant impact.

In Model 2, incorporating noise exposure, medium noise exposure (OR = 0.503, P < 0.001) and high noise exposure (OR = 0.275, P < 0.001) were inhibiting factors for high fertility intention, while the effect magnitude of baseline variables showed no significant change, indicating that the impact of noise exposure was not significantly confounded by demographic and sociological characteristics.

In the full-variable Model 3, medium noise exposure (OR = 0.478) and high noise exposure (OR = 0.246, both P < 0.001) remained significant inhibiting factors for high fertility intention. Working in a high‑noise department was also an independent inhibiting factor (OR = 0.412, P < 0.001). Meanwhile, higher positive mood (OR = 1.071, P = 0.014) and living with the spouse and children (OR = 2.170, P < 0.001) were promoting factors for fertility intention. Conversely, work experience of 4–10 years (OR = 0.582, P = 0.029) and daily working hours exceeding 8 h (9–10 h OR = 0.624, 11–12 h OR = 0.357, and >12 h OR = 0.281, all P < 0.05) were inhibiting factors for high fertility intention. BMI levels showed no significant impact on fertility intention (all P > 0.05) [Table 4].

Table 4.

Hierarchical logistic regression analysis results of fertility intention (>2 children = 1)

Independent variable	Reference group	Model 1 (baseline model)		Model 2 (core variable model)		Model 3 (full variable model)
		OR (95% CI)	P	OR (95% CI)	P	OR (95% CI)	P
Age	20–30 years old
	31–40 years old	1.523 (1.031–2.251)	0.035*	1.507 (1.020–2.223)	0.040*	1.489 (1.005–2.205)	0.047*
	41–49 years old	1.632 (1.015–2.623)	0.043*	1.618 (1.003–2.610)	0.048*	1.596 (0.987–2.582)	0.056
Marital status	Unmarried
	First marriage	2.847 (2.012–4.035)	<0.001	2.813 (1.985–3.982)	<0.001	2.765 (1.950–3.921)	<0.001
	Re-marriage or divorce	1.982 (0.654–6.018)	0.228	1.953 (0.645–5.917)	0.235	1.901 (0.628–5.754)	0.251
After-tax annual income	Below 100,000
	100,000–300,000	1.125 (0.803–1.574)	0.486	1.118 (0.797–1.569)	0.501	1.103 (0.785–1.548)	0.572
	300,000–500,000	1.201 (0.773–1.864)	0.415	1.189 (0.765–1.847)	0.440	1.172 (0.753–1.822)	0.488
	More than 500,000	1.326 (0.721–2.439)	0.361	1.308 (0.710–2.407)	0.383	1.285 (0.698–2.365)	0.421
Status of owner-occupied property	None (renting a house)
	1	1.083 (0.769–1.527)	0.658	1.075 (0.763–1.516)	0.682	1.059 (0.750–1.496)	0.745
	≥2	1.152 (0.689–1.923)	0.583	1.141 (0.680–1.913)	0.605	1.128 (0.669–1.889)	0.647
Noise exposure level	Low-exposure group (<50 dB[A])
	Medium-exposure group	-	-	0.503 (0.352–0.718)	<0.001	0.478 (0.334–0.683)	<0.001
	High-exposure group	-	-	0.275 (0.182–0.415)	<0.001	0.246 (0.161–0.376)	<0.001
Department	Low–medium noise department
	High‑noise department (ICU/ED/OR)	-	-	-	-	0.412 (0.256–0.663)	<0.001
Psychological state (GWBS total score)	-	-	-	-	-	1.071 (1.015–1.130)	0.014*
	Pleasure or happy mood	-	-	-	-	1.085 (1.021–1.153)	0.008*
	Relaxation or tension	-	-	-	-	1.042 (0.983–1.105)	0.176
	Health concerns	-	-	-	-	1.036 (0.975–1.099)	0.251
	Energy	-	-	-	-	1.058 (0.997–1.123)	0.063
	Satisfaction with life	-	-	-	-	1.049 (0.988–1.114)	0.127
	Optimism about the future	-	-	-	-	1.053 (0.991–1.119)	0.108
Family living pattern	Living alone
	Living with spouse and children	-	-	-	-	2.170 (1.345–3.498)	0.001*
	Living with parents	-	-	-	-	1.283 (0.821–2.009)	0.273
	Living with one’s spouse’s parents	-	-	-	-	1.195 (0.721–1.979)	0.492
Working years	1–3
	4–10	-	-	-	-	0.582 (0.358–0.947)	0.029*
	11–20	-	-	-	-	0.615 (0.348–1.084)	0.092
	>20	-	-	-	-	0.653 (0.287–1.485)	0.307
Daily working hours	≤8
	9–10	-	-	-	-	0.624 (0.432–0.899)	0.011*
	11–12	-	-	-	-	0.357 (0.218–0.584)	<0.001
	>12	-	-	-	-	0.281 (0.135–0.585)	<0.001
BMI	<18.5
	18.5–23.9	-	-	-	-	1.032 (0.601–1.776)	0.903
	24.0–27.9	-	-	-	-	1.058 (0.607–1.850)	0.845
	≥28.0	-	-	-	-	1.102 (0.489–2.475)	0.815

