Low Dietary Variety Is Associated with Incident Frailty in Older Adults during the Coronavirus Disease 2019 Pandemic: A Prospective Cohort Study in Japan

**Background:** The COVID-19 pandemic led to social restrictions that may have reduced motivation to maintain a healthy diet among older adults. Dietary variety—intake of diverse food groups—is known to prevent frailty, but its relationship with frailty during prolonged pandemic restrictions had not been studied. This one-year follow-up study examined the association between dietary variety and frailty in community-dwelling older Japanese adults during the pandemic.

**Methods:** This prospective cohort study was conducted in Japan. Baseline surveys were mailed in August 2020 to 4996 randomly selected community-dwelling adults aged ≥65 years. Follow-up surveys were mailed in August 2021 to 1635 respondents. After excluding 170 subjects who were frail at baseline and those with missing data, 1008 subjects were analyzed. Dietary variety was assessed using a 10-food-group questionnaire (meats, fish/shellfish, eggs, soybeans/soy products, milk/dairy, seaweeds, vegetables, fruits, potatoes, oils). Each food group consumed "almost every day" scored 1 point (range 0–10). The cumulative mean of baseline and follow-up scores was used. Scores ≤3, >3 to <6, and ≥6 indicated low, mid, and high dietary variety, respectively. Frailty was assessed using a 5-item screening tool (weight loss, slowed walking, reduced walking frequency, memory complaints, fatigue). Scores ≥3 indicated frailty. Logistic and linear regression analyses adjusted for sex, age, BMI, smoking, alcohol use, living alone, and chronic conditions (hypertension, diabetes, hyperlipidemia, stroke, cardiac disease).

**Key Results:** Of 1008 subjects (mean age 73.8 years in non-frail, 75.7 years in frail), 113 (11.2%) developed frailty over one year, corresponding to 112.1 cases per 1000 person-years. The mean dietary variety score was significantly lower in subjects who became frail compared to those who remained non-frail (p = 0.048). Linear regression showed a significant inverse association between dietary variety score and frailty score (β, −0.032; 95% CI, −0.064 to −0.001; p = 0.046). This association remained significant after adjusting for sex and age (β, −0.051; 95% CI, −0.083 to −0.019; p = 0.002) and in the fully adjusted model (β, −0.045; 95% CI, −0.078 to −0.012; p = 0.015). Sensitivity analysis excluding subjects with very low scores (<1) confirmed the association (β, −0.047; 95% CI, −0.081 to −0.013; p = 0.008). Logistic regression showed that compared to high dietary variety, low dietary variety was associated with increased odds of frailty in the age- and sex-adjusted model (OR, 1.911; 95% CI, 1.066–3.426; p = 0.030) and fully adjusted model (OR, 1.877; 95% CI, 1.034–3.409; p = 0.039). Among specific food groups, meat intake decreased significantly in frail subjects (p = 0.031), and milk/dairy intake decreased significantly in all subjects (p = 0.032). Over half of participants reported that COVID-19 hindered social activities and interactions with family and friends.

**Clinical Implications:** Low dietary variety during the COVID-19 pandemic was independently associated with increased frailty risk in community-dwelling older adults. The frailty incidence rate (11.2% over one year) was higher than pre-pandemic estimates (8.7% in Japan, 13.6% in meta-analyses), suggesting pandemic-related restrictions may accelerate frailty through reduced dietary diversity. The findings highlight the need for targeted dietary support for older adults during prolonged social restrictions. Limitations include a low follow-up rate (potential nonresponse bias), self-reported measures (possible overestimation of associations), lack of pre-pandemic dietary data, and a short one-year observation period. Further follow-up is needed as the pandemic persists.