- Email: [email protected]
Changes in Frailty Status and Risk of Depression: Results From the Progetto Veneto Anziani Longitudinal Study
Accepted Manuscript Title: Changes in Frailty Status and Risk of Depression: Results From the Pro.V.A. Longitudinal Study Author: Marina De Rui, Nicola Veronese, Caterina Trevisan, Sara Carraro, Linda Berton, Stefania Maggi, Sabina Zambon, Maria Chiara Corti, Giovannella Baggio, Brendon Stubbs, Egle Perissinotto, Gaetano Crepaldi, Enzo Manzato, Giuseppe Sergi PII: DOI: Reference:
S1064-7481(16)30296-2 http://dx.doi.org/doi: 10.1016/j.jagp.2016.11.003 AMGP 715
To appear in:
The American Journal of Geriatric Psychiatry
Received date: Revised date: Accepted date:
24-8-2016 3-11-2016 4-11-2016
Please cite this article as: Marina De Rui, Nicola Veronese, Caterina Trevisan, Sara Carraro, Linda Berton, Stefania Maggi, Sabina Zambon, Maria Chiara Corti, Giovannella Baggio, Brendon Stubbs, Egle Perissinotto, Gaetano Crepaldi, Enzo Manzato, Giuseppe Sergi, Changes in Frailty Status and Risk of Depression: Results From the Pro.V.A. Longitudinal Study, The American Journal of Geriatric Psychiatry (2016), http://dx.doi.org/doi: 10.1016/j.jagp.2016.11.003. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Word count: 3206
CHANGES IN FRAILTY STATUS AND RISK OF DEPRESSION: RESULTS FROM THE PRO.V.A. LONGITUDINAL STUDY Marina De Rui1, MD, Nicola Veronese1, MD, Caterina Trevisan1, MD, Sara Carraro1, MD, Linda Berton1, MD, Stefania Maggi2, MD, PhD, Sabina Zambon3, MD, Maria Chiara Corti4, MD, Giovannella Baggio5, MD, Brendon Stubbs6, MD, Egle Perissinotto7, ScD, Gaetano Crepaldi2, MD, PhD, Enzo Manzato1,2, MD, PhD, Giuseppe Sergi1, MD, PhD. 1
Department of Medicine - DIMED – Geriatrics Division, University of Padova, Italy.
2
National Research Council, Aging Branch, Institute of Neuroscience, Padova, Italy.
3
Department of Medicine DIMED, Clinica Medica I, University of Padova, Padova, Italy
4
Division of Health Care Planning and Evaluation of the Regione Veneto, Venice, Italy.
5
Internal Medicine Division, Azienda Ospedaliera, Padova, Italy.
6
Health Service and Population Research Department, Institute of Psychiatry, Psychology and
Neuroscience (IoPPN), King's College London, De Crespigny Park, London Box SE5 8AF, United Kingdom; Physiotherapy Department, South London and Maudsley NHS Foundation Trust, Denmark Hill, London SE5 8AZ, United Kingdom. 7
Department of Cardiac, Thoracic and Vascular Sciences, Unit of Biostatistics, Epidemiology
and Public Health, University of Padova, Italy.
Page 1 of 23
Corresponding Author
De Rui Marina
Clinica Geriatrica - Ospedale Giustinianeo
Via Giustiniani 2, 35128, Padova, Italy
Phone: +390498218492; Fax: +390498211218; e-mail: [email protected]
Conflicts of Interest and Source of Funding: The authors have no conflicts of interest to declare.
Funding source: This work was supported by the Fondazione Cassa di Risparmio di Padova e Rovigo; University of Padova; the Azienda Unità Locale Socio Sanitaria 15 and 18 of the Veneto Region; and a grant from the Veneto Regional Authority (Ricerca Sanitaria Finalizzata n.156/03) (data collection phase). The data analysis phase was also financed by a grant from the University of Padova (Population aging - economics, health, retirement and the welfare state POPA_EHR).
Page 2 of 23
ABSTRACT
Objective: Evaluate whether pre-frailty was associated with the risk of developing depression and if longitudinal changes in frailty status corresponded to changes in incident depression during follow up.
Design: population-based prospective cohort study conducted for 4.4 years.
Setting: two separate geographical areas near the city of Padua in the Veneto Region of Northern Italy. Participants: 891 non-depressed non-frail community dwelling Italian subjects aged ≥65 (46.6% men) belonging to the Progetto Veneto Anziani study.
Measurements: Depression was defined according to the Geriatric Depression Scale and was confirmed by geriatricians skilled in psychogeriatric medicine. Pre-frailty was defined by the presence of 1 or 2 among Fried criteria.
Results: The incidence rate of depression was 13.3% among subjects improving their frailty status at follow up (n=15), 15.0% in those who remained stable (n=79), and 26.7% among worsening participants (n=67) (p = 0.001). Pre-frailty at baseline did not predict the onset of depression (HR 0.82, 95% CI 0.55-1.21; Wald df=1; p=0.43), but a deterioration during follow up in at least one additional frailty criteria was associated with a significantly higher risk (HR 1.95, 95% CI 1.32-2.89; Wald df=2; p=0.01). Improvement in frailty status was not associated with the risk of incident depression (HR 0.71, 95% CI 0.35-1.42; Wald df=2; p=0.28).
Page 3 of 23
Conclusion: Our data did not offer evidence that pre-frailty per se predisposes to the onset of depression, but worsening in frailty status is associated with almost a twofold increased risk of incident depression, irrespective from the initial level of impairment.
Keywords: frailty, elderly, depression, pre-frailty, aging
Page 4 of 23
INTRODUCTION Frailty is a condition characterized by an increased vulnerability to stressors due to a reduction in the functional reserve of various organs/systems which become unable to withstand life events (1). The prevalence of frailty syndrome is reported to be between 10 and 20% among community-dwelling older adults (2).
