Overweight, Obesity, and Depression in Multimorbid Older Adults: Prevalence, Diagnostic Agreement, and Associated Factors in Primary Care-Results from a Multicenter Observational Study

Abstract

Background/Objectives: Obesity and depression, in conjunction with multimorbidity, are interconnected conditions increasingly managed in general practitioner (GP) settings, yet these associations remain insufficiently studied in older patients. This study investigates the prevalence of depression across different body mass index (BMI) classes and includes age and gender differences in multimorbid older patients, offering a novel perspective on subgroup-specific patterns. Further the agreement between GP depression diagnoses and the Geriatric Depression Scale (GDS) is studied and patient-specific factors that may affect the agreement are explored, aiming to improve future diagnostics for vulnerable subgroups. Methods: Data were provided by the baseline assessment of the MultiCare Study, a prospective multicenter observational cohort of multimorbid patients aged 65+ years recruited from 158 GP practices across eight study centers in Germany. Data from 2568 study participants were analyzed based on GP-coded International Classification of Diseases (ICD) diagnoses, structured GP questionnaires, and patient questionnaires. Assessments included data on the BMI and depression (15 item version of the GDS). Agreement between GP diagnoses of depression and GDS assessment was measured using Cohen's kappa. Four logistic regression models were used to examine the effects of patient-specific factors on the agreement of depression diagnosis (match or mismatch). Results: GPs diagnosed depression in 17.3% of cases, compared to the detection of depressive symptoms in 12.4% of the patients by GDS (cut-off ≥ 6 points). The highest prevalence rates were observed in patients with obesity class III (25.0% by GP; 21.7% by GDS). Women were significantly more likely to receive a depression diagnosis by a GP across most BMI classes (except obesity classes II and III). The detection of depressive symptoms by GDS was significantly more prevalent in older multimorbid obese patients (≥75 years), except for patients with obesity class III. The overall agreement between GP diagnosis and GDS assessment was weak (κ = 0.156, p < 0.001). The highest agreement was found for people with obesity class III (κ = 0.256, p < 0.05). Factors associated with a True Positive depression diagnosis (match by both GDS and GP) were female gender (odds ratio (OR) = 1.83, p < 0.05), widowhood (OR = 2.43, p < 0.01), limited daily living skills (OR = 3.14, p < 0.001), and a higher level of education (OR = 2.48, p < 0.01). A significantly lower likelihood of a False Negative depression diagnosis was found for patients with obesity class III. Conclusions: This study highlights the significant prevalence of depression among multimorbid older adults across different BMI classes, particularly in those with obesity class III. The weak diagnostic agreement between GP diagnosis and GDS assessment suggests a need for improved diagnostic practices in primary care. Implementing standardized screening tools and fostering collaboration with mental health specialists could enhance the identification and management of depression in this vulnerable population.

Keywords: Geriatric Depression Scale; body mass index; depression; late life; multimorbidity; obesity; primary care.

Figure 1

Sample flowchart; total sample of the MultiCare baseline survey.

Figure 2

Agreement between GP diagnosis and GDS assessment: percentage distribution of True/False Positives and Negatives.

Figure 3

Agreement between GP diagnosis and GDS assessment: Cohen’s kappa with 95% CI as error bars.