PRECISE - Objective Measures and Observations of Physical Exposures Among Pregnant Employees in the Danish Workforce

Occupational physical activity during pregnancy may decrease blood supply to the uterus and be connected to negative pregnancy outcomes. Every year, this notion prompts many pregnant women to be absent from their workplace in order to protect themselves and their unborn children. However, little is known about the actual extent and impact of physical workloads among expecting mothers in Denmark and whether this high rate of pregnancy related absence is necessary. In consequence, current guidelines from the Danish Working Environment Agency have omitted former limits on lifting during pregnancy due to lack of evidence. (1) Former studies have mainly been based on self-reported data with no measurements of exposures. A comprehensive evaluation of physical exposures in Danish workplaces and potential consequences for pregnant women will provide a rational basis for organization of work with improved options for prevention of adverse health effects. By use of measurements, observations and real-time self-report, the overall aim of the project is to attain precise data on physical workloads (lifting, standing/walking and forward bending) among pregnant employees in the Danish workforce, and investigate how these relate to discomforts, complications and negative pregnancy outcomes.

Background and impact In the Scandinavian countries, 75-80% of women are a part of the workforce and many of these are of reproductive age.(2) Physical workloads are a recurring issue in risk assessment and counselling of pregnant women regarding work. Epidemiological studies have indicated an association with preterm birth and foetal death, (3-9) but have been based on self-report mostly with a retrospective design with obvious limitations regarding recall bias, crude and no time specific exposure assessments. Few studies have applied Job Exposure Matrices (JEMs) for exposure assessments, but these studies have been small and/or based on self-report or expert assessment.(4, 10-14) Overall, previous studies have been limited by misclassification of exposures and a lack of exposure contrast with few highly exposed pregnant women. Pregnancy discomforts are common and often reason for absence during pregnancy.(15-16) Since pregnancy discomforts only in extreme cases involve hospital contacts and thereby entry in the Danish registries, studies rely on self-report. To be able to prevent absence or other negative consequences, we need more knowledge on the importance of physical workloads in relation to development or aggravation of pregnancy discomforts. Results from PRECISE will help improve current knowledge on physical workloads during pregnancy and advance national guidelines, which potentially can prevent negative outcomes in pregnancies and reduce worries and absence among pregnant employees. Further, the construction of trimester-specific JEMs based on objective exposure assessments will expand and support research in occupational reproduction and enable future investigations of associations with other outcomes in pregnancy and offspring. Qualifications and previous related activities of the applicant PRECISE is a natural extension of the applicant Luise Mølenberg Begtrup´s commitment to secure safe working environments for mothers and offspring. In both her research and clinical practice Dr. Begtrup continously strive to improve guidance of healthcare professionals, employers and decision makers and provide information, advice and work planning for pregnant women. She is PI for projects focusing on the wellbeing of pregnant women and their retention at workplaces and was pivotal in the establishment of the nationwide occupational generation cohort DOC\*X-G, and in research of night work and pregnancy outcomes. Dr. Begtrup has been involved in the National Board of Health´s update for maternal care, and recently finished a review to the Danish Working Environment Authority about work-related ergonomic exposures and pregnancy. Research plan The project period of four years is organized in six interrelated work packages (WP1-WP6). The PhD-student covered by this application is planned to start 01/03/2023 and will be recruited through postings through the Capital Region of Denmark and the University of Copenhagen. WP1 Preparation (01/03/2022-01/02/2023): Obtainment of permits, recruitment and enrolment of two PhD-students, training of student assistants and preparation of project information and development of questionnaires and observer-courses. WP2 Recruitment of working pregnant women and colleagues (01/12/2022-30/04/2024): The target group is working pregnant women. Further, a smaller group of non-pregnant women from the same workplaces and with the same work tasks (colleagues) will be included. Around 95% of pregnant women in Denmark have a 1st trimester ultrasound scan.(17) To ensure a representative population, the recruitment to the project takes place at four obstetric departments. After giving informed written consent for participation, the pregnant women receive a baseline questionnaire followed by short weekly questionnaires by SMS until maternity leave. Based on the distribution of pregnant women in the Danish workforce and existing knowledge of occupations and physical exposures,(18-19) we have identified 42 primary occupations of interest. We will prioritize occupations with a high prevalence of pregnant women and secure representation of both high and low exposure to physical workloads. In this way, we ensure adequate coverage of physical workload exposures for most pregnant women in Denmark. Through the participating pregnant women, we will contact non-pregnant colleagues with similar work tasks. In the event of inadequate recruitment of colleagues, we will, with the support of members in our stakeholder group, contact other relevant workplaces for additional recruitment. Recruitment will last six months in each obstetric department. We expect to invite 6,800 working pregnant women.