Prepandemic Risk Factors for Disabling Long COVID: A Prospective Cohort Analysis

Disabling long COVID, characterised by persistent symptoms that limit daily functioning, has emerged as an important public health concern. However, prospective evidence on predisposing risk factors remains limited.

This study used prospective data from the UK Household Longitudinal Study, linking prepandemic baseline information collected in Wave 10 (2018-19) with follow-up data from Wave 14 (2022-23). The analytic sample comprised 12,033 adults aged ≥ 16 years who participated in both waves and self-reported a positive COVID-19 test at follow-up. The primary outcome, disabling long COVID, was defined as symptoms lasting more than 12 weeks that impaired day-to-day activities. Prepandemic sociodemographic, health and psychosocial factors assessed at baseline were included as predictors. Associations were estimated using modified Poisson regression with robust standard errors to calculate adjusted relative risks (RRs).

Disabling long COVID was reported by 690 individuals (5.7%). Higher risk was observed among women (RR 1.26; 95% CI 1.08-1.48) and adults aged 30-49 (RR 1.38; 95% CI 1.10-1.73) or 50-69 (RR 1.28; 95% CI 1.01-1.62) years, compared with those aged 16-29 years. Additional risk factors included pre-existing health conditions (RR 1.31; 95% CI 1.10-1.56), poor self-rated health (RR 1.78; 95% CI 1.40-2.25), psychological distress (RR 1.44; 95% CI 1.21-1.72) and poorer sleep quality (fairly bad: RR 1.92; 95% CI 1.45-2.56; very bad: RR 1.96; 95% CI 1.37-2.81), compared with very good sleep quality. Compared with non-White participants, White participants had lower risk (RR 0.75; 95% CI 0.61-0.92), while moderate (RR 0.76; 95% CI 0.62-0.93) and high (RR 0.81; 95% CI 0.67-0.98) income satisfaction, compared with low-income satisfaction, were protective. Stratified analyses showed that the effects of rural residence (p for interaction = 0.011) and income satisfaction (p = 0.009) differed significantly by sex, with weaker evidence for age (p = 0.095) and self-rated health (p = 0.061).

Prepandemic health, socioeconomic and psychological vulnerabilities were independently associated with disabling long COVID, with distinct sex-specific patterns of risk.