According with Fried et al. (3), the core elements of the frailty phenotype are unintentional weight loss, weakness, exhaustion, slowness, and low physical activity. In the progression from a state of non-frailty, or robustness, to frailty there is frequently an intermediate condition called pre frailty and defined as the presence of one or two of the above criteria. Although the majority of older people gradually become frail, the dynamic nature of this pathway is also potentially reversible, especially in the earlier stages, (i.e. for the pre-frail condition) (4).
Frailty has been associated with an increased risk of several deleterious outcomes in older people, including disability, hospitalization, and institutionalization (5). More recently, a growing body of evidence has suggested that frailty could be considered as an independent risk factor for pathologies like cardiovascular diseases and diabetes (6,7).
Scientific evidence has for some time suggested that an interplay exists between frailty and depression, with some research finding that frailty predisposes to depression (8,9) and others reporting that depression accelerates the progression of frailty (6,10-14). Differently, the relationship between pre-frailty and depression is less clear, but, due to its reversibility, pre frailty may represent a potential window of opportunity to reduce the mental health impact.
Page 5 of 23
However, frailty is a dynamic process and subjects can move through frailty classes during lifetime. For instance, Gill et al. (15) found that in 18 months, 12% of pre-frail elders transitioned to robustness and 25% transitioned to frailty while 23% of frail elders transitioned to pre-frailty. So it would be interesting to understand if changes (worsening or improvement) in frailty items, influences the relationship with incident depression. However, this topic is yet to be adequately explored in the literature.
The aim of the present study was to evaluate whether pre frailty was associated with the risk of developing depression in a population-based sample of community-dwelling elderly subjects. In addition, we investigated whether changes in frailty status, expressed as the change in the number of frailty items during follow up, is associated with changes in depression among people with pre frailty.
METHODS
Study population
Data for this analysis were drawn from the Progetto Veneto Anziani (Pro.V.A.), an observational cohort study of an Italian population of older men and women living in two separate geographical areas near the city of Padua in the Veneto Region of Northern Italy. Sampling procedure, study design, and data collection method have been extensively described elsewhere (16). Briefly, the baseline Pro.V.A. study population consisted of 3,099 age- and sex-stratified, community-dwelling Caucasian participants aged ≥65 years (1854 women and 1245 men) randomly selected between 1995 and 1997 using a multistage stratification method. No exclusion
Page 6 of 23
criteria were used. Follow up assessments were conducted after 4 years (mean follow up 4.4±1.1 years). At baseline and at follow up, information was collected on participants’ formal education, living status (i.e. living alone or not), smoking and regular exercising habits during an in-person interview. Smoking habits were classified as “never/former” (for at least a year in the past) versus “current” smokers. Educational level (the total number of years of school attended) was dichotomized as ≤5 versus >5 years of schooling. Monthly income was recorded in the database in Italian liras and dichotomized using the cut-off of 500,000 Italian liras, which would correspond to about 800 Euros taking into account ISTAT monetary revaluation coefficients (17). Alcohol drinking was categorized as “yes” vs. “no” in the previous month. Participants were asked to report how many hours a week they had spent on practicing physical activity in the previous month; this information was dichotomized in <4 versus ≥4 hours/week. Disability was defined as the inability or need for assistance to perform one or more activities of daily living (ADL): bathing, dressing, eating, using the toilet, continence or transferring (18).
Body weight was measured on a calibrated balance scale. Height was measured with a stadiometer to the nearest 0.5 cm. Body mass index (BMI) was calculated as weight (in kilograms) divided by height (in meters) squared.
Any medical conditions were identified by board-certified physicians involved in the study, who examined all the clinical information collected on each participant, including clinical history, self-reported symptoms (using standardized questionnaires), medical and hospital records, blood tests, and a physical examination. The major diseases considered were any of the following: diabetes, cardiovascular diseases (CVD: congestive heart failure, angina and myocardial
Page 7 of 23
infarction, stroke, and peripheral artery disease), hip fracture, hand and lower limbs osteoarthritis, cancer, and vision and hearing impairment.
For the purpose of the present study, 635 participants were excluded because of cognitive impairment defined as a score in the Mini-Mental State Examination (MMSE) (19) below 26 and thus the Geriatric Depression Scale may have been unreliable (20). In addition, people with depression (n=1141), severe neurological diseases (i.e. people with terminal neurological disease such as multiple sclerosis or stroke) (n=43) or frailty (n=19) at baseline were not included. A further 370 people did not have complete follow up data. As shown in Figure 1, 891 participants constituted the final analytical sample for this study.
The present study was approved by the Human Studies Committee of Padua University and the Veneto Region’s Local Health Units (USSL) No.15 and No.18, and all subjects gave informed consent before their participation.
Definition of depression
Any presence of depressive symptoms was assessed at the baseline and at the follow-up with the Geriatric Depression Scale (GDS) (21), a 30-item self-reporting tool for identifying depression that has been extensively validated for use in the elderly. GDS scores range from 0 to 30, with a score of ≥11 being utilized as cut off for depression (21). The diagnosis of depression was confirmed by geriatricians skilled in psychogeriatric medicine using a standardized questionnaire checking also additional relevant information such as signs and symptoms, medical records, and medication use (16).
Page 8 of 23
Definition of pre-frailty
For the definition of pre-frailty and frailty we adopted a slightly modified version of the criteria set by Fried (3) as follows:
1. unintentional weight loss: self-reported weight loss above 5 kg over the past year not due to voluntary calorie restriction; 2. exhaustion: answering “No” to the question “Do you feel full of energy?” plus GDS score ≥ 10 (20); 3. low-energy expenditure: energy expenditure for leisure-time activities in an ideal week of the previous month below 383 kcal/week in males and 270 kcal/week in females (3,22); 4. weakness: handgrip strength below the sex and BMI cut-offs proposed by Fried3. Handgrip strength was measured in Kg using a JAMAR hand-held dynamometer (BK7498, Fred Sammons, Inc.). The best result obtained at three attempts at the dominant side was used for analyses; 5. slow gait speed: walking speed measured in meters/second below the sex and BMI cutoffs proposed by Fried (3). Gait speed was defined as the best performance achieved in two walks at usual pace along a 4-meter corridor. Participants were allowed to use canes or walkers.
