Determinants of weight, psychological status, food contemplation and lifestyle changes in patients with obesity during the COVID-19 lockdown: a nationwide survey using multiple correspondence analysis

The study investigates how the COVID-19 lockdown in Italy (March 11–May 18, 2020) affected weight control, lifestyles, and psychological wellbeing among patients with obesity followed by Obesity Centers (OCs). The context includes widespread restrictions (work-from-home mandates, closure of public spaces) and known associations between obesity and worse COVID-19 outcomes. The authors hypothesized that lockdown restrictions negatively affected weight control in PWO and aimed to: (1) identify weight and lifestyle changes during lockdown; (2) compare mental and physical wellbeing during lockdown versus prepandemic; (3) identify determinants of weight change, lifestyle modifications, and psychological distress; and (4) assess the impact of maintaining communication with OCs via telemedicine on weight, lifestyle, and wellbeing.

Background evidence cited includes: obesity’s association with more severe COVID-19 outcomes (higher hospitalization, ICU care, mortality); lockdown-related lifestyle disruptions (reduced physical activity, social isolation) and psychological distress (fear, anxiety, depression) linked to increased food intake and weight gain; similar reports from other countries under restrictions; and the Italian obesity care structure (primary care, specialist OCs, bariatric/rehabilitation). Prior small studies in PWO suggested that loneliness and working from home predicted weight gain via emotional eating and reduced PA; increased anxiety/depression and stress eating hindered weight loss goals. Education level was not a consistent predictor of weight change in prior work.

Design: Multicenter prospective survey among PWO whose outpatient follow-up visits at Italian ADI network OCs were postponed due to lockdown. Twenty-six centers participated. Ethics: Approved by the Ethics Committee of the Local Health Authority of Brindisi; informed consent obtained. Data collection: Structured 77-item questionnaire administered by trained personnel via phone or completed online between May 2 and June 25, 2020 (including two weeks post-lockdown to allow completion). Non-responders were re-contacted once. Variables were mostly categorical; age, height, and weight were numeric. Participants were instructed to weigh themselves in the morning in light clothes; pre-lockdown comparison weight was taken from last clinic measurement, or last self-report immediately before lockdown if intervening changes occurred. BMI categories: overweight (25.0–29.9), obesity class I (30.0–34.9), class II (35.0–39.9), class III (≥40.0) kg/m². Questionnaire sections analyzed: (1) demographics (sex, age, education, employment), region COVID-19 incidence tier, anthropometrics, work type; (2) work changes during lockdown, emotional difficulties (type and intensity), sleep modifications, psychophysical wellbeing, body dissatisfaction; (3) physical activity (changes, location), being on diet therapy and adherence, weight; (4) telemedicine contact with OC (phone, video, email, social media), willingness to use obesity medications, and perceived importance of obesity. Dietary pattern and cooking details were collected but not analyzed in this paper. Regional COVID-19 exposure: Regions classified by cumulative standardized daily incidence/100,000 by end of June 2020: high (>1,000), medium (200–999), low (<200). Statistical analysis: Descriptive statistics (means ± SD, medians, frequencies); group comparisons via ANOVA or chi-square. Multiple correspondence analysis (MCA) was used to reduce high-dimensional categorical data to two dimensions (~80% variance) and visualize associations among categories (weight change, working at home, confinement, anxiety level, job status during lockdown, emotional difficulties and types). Emphasis on socioeconomic variables. Multiple linear regression modeled weight difference (post-lockdown minus pre-lockdown weight). Independent variables were categorical, chosen for substantive contribution and variance explained; multicollinearity assessed via variance inflation factor (<2 for all). Age and sex included as covariates.

