Objectives Occupational physical activity (OPA) during pregnancy has been linked to adverse pregnancy outcomes, but crude exposure assessment remains an issue in causal inference. We aimed to develop a quantitative trimester-specific job exposure matrix (JEM) for standing, walking, and forward bending among pregnant workers.

Methods Accelerometer measurements from 403 female workers across 109 DISCO-08 job codes were obtained in Denmark between January 2023 and June 2024. Full workdays were measured during two weeks among pregnant workers and one week among non-pregnant workers. We used linear mixed-effects models to estimate exposure levels of occupational standing, walking, and forward bending for all 1171 DISCO-08 codes, including age, trimester, and expert ratings as fixed effects, and job codes and workers as random effects.

Results The between-job variances relative to total variances were 56% for standing, 51% for walking, and 45% for forward bending. The fixed effect trimester reduced standing time by 0.38 hours during the 3rd trimester compared to non-pregnant participants, whereas no differences were observed for walking or forward bending. Based on the trimester-specific JEM for occupational standing time, bakers had the highest exposure (range from non-pregnant to 3rd trimester, 5.41–5.03 hours/workday). For walking and forward bending, the highest exposed jobs from the pregnancy-specific JEM were waiters (1.76 hours/workday) and livestock/dairy producers (1.24 hours/workday), respectively.

Conclusions The JEM enhances independent objective exposure assessment in epidemiological studies of OPA and pregnancy outcomes and may advance guidelines and potentially prevent adverse pregnancy outcomes.

Key terms adverse pregnancy outcome; exposure assessment; forward bending; JEM; job exposure matrix; linear mixed-effects model; occupational physical activity; PRECISE JEM; pregnancy; pregnant; standing; walking

• Supplementary material