Subjects unable to perform handgrip or the walking item were considered like having weakness or slow gait speed, respectively. At baseline, participants were classified as frail if they met 3 or more of the 5 criteria, as pre-frail if they met 1 or 2, and as robust if they met none of the criteria.
Page 9 of 23
However, using the inclusion/exclusion criteria cited above, no frail participants were included in the present study.
Statistical analysis
For the purpose of the study, the number of frailty items satisfied by each participant at baseline was compared with those recorded at follow up, calling this variable “change in frailty status”. This variable was then used to divide subjects in three subgroups: if the number of satisfied criteria at follow up was lower than at the baseline the subject was considered as “improved” (Group 1), if the number was the same the subject was considered as “stable” (Group 2), and if it was higher the subject was considered as “worsened” (Group 3). Participants’ characteristics were expressed as means ± standard deviations for continuous variables, and counts and percentages for categorical variables. For continuous variables, normal distributions were tested using the Shapiro-Wilk test. Age- and gender-adjusted p values were calculated as follows: for continuous variables the differences between the means of the covariates were analysed using a general linear model; for categorical variables logistic regression was applied. Bonferroni’s correction was applied in all comparisons. Cox’s proportional hazard models were used to assess associations between the three groups of subjects and incident depression by plotting the Schoenfeld residuals versus time. Known factors associated with frailty and/or depression were analysed as covariates in the univariate analysis. To explore whether a variable should be included as a predictor in the final model, the log-rank test of equality across strata was performed for all the categorical variables and Cox’s univariate
Page 10 of 23
proportional hazards regression for all the continuous variables. The predictors included in the final model were all the variables reaching a p<0.20 in the univariate analyses. Collinearity among covariates was quantified by the variance inflation factor (VIF) among the covariates, but no variable was excluded for this reason. Hazard ratios (HRs) and 95% confidence intervals (CIs) were used to compare depression rates across the tree groups of subjects taking Group 2 (=stable) as reference. We conducted several sensitivity analyses stratifying for clinical factors independently associated with depression in the final model to test the robustness of the results to check if pre-frailty can depend on any condition, but any of these factors appear to be relevant.
All analyses were performed using the SPSS 23.0 for Windows (SPSS Inc., Chicago, Illinois). All statistical tests were two-tailed and statistical significance was assumed for a p-value <0.05
RESULTS Baseline characteristics
Among 891 subjects included in the study, 44.0% were pre-frail and the remaining were robust. Due to the inclusion criteria, none of the participants were frail or depressed at baseline. Over the course of the study, 59% (n = 527) maintained their frailty status, whilst 13% improved (n = 113) and 28% (n = 251) worsened in at least one frailty item. Table 1 shows the baseline characteristics of the subjects divided in three groups by the change in the number of frailty items during follow up. Subjects who worsened their frailty status at follow up (Group 3), were significantly older than the other participants at baseline. No significant differences were observed among the three groups regarding general and anthropometric characteristics
Page 11 of 23
(educational level, income, physical activity, smoking, alcohol drinking, disability, BMI and living status).
At baseline the GDS score was similar in the three groups (6.7±2.6 in Group 1, 6.3±2.1 in Group 2, and 6.6±2.1 in Group 3, df=1206; p = 0.06).
Regarding comorbidities, the prevalence of diabetes, CVD, visual and hearing impairment was significantly higher in Group 3 subjects compared to the others.
Follow-up data
During the follow up,167 new cases (70 men and 97 women) of depression were recorded. Across the three groups, the cumulative incidence rate of depression was 13.3% among subjects improving their frailty status at follow up (Group 1), 15.0% in the stable group (Group 2), and 26.7% among those with worsening frailty status (Group 3) (age- and sex-adjusted p = 0.001). Using Cox’s regression analysis, and adjusting for baseline potential confounders, the presence of pre-frailty at baseline did not predict the onset of depression (HR 0.82, 95% CI 0.55-1.21; Wald df=1; p=0.43). Taking subjects who remained stable at follow up as reference, after adjusting for sex and age (Figure 2a) and other potential confounders (including pre-frailty) (Figure 2b), improvement in frailty status were not significantly associated with the risk of incident depression (HR 0.71, 95% CI 0.35-1.42; Wald df=2; p=0.28). On the contrary, worsening in at least 1 additional frailty criteria was associated with a significantly higher risk of incident depression (HR 1.95, 95% CI 1.32-2.89; Wald df=2; p=0.01).
Page 12 of 23
DISCUSSION
The present study utilizing a large sample of community-dwelling Caucasian elderly subjects, offered no evidence that, contrary to our hypothesis, pre-frailty per se did not predict the onset of depression over a 4.4-year follow up. Conversely, worsening in frailty status from baseline influenced the risk of incident depression. In particular, participants who accumulated during the observation period at least one more frailty criterion, had almost a two-fold increased risk of depression, also independently from their frailty status at baseline.
The main strength of the present study lies in the fact that we focused on robust and pre-frail individuals, in which the risk of developing depression in the long-term has not been adequately explored. Whilst there is a number of cross-sectional studies in the literature suggesting that frailty and depression are interrelated (2,8,9,11,23), due to the fact they are cross-sectional, it is not possible to disentangle the directionality of this association. Moreover, considering that negative depressive symptoms can overlap with frailty criteria, results from cross-sectional studies can be difficult to interpret with confidence. We chose to study the risk of incident depression in not yet frail subjects mainly for three reasons: first, because the study model would be clearer if only non-depressed and not yet frail individuals were selected. Second, because, being robust and pre-frail are two key categories and there is minimal data considering the interplay between the two. Finally, pre-frailty is potentially reversible and may have important clinical preventive implication.