Note: GWBS is General Well-being Scale. High‑noise department: ICU, emergency department, operating room. BMI is body mass index. “-” means that the model does not include indicators and has no values; *P means <0.05.

Directed Acyclic Graph for Conceptualising Associations and Confounder Selection

DAG plots were constructed to elucidate the hypothesised causal relationships among occupational noise exposure, psychological state, fertility intention, and relevant covariates and to guide covariate selection for multivariable regression models. As shown in Figure 1, DAG confirmed that sector-level clustering was not a major source of confounding and that no substantial collision bias or overadjustment occurred in the final model.

Figure 1.

DAG of noise exposure, psychological status, and fertility intention.

Mediation Analysis of Noise Exposure, Psychological State, and Fertility Intention

The mediating role of psychological state in the association between noise exposure and fertility intention was clarified by constructing and testing a logistic regression-based mediation model using the bootstrap method (5000 resamples), with unstandardised regression coefficients (B) reported for all paths.

Independent variable: noise exposure level (ordinal categorical variable: low exposure = 0, moderate exposure = 1, and high exposure = 2); mediating variable: psychological state (continuous variable: total score of the GWBS); and dependent variable: fertility intention (binary variable: willing to have >2 children = 1, otherwise = 0).

The results showed that all paths of the mediation model were statistically significant (P < 0.001), and psychological state played a partial mediating role between noise exposure and fertility intention.

Path a (Noise exposure → Psychological state): B = −0.312, 95% CI: −0.401 to −0.223, P < 0.001 (indicating that higher noise exposure was significantly associated with lower GWBS total scores, i.e., poorer psychological status);

Path b (Psychological state → Fertility intention): B = 0.430, 95% CI: 0.328–0.532, P < 0.001 (indicating that better psychological status was significantly associated with higher willingness to have >2 children);

Path c (Total effect: Noise exposure → Fertility intention): B = −0.526, 95% CI: −0.718 to −0.334, P < 0.001;

Path c’ (Direct effect: Noise exposure → Fertility intention): B = −0.392, 95% CI: −0.575 to −0.209, P < 0.001;

Indirect effect (Path a × Path b): effect = −0.134, 95% CI: −0.206 to −0.062, accounting for 23.32% of the total effect. See Table 5 and Figure 2.

Table 5.

Mediation effect analysis results of psychological state (GWBS total score) between noise exposure and fertility intention (bootstrap = 5000 times)

Effect Type	Estimate	95% confidence interval (95% CI)	P	Proportion of mediating effect (%)
Total effect (Noise → Fertility intention)	−0.526	−0.718 to −0.334	<0.001	-
Direct effect (Noise → Fertility intention)	−0.392	−0.718 to −0.334	<0.001	-
Indirect effect (Noise → Psychological state → Fertility intention)	−0.134	−0.206 to −0.062	<0.001	23.32

Note: Solid arrows indicate hypothesised directional associations. Confounders included age, marital status, annual income, homeownership, clinical department, working years, daily working hours, and family living pattern. Psychological status was treated as a mediator in the pathway from noise exposure to fertility intention.

Figure 2.

Mediation model of psychological status between noise exposure and fertility intention.

