Generalized anxiety disorder in older adults: Prevalence, associated factors, and treatment outcomes

Generalized anxiety disorder (GAD) occurs commonly in older adults and is associated with several negative outcomes, including personal distress, disability burden and increased use of health services. The main goal of this thesis was to provide a comprehensive picture of the prevalence, associated factors and treatment outcomes of GAD in community-dwelling older adults, starting at a macro epidemiologic level and finishing on the micro level of neuropsychological functioning. In order to achieve this goal, two population-based datasets were accessed. The National Survey of Mental Health and Well Being (NSMHW2) is a representative survey of 8,841 Australians aged 16-85 years conducted by the Australian Bureau of Statistics (ABS) in 2007. The Longitudinal Assessment of Women (LAW) study was undertaken to study health determinants in middle-aged and older women in Brisbane, Australia, and at inception, in 2001, had 511 participants. This thesis is structured around seven chapters. The first chapter is introductory and presents a general introduction to GAD and summarises the literature on GAD in older adults. Chapter 1 also provides an outline of the samples and goals for the thesis. The second chapter addresses the prevalence and associated factors of current GAD in community-dwelling Australians, analysing data from the NSMHW2. Data from participants aged between 55 and 85 years (N=3,305, 54% female) were considered for the analyses, which applied survey data routines. Eighty-four participants fulfilled the criteria for 12-month GAD, with a weighted prevalence of 2.8% (95% CI: 2.0, 3.7). Multivariate logistic regression analyses indicated that older age, functional limitations, comorbid depression, and family history of anxiety or depression were the factors most significantly associated with 12-month GAD. The third and fourth chapters discuss nosological and phenomenological dimensions of GAD, also using the NSMHW2 dataset. Chapter 3 presents a study about the age at onset of GAD. Its goal was to examine the impact of having experienced a first GAD episode earlier in life, for community-dwelling adults aged 55 years and older who had a lifetime diagnosis of GAD (N=228, 65% female). Early onset was defined through a data driven approach, using the median age at onset for the entire NSMHW2 cohort, and groups were compared using χ², t-test, and logistic regression procedures. Physical abuse during childhood, lifetime diagnosis of dysthymia, and number of GAD episodes were all associated with increased odds of being diagnosed with GAD earlier in life, after adjusting for current age and 12-month GAD. Chapter 4 elaborates upon worry as a central dimension of GAD, after which a detailed analysis of lifespan self-reported worry is presented, using all participants that identified themselves as worriers (N=3,735, 16-85 years, 61% female). Findings indicated an overall decrease in worry count with advancing age, as well as a developmental distribution of worry content and a quantitative but not qualitative distinction between normal and pathological worriers. Compared with younger adults and after adjusting for sociodemographic and clinical factors, older adults reported fewer worries and a lower likelihood of worrying about interpersonal relations, health, work, and miscellaneous topics, but a higher likelihood of worrying about the health and welfare of loved ones. Chapter 5 addresses the association between GAD and cognitive functioning in the LAW study cohort. Previous findings regarding the impact of anxiety symptoms and GAD on cognitive functioning are reviewed, followed by a study where these hypotheses are investigated. Results obtained on a neuropsychological battery were examined according to participants’ GAD status (current GAD, n=9; worrier but not GAD, n=9; non worrier, n=27, with participants from the non-GAD groups matched for age (+/- 2 years) and education with participants diagnosed with GAD). Those diagnosed with current GAD scored significantly worse on tasks thought to activate the pre-frontal cortex, namely inhibition and verbal fluency tasks. The sixth chapter provides a detailed perspective on current interventions for late-life GAD, both at a qualitative and quantitative level. Included trials were assessed according to their intervention type (pharmacological or psychotherapeutic). Random effects analyses were used to obtain the odds of response for each trial as well as the pooled results, and heterogeneity was addressed by the Q and I² statistics. Twenty-seven trials (14 pharmacological, 13 psychotherapeutic) fulfilled the inclusion criteria, reporting results from 2,373 baseline participants. Pooled treatment effects for pharmacological and psychotherapeutic trials were similar, with findings favouring active interventions over control conditions. For psychotherapeutic trials there were no significant results when an active condition (e.g., discussion group) was used in the control arm. The seventh chapter brings together these findings, providing an integrated perspective of this project and the implications both for research and clinical practice. First, understanding the characteristics associated with late life GAD promotes the identification of those at a higher risk and who could benefit more from tailored prevention and intervention programs. In addition, older adults seem to benefit from non-specific models of treatment, such as discussion groups, as much as from formal interventions such as cognitive behaviour therapy. This observation might have implications for the development of intervention strategies more suited to the needs of older people. Findings regarding the worry pattern make us question the extent to which the existing diagnostic criteria apply throughout the lifespan and whether age-related specifications should be introduced. Furthermore, these results suggest that the combined use of clinical and neuropsychological information might result in a more comprehensive assessment, eventually providing additional guidelines for the diagnosis of GAD. Strengths and limitations are also discussed, along with suggestions for future research