Page 13 of 23
Regarding the relation between pre-frailty and depression, to the best of our knowledge, only two longitudinal studies (8,9) are available. Our findings are in line with those of Makizako et al. (9) who did not find a significant association between the two conditions in 3025 elderly Japanese people over a 15-month follow up after adjusting for potential confounders (OR 1.38, 95% CI 0.93-2.05). Conversely, Feng et al. (8) in a sample of 1827 middle-aged and elderly Chinese persons found a significant association (OR 2.26, 95% CI 1.12-4.57) between pre-frailty and depression over a 4-year follow up. A possible explanation for these differences could be that Feng et al. (8) did not include MMSE score among the exclusion criteria whereas Mikizako et al. (9) excluded subjects with a MMSE score <18/30 as well as we excluded all participants with a MMSE score < 24/30. Including people with a possible diagnosis of dementia poses diagnostic uncertainties since that in older adults depressive symptoms may mimicry cognitive deficits (24). Moreover, scales for the evaluation of depression in the elderly as the Geriatric Depression Scale do not maintain their validity in populations that contain patients with dementia (20).
As regards the change in frailty status, we observed that the majority of participants remained stable whereas 41% experienced a change from baseline frailty status. Among subjects that changed their status, about a third of them experienced an improvement, and the remaining two thirds deteriorated. The proportion of worsening subjects is higher compared to that reported by Lee et al. (4) but they included also frail subjects, who clearly cannot worsen further.
Regarding the relation between changes of frailty status and incidence of depression, to our knowledge, in the literature studies are lacking. Not necessarily accumulation of one or more frailty criteria meant that the subject became frail. Our findings seem to suggest that pre-frail elderly subjects are at higher risk for depression when they increase their level of frailty. In fact,
Page 14 of 23
subjects who worsened in terms of frailty, at baseline had a greater physical comorbidity, as evidenced by the higher prevalence of diabetes, cardiovascular diseases, osteoarthritis and sensory deficits. However, change in frailty status remained a predictor of incident depression even after adjustment for potential confounders, including comorbidities. In this sense, new onset depression could be a consequence of a negative modification of subject’s basal condition to which he/she had previously adapted. Subjective perception of physical worsening may suggest getting closer to death and impairment in motor functions favor social isolation. However, this finding does not exclude that accumulation and progression of chronic comorbidities may add an important burden to the onset of depression.
Considering that multimorbidity contributes and predisposes to the frailty phenotype, the practical implication of this finding could be that trying to improve or just preserving patients clinical and functional status may reduce the long-term risk of developing depression. And recalling the vicious circle in which comorbidity precipitates depression and depression predisposes to comorbidity and hence to disability, frailty and death, delaying or preventing worsening of frailty status, and treating depression where present, may reduce also health care costs and mortality.
The present study has a number of limitations. First we were unable to assess changes in the number of frailty items at shorter intervals than the 4.4-year follow up. Moreover we did not have any information if during their life these participants suffered from any form of depression. We were able to ascertain depression only at the time of evaluation, thus, it is possible that some cases of incident depression were relapses of a previously diagnosed depression. Second, unintentional weight loss was self-reported so it may be not so accurate. However, this criterion
Page 15 of 23
was the same as proposed by Fried et al. (3) in their original work. Another limit of the present study lies in that daily physical activity level was self-reported by participants during face-to face interviews, so the chances of an under- or over-reporting bias should be taken into account. Moreover, we had a number of people excluded at baseline which may lead to an underestimation of the incidence of depression in the general population. Third, both depression and frailty are conditions more frequent in older women than in men. Unfortunately, due to the limited sample size of our study, we were not able to stratify our analyses for gender. Future studies are so needed to detect specific gender differences. Finally, we used a slightly modified version of frailty index proposed by Fried et al. (3), and adapting the frailty criteria can influence the quality of the composite score and potentially introduce bias (25).
In conclusion our study confirmed that the dynamism of frailty pathway entails, at least in the early stages (i.e. pre-frailty), the possibility to reverse to a lower level of impairment. Our data did not offer evidence that pre-frailty per se predisposes to the onset of depression, but negative changes (i.e. worsening) in frailty level almost doubles the risk of incident depression, irrespective from the initial level of impairment. These findings point to the importance of improving or at least trying to preserve elderly people’s functionality in order to reduce comorbidities accumulation and functional deficits that would drag the subjects in an escalation of disease vulnerability, depression, worsening of frailty status and death. Future randomized clinical trials are needed to understand the clinical implication of treating frailty or pre-frailty in the prevention of new onset depression in the elderly.
ACKNOWLEDGEMENTS
Page 16 of 23
This work was supported by the Fondazione Cassa di Risparmio di Padova e Rovigo; University of Padova; the Azienda Unità Locale Socio Sanitaria 15 and 18 of the Veneto Region; a grant from the Veneto Regional Authority (Ricerca Sanitaria Finalizzata n.156/03), and a grant from the University of Padova (Population aging - economics, health, retirement and the welfare state - POPA_EHR).
Page 17 of 23
REFERENCES
1. Cigolle CT, Ofstedal MB, Tian Z, et al. Comparing models of frailty: the Health and Retirement Study. J Am Geriatr Soc 2009;57:830-9. 2. Buigues C, Padilla-Sánchez C, Garrido JF, et al. The relationship between depression and frailty syndrome: a systematic review. Aging Ment Health 2015;19:762-72. 3. Fried LP, Tangen CM, Walston J, et al.; Cardiovascular Health Study Collaborative Research Group. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci 2001;56:M146-56. 4. Lee JS, Auyeung TW, Leung J, et al. Transitions in frailty states among communityliving older adults and their associated factors. J Am Med Dir Assoc 2014;15(4):281-6. 5. Fugate Woods N, LaCroix AZ, Gray SL, et al. Frailty: Emergence and consequences in women aged 65 and older in the Women’s Health Initiative Observational Study. J Am Geriatr Soc 2005;53:1321e1330. 6. Sergi G, Veronese N, Fontana L, et al. Pre-frailty and risk of cardiovascular disease in elderly men and women: the Pro.V.A. study. J Am Coll Cardiol 2015;65:976-83. 7. Veronese N, Stubbs B, Fontana L, et al. Frailty Is Associated with an Increased Risk of Incident Type 2 Diabetes in the Elderly. J Am Med Dir Assoc 2016;17:902-7. 8. Feng L, Nyunt MS, Feng L, et al. Frailty predicts new and persistent depressive symptoms among community-dwelling older adults: findings from Singapore longitudinal aging study. J Am Med Dir Assoc 2014;15:76.e7-76.e12.