DISCUSSION

Fertility Rate Decline Linked to Reduced Childbearing Willingness and Noise Exposure

The ideal number of children is a commonly used dimension in current research on fertility intentions,[14] and it is most closely related to fertility behaviour. This study observed that childbearing-age female healthcare workers in Shenzhen generally showed relatively low willingness to have >2 children, with only 25.20% hoping to have >2 children, and there was no obvious gender bias. This result may be related to the preference for small families in contemporary society and the gradual popularisation of gender equality values.[15,16] In addition, the results of this study indicate a significant negative correlation between noise exposure level and the willingness to have more than two children (P < 0.001); staff in high‑noise departments had higher noise exposure, poorer psychological status, and lower fertility intention, indicating concentrated vulnerability in acoustically stressful work areas. Specifically, multivariate logistic regression analysis showed that high noise exposure was significantly associated with lower fertility intention (OR = 0.246, 95% CI = 0.161–0.376, P < 0.001) after adjusting for age, education level, marital status, and household income. This finding reveals the potential impact of noise exposure in the occupational environment on fertility decisions. For medical staff, noise exposure, as an occupational stressor, combined with work pressure, shift system, and concerns about occupational exposure during pregnancy,[17] further reduces the enthusiasm to have more than two children.[18] If the persistent low willingness to have more than two children is not improved, then it may have a long-term impact on the population structure and socio-economic development of Shenzhen.[19,20] Therefore, policymakers and planners must conduct in-depth analysis of these impacts and develop corresponding long-term response strategies, such as strengthening education and training, optimising the labour market structure, and promoting the development of social support systems.

Synergistic Effects of Noise Exposure, Psychological State, and Family Support on Fertility Intention

Univariate analysis shows that multiple factors are involved in the formation of the willingness to have more than two children. Age, position, professional title, education level, economic status, work intensity, health status, family relations, and other aspects jointly affect the fertility decision-making of female medical staff. Willingness to have more than two children in the 31–40 age group was significantly higher than in the 20–30 age group (OR = 1.489, P = 0.047) after adjusting for confounding factors. This result may be related to the fact that older women have more stable economic foundations, stronger family support, and clearer fertility plans. Nurses are more willing to have more than two children than doctors and medical technicians are, which may be related to differences in work pressure and career development paths among different occupations. The low-income group and the group without housing have a relatively higher willingness to have more than two children, which is contrary to the traditional cognition that economic conditions restrict fertility intention. This result may have occurred because the sample group has a strong demand for family support, or it is affected by the expectation of policy support. Good health status, harmonious family relations, and appropriate work intensity are also conducive to improving the willingness to have more than two children.

Studies have found that noise exposure, family living patterns (such as living with children and spouses), and psychological state levels (especially the ‘pleasure or happy mood’ dimension) are independent influencing factors of willingness to have more than two children, and occupational noise exposure has a significant negative correlation with fertility intention—a higher noise exposure level corresponds to lower willingness of female medical staff to have more than two children. Moreover, psychological state plays a partial mediating role in the negative association between noise exposure and fertility intention: Long-term exposure to high-intensity occupational noise not only exerts a direct inhibitory effect on fertility intention by increasing the occupational psychological burden of medical staff but also indirectly reduces fertility willingness by impairing their psychological state (e.g., reducing positive emotions and inducing anxiety symptoms). This study confirms this mechanism through logistic regression-based mediation analysis, with the indirect effect estimate standing at −0.134 (95% CI: −0.206 to −0.062), accounting for 23.32% of the total effect. This finding indicates that a positive psychological state can significantly enhance an individual’s ability to cope with the challenges of having more than two children.[21] Specifically, individuals with a pleasant mood may be more optimistic about the pressure of having more than two children and have a higher subjective well-being through family support (such as spousal emotional support), thereby forming a virtuous cycle and enhancing their confidence in having more than two children.[22] At the same time, mediation effect analysis confirms that noise exposure indirectly affects the willingness to have more than two children by inhibiting mood (indirect effect = −0.134, 95% CI: −0.206 to −0.062), while a direct influence pathway also exists. Long-term noise exposure will trigger a physiological stress response in the human body, leading to increased cortisol and decreased sleep quality, which in turn leads to psychological problems such as irritability, anxiety, and depression.[23] Studies have shown that occupational noise exposure is associated with hearing loss, fatigue, and burnout.[24] The decrease in mood is correlated with reduced confidence and enthusiasm in coping with the pressure of having more than two children. Medical staff in Shenzhen generally face the dual challenges of high-intensity work pressure and noise exposure, which are further associated with poorer psychological states and indirectly correlated with reduced willingness to have more than two children. Therefore, policymakers should prioritise optimising the working environment for medical staff, addressing noise exposure issues, and increasing psychological counselling services to improve medical staff’s mental well-being. Furthermore, considering the inhibitory effect of Shenzhen’s high cost of living on the willingness to have more than two children, this effect can be alleviated through economic measures such as tax breaks and priority allocation of affordable housing.

