Background
Chronic physical conditions and mental disorders frequently co-occur and are independently associated with increased risk of suicidal behavior. However epidemiological studies often focus on conditions separately and predominantly based on cross-sectional relationships. There is limited evidence that investigates the longitudinal development of mental-chronic illness comorbidity structure at the population level. This study investigates time trend patterns of comorbidity between mental health measures and chronic physical illnesses in Australia.
Methods
The data were obtained in waves 9, 13, 17, and 21 of the Household, Income and Labour Dynamics in Australia (HILDA) Survey. A balanced longitudinal panel was created by retaining individuals who participated in all four waves (6,340 per wave), with missing values removed before analysis. Chronic physical conditions included arthritis or osteoporosis, asthma, cancer, chronic bronchitis or emphysema, type-1 and type-2 diabetes, high blood pressure, heart disease, and other serious circulatory conditions. Mental health measures included diagnosed depression or anxiety, other mental illness, psychological distress, and selected emotional symptoms. For each wave, regularized Ising models using logistic regression with L1 (eLASSO) penalization were estimated to identify non-zero conditional associations. Edge strength was defined as the absolute value of the regularized coefficient, and networks were compared across waves to identify persistent associations.
Results
Across the four waves, the extent of non-zero mental-chronic associations increased by 44 to 59, and it can be stated that there was a progressive growth in cross-domain connections during the 12-year period. Mean edge strength decreased between Wave 9 (0.275) to Wave 21 (0.193), whereas the maximum edge weights were still high (0.890-0.994), indicating that whereas additional associations were formed, the central tendency of individual ties weakened. There were several cross-domain links that were present throughout the waves. Both depression or anxiety and psychological distress were associated with cardiovascular, respiratory and metabolic conditions (high blood pressure, heart disease, chronic bronchitis or emphysema, type-2 diabetes and arthritis or osteoporosis). The presence of these stable associations implies that mental and chronic physical conditions can be structurally interrelated in the population, with the central nodes of the longitudinal comorbidity network being the state of depression and distress.
Conclusion
Mental and chronic physical conditions form stable yet progressively expanding comorbidity networks over time. Depression and psychological distress consistently emerged as central cross-domain connectors. These findings highlight the need for integrated, longitudinal approaches to comorbidity, supporting coordinated prevention and care strategies that simultaneously address mental and physical health in population settings.