Page 18 of 23
9. Makizako H, Shimada H, Doi T, Yoshida D, et al. Physical frailty predicts incident depressive symptoms in elderly people: prospective findings from the Obu Study of Health Promotion for the Elderly. J Am Med Dir Assoc 2015;16:194-9. 10. Rovner BW, German PS, Brant LJ, et al. Depression and mortality in nursing homes. JAMA 1991;265:993-6. Erratum in: JAMA 1991;265:2672. 11. Penninx BW, Guralnik JM, Ferrucci L, et al. Depressive symptoms and physical decline in community-dwelling older persons.JAMA 1998;279:1720-6. 12. Whooley MA, Kip KE, Cauley JA, et al. Depression, falls, and risk of fracture in older women. Study of Osteoporotic Fractures Research Group. Arch Intern Med 1999;159:484-90. 13. Bruce ML. Depression and disability in late life: directions for future research. Am J Geriatr Psychiatry 2001;9:102-12. 14. Andersen K, Lolk A, Kragh-Sørensen P, et al. Depression and the risk of Alzheimer disease. Epidemiology2005;16:233-8. 15. Gill TM, Gahbauer EA, Allore HG, et al. Transitions between frailty states among communityliving older persons. Arch Intern Med. 2006; 166:418–423. 16. Corti MC, Guralnik JM, Sartori L, et al. The effect of cardiovascular and osteoarticular diseases on disability in older Italian men and women: rationale, design, and sample characteristics of the Progetto Veneto Anziani (PRO.V.A.) study. J Am Geriatr Soc 2002;50:1535-40. 17. http://www.istat.it/it/archivio/30440. Accessed April2, 2016. 18. Katz S, Ford A, Moskowitz R. The index of ADL: a standardized measure of biological and psychosocial function. JAMA 1963;185:914-9.
Page 19 of 23
19. Folstein MF, Folstein SE, McHugh PR. "Mini-mental state". A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189-98. 20. Burke WJ, Houston MJ, Boust SJ, et al. Use of the Geriatric Depression Scale in dementia of the Alzheimer type. J Am Geriatr Soc 1989; 37:856e860. 21. Yesavage JA, Brink TL, Rose TL, et al. Development and validation of a geriatric depression screening scale: a preliminary report. J Psychiatr Res 19821983;17:37-49. 22. Gary R. Evaluation of frailty in older adults with cardiovascular disease: incorporating physical performance measures. J Cardiovasc Nurs 2012;27:120-31. 23. Collard RM, Arts M, Comijs HC, et al. The role of frailty in the association between depression and somatic comorbidity: results from baseline data of an ongoing prospective cohort study. Int J Nurs Stud 2015;52:188-96. 24. Leyhe T, Reynolds CF 3rd, Melcher T, et al. A common challenge in older adults: Classification, overlap, and therapy of depression and dementia. Alzheimers Dement 2016; [Epub ahead of print] doi: 10.1016/j.jalz.2016.08.007. 25. Theou O, Cann L, Blodgett J, et al. Modifications to the frailty phenotype criteria: Systematic review of the current literature and investigation of 262 frailty phenotypes in the survey of health, ageing, and retirement in Europe. Ageing Res Rev 2015;21:78e94.
Page 20 of 23
FIGURE LEGENDS Figure 1. Flow chart of patients’ selection process.
Figure 2. Association between change in frailty status and risk of incident depression afteradjusting for sex and age (Figure 2a) and also for other potential confounders (Figure 2b).
Figure 2 footnotes: in Figure 2b, HR are adjusted for sex, age, monthly income, current smoking, GDS score, MMSE score, diabetes, cardiovascular diseases, osteoarthritis, and visual and hearing impairment.
Page 21 of 23
Table 1. Baseline characteristics of the sample divided by changes in frailty status from baseline to follow up. Patients’ characteristics
Group 1 - Improved
Group 2 - Stable
Group 3 - Worsened
p-value
(n = 113)
(n = 527)
(n = 251)
for overall comparison*†^
Age (years)
70.9±5.3
71.2±5.2
73.5±6.1
<0.0001**
Women (%)
69.0
49.9
53.8
0.001
Educational level > 5 years (%)
79.6
75.0
80.5
0.34
Monthly income ≥ 800 € (%)
61.1
52.0
57.4
0.18
Physical activity ≥4 h/week (%)
27.4
34.5
33.3
0.35
Current smoking (%)
6.2
12.0
12.0
0.19
Alcohol drinking (%)
72.6
74.6
72.5
0.79
Disability (%)
25.7
21.1
25.1
0.34
BMI (Kg/m2)
28.2±2
27.9±4.2
28.0±4.4
0.44
18.6
13.1
14.5
0.31
GDS score
6.7±2.6
6.3±2.1
6.6±2.1
0.06
MMSE score
26.9±2.0
27.4±1.8
27.2±1.8
0.04
General and anthropometric characteristics
Living alone (%) Cognitive functions
Comorbidities (%)
Page 22 of 23
Diabetes
11.5
10.4
17.9
0.02
CVD
13.3
9.7
18.3
0.004
Hip fracture
2.7
2.1
2.0
0.92
Hand osteoarthritis
13.3
11.4
15.9
0.21
Limbs osteoarthritis
19.5
16.1
23.1
0.06
Cancer
6.2
5.1
6.4
0.75
Visual impairment
28.3
24.9
36.7
0.003
Hearing impairment
56.6
63.0
73.3
0.002
Numbers are mean values ± standard deviations or percentages (%), as appropriate. GDS: Geriatric Depression Scale; MMSE: Mini-Mental State Examination; BMI: body mass index; CVD: cardiovascular diseases. *Unless otherwise specified, p values are adjusted for age and sex using a general linear model or logistic regression, as appropriate. †
The test with df=2
^df for the F-tests: df=2, 1206. **Univariate ANOVA (ANalysis Of Variance), not adjusted for age.