(20) With a 45% participation rate, we will be able to enrol 3,000 women answering questionnaires. To make a valid average exposure estimate, we need measurements/observations on 5-10 pregnant women/occupation.(21) Thus, measurements/observations will be performed on ≥ 200 pregnant women and 100 non-pregnant women. WP3 Data-collection (01/01/2023-30/06/2024): The baseline questionnaire contains questions about lifestyle, previous pregnancies, chronic diseases (including low back pain), occupation and working hours. The weekly questionnaire contains questions about absence, physical workloads, and pregnancy discomforts. Questionnaires are sent by SMS via REDCap, a secure web application for data collection licensed for use in the Capital Region. Exposure to occupational standing/walking and bending are measured by use of triaxial accelerometers (Axivity sensors) worn on the back and thigh for 24 hours/day in optimally seven days. (22-25) The validated program Acti4 is used to analyze the data.(26) Exposure to lifting is quantified through continuous observations during a workday with registration of number of lifts, weight of burdens, and number of person-liftings. Observations are made by the PhD students and student assistants after completing a 2-hour course to minimize inter-observer differences. All measurements and observations among pregnant women will take place during the 2nd and 3rd trimester while performing their usual work. WP4 Studies on exposure to physical workloads and pregnancy discomforts using real-time self-report (01/09/2023-01/10/2024): Each PhD project will conduct a cohort study investigating associations between exposure to physical workloads and pregnancy discomforts. The PhD project at hand will focus on exposure to standing/walking and bending and self-reported pregnancy discomforts. Methods Standing/walking (hours/day) and bending (min/day) are measured by weekly self-report. To avoid reverse causation and ensure correct timing only exposures until the week before outcome are used. Exposure is measured as Immidiate exposure in the week preceding outcome: Average time with standing/walking work (hours/day) and bending more than 30 degrees (min/day) and Long term exposure measured as number of weeks with: Average standing/walking work \>4 hours/day and bending ≥30 degrees \>1 hour/day. Pregnancy discomforts include pelvic problems, low back pain, and Braxton Hick contractions are measured by weekly self-reports. To account for repeated measurements on participants mixed logistic regression adjusted for relevant confounders is used. A problem in former studies is the potential healthy worker effect, which we in this study adress by G-estimation.(27) With an expected 3000 pregnancies, of which around 1/3 is exposed,(7,28) a prevalence of pregnancy discomforts of 14-53%(15,28) and a desired statistical power of 0.8, it will be possible to establish a 16% increased risk of pregnancy discomforts among exposed pregnancies. WP5 Construction of trimester-specific Job exposure matrices (JEMs) and studies on exposure to physical workloads and negative pregnancy outcomes (01/08/2024-01/08/2025): Two trimester specific JEMs (LIFTING-JEM AND POSITION-JEM) will be constructed from the collected measurement data in WP3 and validated and described in a paper. Further each PhD project will conduct a cohort study of associations between combined exposure to physical workloads and negative pregnancy outcomes. The PhD project at hand will focus on a POSITION-JEM and investigate risk of preterm birth. Methods A trimester specific POSITION-JEM is constructed as a cross-tabulation of jobs defined by DISCO-88/08 codes and exposure to standing/walking and forward bending for each trimester using non-pregnant (1st trimester) and pregnant (2nd and 3rd trimester) levels. Exposures are calculated as group averages within jobs (mean/SD) by BLUPS in linear mixed modelling for cumulative duration of standing/walking and forward bending at work. JEMs are validated by looking at variance and contrast between jobs. Sufficient contrast between occupational groups is crucial and analyses will only be performed to the extent that such a contrast can be demonstrated. The variation within jobs is not taken into account in a JEM, which results misclassification of individual exposure. However, use of group-based exposure results in Berkson error rather than classical error in the statistical analysis, and as such do not attenuate the risk measure.(29) Both trimester-specific JEMs will be applied in the nationwide occupational generation cohort DOC\*X-G, which links information on larbour market attachment and pregnancies in the entire Danish Workforce 1976-2018. Cox regression with time dependent exposure is used to investigate associations between immediate (preceding week) and longterm (cumulated) exposure during pregnancy and the risk of preterm birth. Preterm birth (birth in weeks 23-36) is calculated by gestational age at birth from DOC\*X-G in 2000-2019 (roughly 1,000,000 births of which 7% are expected to be preterm).(30) The analyses will take into account possible interactions between exposures and absence before maternity leave . WP6 Writing of PhD.dissertations, completion of the project and dissemination of results (01/06/2025-28/02/2026).