Introduction: Evidence from Japan on how psychological distress changes with time since spousal loss in the general population is limited. Prior reports rarely index years since bereavement or separate early from later stages.
Purpose: To describe married-referenced differences in psychological distress among widowed older adults in Japan and to examine gradients by time since bereavement, with attention to socioeconomic and health correlates.
Method: We conducted a secondary analysis of the National Survey of the Middle-aged and Older (Wave 1, 2010; SSJDA). Eligible respondents were aged 55-84 with complete data (N=2,750). Psychological distress was assessed using a five-item form of the Kessler scale (K6); scores were rescaled by 1.2 to the conventional 0-24 metric. We summarized crude proportions meeting thresholds for moderate (K6≥5) and serious (K6≥13) distress. Analyses focused on widowed versus continuously married respondents, stratified by sex and bereavement duration in four bins: <3 years, 3-<6 years, 6-<10 years, and ≥10 years. Married-referenced contrasts were reported as percentages, percentage-point differences, and prevalence ratios, alongside age-stratified summaries.
Results: Compared with the continuously married, widowed adults showed higher proportions of moderate distress. Within widowhood, elevation was most pronounced in the early period (<3 years) and remained raised at 3-<6 years, with attenuation by 6-<10 years and ≥10 years. A small subset met the serious-distress threshold, concentrated nearer to the loss. Among men, the early spike appeared sharper with a steeper subsequent decline; among women, levels remained comparatively elevated even ≥10 years after bereavement. Widowed respondents were more frequently in lower-income and non-employment categories and reported poorer self-rated health.
Conclusions: Distress following spousal loss shows a time-sensitive elevation that partly diminishes over years, with sex-specific trajectories and socioeconomic patterning. These descriptive findings provide a baseline for preregistered, minimally adjusted estimation in subsequent work.