Sample and baseline: 1,232 respondents from 26 OCs; 72% female; mean age ~50 years; mean BMI ~34.7 kg/m²; 41% obesity class II–III. Over half lived in southern Italy; education predominantly middle (29%) or high school (41%); main work statuses: employed (37%), homemakers (16%), retired (15%). Weight change during lockdown: 48.8% gained weight, 27.1% lost weight, 24.1% maintained weight. Mean overall weight change was +2.3 ± 4.8 kg. Among gainers, mean gain was +4.0 ± 2.4 kg (about +4.2% ± 5.4%). No significant differences in weight change by age, sex, education, work-from-home status, or regional COVID-19 incidence tier. Lifestyle and psychological changes: About 37% reported increased emotional difficulties (predominantly fear and dissatisfaction). 61% reduced physical activity; 55% reported changes in sleep quality/quantity (e.g., insomnia or early awakening). Body dissatisfaction reported by ~80%. Correlates of weight gain (p < 0.001 across comparisons): Decreased psychophysical wellbeing (69.6% among gainers), increased emotional difficulties (62%), sleep changes (54%), reduced PA (56%), and difficulty following diet therapy (68%) were more prevalent among weight gainers than those with stable or reduced weight. Those experiencing sleep changes and reduced PA were disproportionately weight gainers. Telemedicine: 62.5% maintained contact with OCs via phone/video/email/social media. Lack of contact strongly correlated with weight gain (57.8% gainers vs 13.9% losers; p < 0.001). Participants willing to use anti-obesity medications were more commonly weight gainers (68.4%; p < 0.001). MCA clusters: Two clusters identified. Cluster 1 (unchanged/improved lifestyle) associated with artisans/traders/farmers, private and public employees, and retirees; linked to unchanged/decreased weight and absence of negative emotions (anger, depression, boredom, dissatisfaction), normal anxiety, and unchanged emotional difficulties. Cluster 2 (worsened lifestyle) associated with unemployment and working from home; linked to depression, boredom, dissatisfaction, weight gain, greater food contemplation, eating more, confinement at home, and interest in obesity medications. Homemakers appeared as a subcluster: weight gain and anger related to confinement. Regression (age- and sex-adjusted; directions): Improved psychophysical wellbeing associated with weight loss; worsened wellbeing associated with weight gain. Increased PA associated with weight loss; PA at home vs reduction also associated with less gain. Increased “thought of food” and greater value attributed to food associated with weight gain; decreased thought of food associated with weight loss. Maintaining contact with the OC associated with lower weight gain, while having no referral OC associated with higher gain. Age and sex were not associated with weight change.

The findings support the hypothesis that COVID-19 lockdown restrictions adversely affected weight control in many PWO and that work status and psychosocial factors were key determinants. Employment conditions influenced outcomes: unemployed, homemakers, and those working from home were more likely to cluster with negative emotional states, increased food preoccupation, reduced PA, and weight gain, whereas public/private employees and retirees tended to maintain or improve lifestyle and weight. The strong association between lack of telemedicine contact and weight gain underscores the importance of continued clinical engagement during disruptions. Psychological distress (fear, dissatisfaction, depression, boredom) and sleep disturbances were closely linked to weight gain, suggesting that emotional dysregulation mediates adverse lifestyle changes. These results align with prior reports of increased anxiety/depression and stress eating in PWO during stay-at-home orders and reinforce the need for integrated behavioral support alongside dietary and PA counseling. Telemedicine appears to mitigate weight gain risk and may be particularly important for socially vulnerable PWO or those with emotional dysregulation, enabling monitoring, coaching, and timely interventions.

Among Italian PWO followed by OCs, nearly half gained weight during the COVID-19 lockdown. Work status (unemployed, homemaker, working from home), emotional difficulties, reduced PA, sleep disturbances, increased food contemplation, and lack of telemedicine contact with OCs were key determinants of weight gain. Sustained access to care via telemedicine should be prioritized during service disruptions to support weight management and wellbeing. Clinical practice should target groups at higher risk with tailored remote interventions, potentially including pharmacotherapy, stress management/mindfulness, and lifestyle coaching. Future research should evaluate similar determinants in other countries and contexts (e.g., economic recessions, seasonal constraints), and further examine the roles of body dissatisfaction and bariatric surgery status in weight trajectories during critical periods.

Potential selection bias: 46% of contacted patients declined participation, limiting representativeness of PWO followed at Italian OCs. Analyses were not adjusted for interview timing/date, which may have attenuated some effect sizes. Weight was self-measured at home for the post-lockdown assessment, which may introduce measurement variability. Dietary pattern and cooking habit data were collected but not analyzed here.