Page 23 of 23
S1064-7481(16)30296-2 http://dx.doi.org/doi: 10.1016/j.jagp.2016.11.003 AMGP 715
To appear in:
The American Journal of Geriatric Psychiatry
Received date: Revised date: Accepted date:
24-8-2016 3-11-2016 4-11-2016
Please cite this article as: Marina De Rui, Nicola Veronese, Caterina Trevisan, Sara Carraro, Linda Berton, Stefania Maggi, Sabina Zambon, Maria Chiara Corti, Giovannella Baggio, Brendon Stubbs, Egle Perissinotto, Gaetano Crepaldi, Enzo Manzato, Giuseppe Sergi, Changes in Frailty Status and Risk of Depression: Results From the Pro.V.A. Longitudinal Study, The American Journal of Geriatric Psychiatry (2016), http://dx.doi.org/doi: 10.1016/j.jagp.2016.11.003. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Word count: 3206
CHANGES IN FRAILTY STATUS AND RISK OF DEPRESSION: RESULTS FROM THE PRO.V.A. LONGITUDINAL STUDY Marina De Rui1, MD, Nicola Veronese1, MD, Caterina Trevisan1, MD, Sara Carraro1, MD, Linda Berton1, MD, Stefania Maggi2, MD, PhD, Sabina Zambon3, MD, Maria Chiara Corti4, MD, Giovannella Baggio5, MD, Brendon Stubbs6, MD, Egle Perissinotto7, ScD, Gaetano Crepaldi2, MD, PhD, Enzo Manzato1,2, MD, PhD, Giuseppe Sergi1, MD, PhD. 1
Department of Medicine - DIMED – Geriatrics Division, University of Padova, Italy.
2
National Research Council, Aging Branch, Institute of Neuroscience, Padova, Italy.
3
Department of Medicine DIMED, Clinica Medica I, University of Padova, Padova, Italy
4
Division of Health Care Planning and Evaluation of the Regione Veneto, Venice, Italy.
5
Internal Medicine Division, Azienda Ospedaliera, Padova, Italy.
6
Health Service and Population Research Department, Institute of Psychiatry, Psychology and
Neuroscience (IoPPN), King's College London, De Crespigny Park, London Box SE5 8AF, United Kingdom; Physiotherapy Department, South London and Maudsley NHS Foundation Trust, Denmark Hill, London SE5 8AZ, United Kingdom. 7
Department of Cardiac, Thoracic and Vascular Sciences, Unit of Biostatistics, Epidemiology
and Public Health, University of Padova, Italy.
Page 1 of 23
Corresponding Author
De Rui Marina
Clinica Geriatrica - Ospedale Giustinianeo
Via Giustiniani 2, 35128, Padova, Italy
Phone: +390498218492; Fax: +390498211218; e-mail: [email protected]
Conflicts of Interest and Source of Funding: The authors have no conflicts of interest to declare.
Funding source: This work was supported by the Fondazione Cassa di Risparmio di Padova e Rovigo; University of Padova; the Azienda Unità Locale Socio Sanitaria 15 and 18 of the Veneto Region; and a grant from the Veneto Regional Authority (Ricerca Sanitaria Finalizzata n.156/03) (data collection phase). The data analysis phase was also financed by a grant from the University of Padova (Population aging - economics, health, retirement and the welfare state POPA_EHR).
Page 2 of 23
ABSTRACT
Objective: Evaluate whether pre-frailty was associated with the risk of developing depression and if longitudinal changes in frailty status corresponded to changes in incident depression during follow up.
Design: population-based prospective cohort study conducted for 4.4 years.
Setting: two separate geographical areas near the city of Padua in the Veneto Region of Northern Italy. Participants: 891 non-depressed non-frail community dwelling Italian subjects aged ≥65 (46.6% men) belonging to the Progetto Veneto Anziani study.
Measurements: Depression was defined according to the Geriatric Depression Scale and was confirmed by geriatricians skilled in psychogeriatric medicine. Pre-frailty was defined by the presence of 1 or 2 among Fried criteria.
Results: The incidence rate of depression was 13.3% among subjects improving their frailty status at follow up (n=15), 15.0% in those who remained stable (n=79), and 26.7% among worsening participants (n=67) (p = 0.001). Pre-frailty at baseline did not predict the onset of depression (HR 0.82, 95% CI 0.55-1.21; Wald df=1; p=0.43), but a deterioration during follow up in at least one additional frailty criteria was associated with a significantly higher risk (HR 1.95, 95% CI 1.32-2.89; Wald df=2; p=0.01). Improvement in frailty status was not associated with the risk of incident depression (HR 0.71, 95% CI 0.35-1.42; Wald df=2; p=0.28).
Page 3 of 23
Conclusion: Our data did not offer evidence that pre-frailty per se predisposes to the onset of depression, but worsening in frailty status is associated with almost a twofold increased risk of incident depression, irrespective from the initial level of impairment.
Keywords: frailty, elderly, depression, pre-frailty, aging
Page 4 of 23
INTRODUCTION Frailty is a condition characterized by an increased vulnerability to stressors due to a reduction in the functional reserve of various organs/systems which become unable to withstand life events (1). The prevalence of frailty syndrome is reported to be between 10 and 20% among community-dwelling older adults (2).
According with Fried et al. (3), the core elements of the frailty phenotype are unintentional weight loss, weakness, exhaustion, slowness, and low physical activity. In the progression from a state of non-frailty, or robustness, to frailty there is frequently an intermediate condition called pre frailty and defined as the presence of one or two of the above criteria. Although the majority of older people gradually become frail, the dynamic nature of this pathway is also potentially reversible, especially in the earlier stages, (i.e. for the pre-frail condition) (4).
