ABSTRACT
BACKGROUND: The associations of doses of vaccine received with symptomatic infection with SARS-CoV-2 and negative conversion rate of viral RNA by BMI, diabetes, and age are unclear. METHODS: Included were adult cases of SARS-CoV-2 infection hospitalized at a makeshift hospital in Shanghai (N = 38,592). Each case received a real-time reverse transcriptase-polymerase chain reaction (RT-PCR) test every day until discharge. Symptomatic cases had ≥1 pre-specified symptoms. Negative conversion time (NCT) was the duration between the specimen collection date associated with the first positive RT-PCR test and the first test date of the two consecutive negative tests at least 24 h apart. BMI-, diabetes- and age-stratified multivariable logistic and Poisson regressions were applied. FINDINGS: Coexistence of overweight/obesity and diabetes was associated with a higher risk of symptomatic infection and a longer NCT compared with coexistence of normal weight and without diabetes, but this association was primarily attributed to underlying comorbidities. Compared with absence of vaccination, booster vaccination, but not full vaccination, was consistently associated with a 42 %-56 % lower odds of symptomatic infection and â¼1.6-1.8 days of shorter NCT across different strata separately for weight and diabetes. Age-stratified analyses found that the effectiveness of booster vaccination did not attenuate with age, except for preventing symptomatic infection among adults with diabetes. INTERPRETATION: BMI and diabetes co-determined their associations with symptomatic infection and NCT. Booster vaccination but not full vaccination was associated a lower risk of symptomatic infection, a shorter NCT or both regardless of BMI, diabetes status and age.