Future Research Directions and Policy Recommendations

In summary, female medical workers of childbearing age in Shenzhen have complex and diverse fertility intentions (especially the willingness to have more than two children). Future research will further expand the sample scope to include more medical institutions and occupational groups to enhance the generalisability of the results. Comparing the differences in the willingness to have more than two children among different occupational groups ensures that a more comprehensive understanding of the impact of occupational characteristics on fertility decisions can be achieved. Specific measures may include establishing psychological counselling centres in medical institutions to provide regular psychological state assessments and interventions; extending flexible parental leave to encourage spousal participation in childcare; providing tax breaks and priority allocation of affordable housing for families with more than two children; optimising the working environment and reducing unnecessary overtime; optimising noise control by adjusting shift scheduling on the basis of noise exposure levels to avoid long periods of continuous work in high-noise environments; and using media to dispel the negative image of having more than two children and to support career development. These measures will help alleviate the fertility pressure on medical workers, stimulate their enthusiasm for childbearing, and promote the long-term balanced development of the population.

Study Limitations

The revised GWBS has inconsistent scoring criteria compared with the original version, making a direct comparison with relevant research results impossible. Future research can adopt the original GWBS combined with an independent occupational noise psychological perception scale. This study did not explore the impact of the inconsistency between subjective noise perception and objective noise exposure on psychological state and fertility intention. The research conclusions on noise exposure are limited to the perspective of objective measurement. The study only selected subjects from a single medical institution, with a single sample source and selection bias, and the results cannot be directly generalised to all childbearing-age female medical staff in Shenzhen. Although high‑noise occupational subgroups were analysed, detailed comparisons among individual high‑noise departments (ICU versus emergency versus operating room) were not performed. Follow-up research needs to expand the sample to different levels and types of medical institutions and to conduct more detailed comparisons in different departments. The dichotomisation of fertility intention ignores the differences in fertility willingness among subgroups such as ‘no fertility intention’ and ‘willing to have one child’, which reduces the comparability of the results and may mask the heterogeneous impact of noise exposure and psychological status on different fertility intention levels.

Given the cross-sectional, single-centre design and potential residual confounding factors, the conclusions of this study are correlational rather than causal. Additionally, this study did not present baseline characteristics of different noise exposure groups, and the evidence for noise–fertility intention association lacks subgroup and sensitivity analyses, limiting the robustness of the conclusions. Cumulative noise exposure dose was not measured, thus making dose–response exploration impossible. Future research should present group baseline characteristics, conduct subgroup/sensitivity analyses, measure cumulative noise dose, and adopt prospective designs to strengthen causal inference.

An important detail to note is that this study adopts a cross-sectional design, which can only establish associations between variables rather than confirm causal relationships. The observed associations between noise exposure, psychological status, and fertility intention do not imply a unidirectional causal effect. For example, the negative association between noise exposure and fertility intention may be confounded by unmeasured factors such as work stress, lifestyle, or personal values, and reverse causality cannot be completely ruled out. Therefore, the conclusions of this study are limited to correlational findings, and the potential causal relationships need to be verified by prospective cohort studies or intervention studies in the future.

CONCLUSION

On the basis of the psychological status measurement results from the revised GWBS, this study provides preliminary evidence that workplace noise exposure may be related to low willingness to have more than two children among childbearing-age female healthcare workers in Shenzhen, possibly through direct psychological pressure and indirect impairment of positive emotions. Family support may serve as a protective factor for fertility intention. Several demographic, occupational, and family factors were associated with fertility intention in this sample. These findings may support the optimisation of occupational noise control, psychological intervention, and family-friendly policies to potentially improve the fertility-related willingness of this population.

Availability of Data and Materials

The original contributions presented in the study are included in the article. Further inquiries can be directed to the corresponding authors.

Author Contributions

WenWen Wan contributes to the study conceptualization, core knowledge content definition, and research design. ChangYu Deng conducted literature retrieval and oversaw clinical research components. SiQi Chen managed experimental research, objective noise measurement, and data collection. He Wang performed statistical analysis, including regression and mediation tests, and ensured data consistency. YueYun Wang took charge of manuscript preparation, synthesizing findings and structuring the academic writing. Authors read and approved the final manuscript, took public responsibility for appropriate portions of the content, and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or completeness of any part of the work were appropriately investigated and resolved.

Ethics Approval and Consent to Participate

All research subjects signed informed consent forms; this study was approved by the Research Ethics Committee of Shenzhen Maternal and Child Healthcare Hospital (SFYLS[2023]013).

Conflicts of Interest

The authors declare no conflicts of interest.

Funding Statement

Nil.