Frailty has been associated with an increased risk of several deleterious outcomes in older people, including disability, hospitalization, and institutionalization (5). More recently, a growing body of evidence has suggested that frailty could be considered as an independent risk factor for pathologies like cardiovascular diseases and diabetes (6,7).
Scientific evidence has for some time suggested that an interplay exists between frailty and depression, with some research finding that frailty predisposes to depression (8,9) and others reporting that depression accelerates the progression of frailty (6,10-14). Differently, the relationship between pre-frailty and depression is less clear, but, due to its reversibility, pre frailty may represent a potential window of opportunity to reduce the mental health impact.
Page 5 of 23
However, frailty is a dynamic process and subjects can move through frailty classes during lifetime. For instance, Gill et al. (15) found that in 18 months, 12% of pre-frail elders transitioned to robustness and 25% transitioned to frailty while 23% of frail elders transitioned to pre-frailty. So it would be interesting to understand if changes (worsening or improvement) in frailty items, influences the relationship with incident depression. However, this topic is yet to be adequately explored in the literature.
The aim of the present study was to evaluate whether pre frailty was associated with the risk of developing depression in a population-based sample of community-dwelling elderly subjects. In addition, we investigated whether changes in frailty status, expressed as the change in the number of frailty items during follow up, is associated with changes in depression among people with pre frailty.
METHODS
Study population
Data for this analysis were drawn from the Progetto Veneto Anziani (Pro.V.A.), an observational cohort study of an Italian population of older men and women living in two separate geographical areas near the city of Padua in the Veneto Region of Northern Italy. Sampling procedure, study design, and data collection method have been extensively described elsewhere (16). Briefly, the baseline Pro.V.A. study population consisted of 3,099 age- and sex-stratified, community-dwelling Caucasian participants aged ≥65 years (1854 women and 1245 men) randomly selected between 1995 and 1997 using a multistage stratification method. No exclusion
Page 6 of 23
criteria were used. Follow up assessments were conducted after 4 years (mean follow up 4.4±1.1 years). At baseline and at follow up, information was collected on participants’ formal education, living status (i.e. living alone or not), smoking and regular exercising habits during an in-person interview. Smoking habits were classified as “never/former” (for at least a year in the past) versus “current” smokers. Educational level (the total number of years of school attended) was dichotomized as ≤5 versus >5 years of schooling. Monthly income was recorded in the database in Italian liras and dichotomized using the cut-off of 500,000 Italian liras, which would correspond to about 800 Euros taking into account ISTAT monetary revaluation coefficients (17). Alcohol drinking was categorized as “yes” vs. “no” in the previous month. Participants were asked to report how many hours a week they had spent on practicing physical activity in the previous month; this information was dichotomized in <4 versus ≥4 hours/week. Disability was defined as the inability or need for assistance to perform one or more activities of daily living (ADL): bathing, dressing, eating, using the toilet, continence or transferring (18).
Body weight was measured on a calibrated balance scale. Height was measured with a stadiometer to the nearest 0.5 cm. Body mass index (BMI) was calculated as weight (in kilograms) divided by height (in meters) squared.
Any medical conditions were identified by board-certified physicians involved in the study, who examined all the clinical information collected on each participant, including clinical history, self-reported symptoms (using standardized questionnaires), medical and hospital records, blood tests, and a physical examination. The major diseases considered were any of the following: diabetes, cardiovascular diseases (CVD: congestive heart failure, angina and myocardial
Page 7 of 23
infarction, stroke, and peripheral artery disease), hip fracture, hand and lower limbs osteoarthritis, cancer, and vision and hearing impairment.
For the purpose of the present study, 635 participants were excluded because of cognitive impairment defined as a score in the Mini-Mental State Examination (MMSE) (19) below 26 and thus the Geriatric Depression Scale may have been unreliable (20). In addition, people with depression (n=1141), severe neurological diseases (i.e. people with terminal neurological disease such as multiple sclerosis or stroke) (n=43) or frailty (n=19) at baseline were not included. A further 370 people did not have complete follow up data. As shown in Figure 1, 891 participants constituted the final analytical sample for this study.
The present study was approved by the Human Studies Committee of Padua University and the Veneto Region’s Local Health Units (USSL) No.15 and No.18, and all subjects gave informed consent before their participation.
Definition of depression
Any presence of depressive symptoms was assessed at the baseline and at the follow-up with the Geriatric Depression Scale (GDS) (21), a 30-item self-reporting tool for identifying depression that has been extensively validated for use in the elderly. GDS scores range from 0 to 30, with a score of ≥11 being utilized as cut off for depression (21). The diagnosis of depression was confirmed by geriatricians skilled in psychogeriatric medicine using a standardized questionnaire checking also additional relevant information such as signs and symptoms, medical records, and medication use (16).
Page 8 of 23
Definition of pre-frailty
For the definition of pre-frailty and frailty we adopted a slightly modified version of the criteria set by Fried (3) as follows:
1. unintentional weight loss: self-reported weight loss above 5 kg over the past year not due to voluntary calorie restriction; 2. exhaustion: answering “No” to the question “Do you feel full of energy?” plus GDS score ≥ 10 (20); 3. low-energy expenditure: energy expenditure for leisure-time activities in an ideal week of the previous month below 383 kcal/week in males and 270 kcal/week in females (3,22); 4. weakness: handgrip strength below the sex and BMI cut-offs proposed by Fried3. Handgrip strength was measured in Kg using a JAMAR hand-held dynamometer (BK7498, Fred Sammons, Inc.). The best result obtained at three attempts at the dominant side was used for analyses; 5. slow gait speed: walking speed measured in meters/second below the sex and BMI cutoffs proposed by Fried (3). Gait speed was defined as the best performance achieved in two walks at usual pace along a 4-meter corridor. Participants were allowed to use canes or walkers.
Subjects unable to perform handgrip or the walking item were considered like having weakness or slow gait speed, respectively. At baseline, participants were classified as frail if they met 3 or more of the 5 criteria, as pre-frail if they met 1 or 2, and as robust if they met none of the criteria.
Page 9 of 23
However, using the inclusion/exclusion criteria cited above, no frail participants were included in the present study.
Statistical analysis
For the purpose of the study, the number of frailty items satisfied by each participant at baseline was compared with those recorded at follow up, calling this variable “change in frailty status”. This variable was then used to divide subjects in three subgroups: if the number of satisfied criteria at follow up was lower than at the baseline the subject was considered as “improved” (Group 1), if the number was the same the subject was considered as “stable” (Group 2), and if it was higher the subject was considered as “worsened” (Group 3). Participants’ characteristics were expressed as means ± standard deviations for continuous variables, and counts and percentages for categorical variables. For continuous variables, normal distributions were tested using the Shapiro-Wilk test. Age- and gender-adjusted p values were calculated as follows: for continuous variables the differences between the means of the covariates were analysed using a general linear model; for categorical variables logistic regression was applied. Bonferroni’s correction was applied in all comparisons. Cox’s proportional hazard models were used to assess associations between the three groups of subjects and incident depression by plotting the Schoenfeld residuals versus time. Known factors associated with frailty and/or depression were analysed as covariates in the univariate analysis. To explore whether a variable should be included as a predictor in the final model, the log-rank test of equality across strata was performed for all the categorical variables and Cox’s univariate
Page 10 of 23
proportional hazards regression for all the continuous variables. The predictors included in the final model were all the variables reaching a p<0.20 in the univariate analyses. Collinearity among covariates was quantified by the variance inflation factor (VIF) among the covariates, but no variable was excluded for this reason. Hazard ratios (HRs) and 95% confidence intervals (CIs) were used to compare depression rates across the tree groups of subjects taking Group 2 (=stable) as reference. We conducted several sensitivity analyses stratifying for clinical factors independently associated with depression in the final model to test the robustness of the results to check if pre-frailty can depend on any condition, but any of these factors appear to be relevant.
All analyses were performed using the SPSS 23.0 for Windows (SPSS Inc., Chicago, Illinois). All statistical tests were two-tailed and statistical significance was assumed for a p-value <0.05
RESULTS Baseline characteristics
Among 891 subjects included in the study, 44.0% were pre-frail and the remaining were robust. Due to the inclusion criteria, none of the participants were frail or depressed at baseline. Over the course of the study, 59% (n = 527) maintained their frailty status, whilst 13% improved (n = 113) and 28% (n = 251) worsened in at least one frailty item. Table 1 shows the baseline characteristics of the subjects divided in three groups by the change in the number of frailty items during follow up. Subjects who worsened their frailty status at follow up (Group 3), were significantly older than the other participants at baseline. No significant differences were observed among the three groups regarding general and anthropometric characteristics
Page 11 of 23
(educational level, income, physical activity, smoking, alcohol drinking, disability, BMI and living status).
At baseline the GDS score was similar in the three groups (6.7±2.6 in Group 1, 6.3±2.1 in Group 2, and 6.6±2.1 in Group 3, df=1206; p = 0.06).
Regarding comorbidities, the prevalence of diabetes, CVD, visual and hearing impairment was significantly higher in Group 3 subjects compared to the others.
Follow-up data
During the follow up,167 new cases (70 men and 97 women) of depression were recorded. Across the three groups, the cumulative incidence rate of depression was 13.3% among subjects improving their frailty status at follow up (Group 1), 15.0% in the stable group (Group 2), and 26.7% among those with worsening frailty status (Group 3) (age- and sex-adjusted p = 0.001). Using Cox’s regression analysis, and adjusting for baseline potential confounders, the presence of pre-frailty at baseline did not predict the onset of depression (HR 0.82, 95% CI 0.55-1.21; Wald df=1; p=0.43). Taking subjects who remained stable at follow up as reference, after adjusting for sex and age (Figure 2a) and other potential confounders (including pre-frailty) (Figure 2b), improvement in frailty status were not significantly associated with the risk of incident depression (HR 0.71, 95% CI 0.35-1.42; Wald df=2; p=0.28). On the contrary, worsening in at least 1 additional frailty criteria was associated with a significantly higher risk of incident depression (HR 1.95, 95% CI 1.32-2.89; Wald df=2; p=0.01).
Page 12 of 23
DISCUSSION
The present study utilizing a large sample of community-dwelling Caucasian elderly subjects, offered no evidence that, contrary to our hypothesis, pre-frailty per se did not predict the onset of depression over a 4.4-year follow up. Conversely, worsening in frailty status from baseline influenced the risk of incident depression. In particular, participants who accumulated during the observation period at least one more frailty criterion, had almost a two-fold increased risk of depression, also independently from their frailty status at baseline.
The main strength of the present study lies in the fact that we focused on robust and pre-frail individuals, in which the risk of developing depression in the long-term has not been adequately explored. Whilst there is a number of cross-sectional studies in the literature suggesting that frailty and depression are interrelated (2,8,9,11,23), due to the fact they are cross-sectional, it is not possible to disentangle the directionality of this association. Moreover, considering that negative depressive symptoms can overlap with frailty criteria, results from cross-sectional studies can be difficult to interpret with confidence. We chose to study the risk of incident depression in not yet frail subjects mainly for three reasons: first, because the study model would be clearer if only non-depressed and not yet frail individuals were selected. Second, because, being robust and pre-frail are two key categories and there is minimal data considering the interplay between the two. Finally, pre-frailty is potentially reversible and may have important clinical preventive implication.
Page 13 of 23
Regarding the relation between pre-frailty and depression, to the best of our knowledge, only two longitudinal studies (8,9) are available. Our findings are in line with those of Makizako et al. (9) who did not find a significant association between the two conditions in 3025 elderly Japanese people over a 15-month follow up after adjusting for potential confounders (OR 1.38, 95% CI 0.93-2.05). Conversely, Feng et al. (8) in a sample of 1827 middle-aged and elderly Chinese persons found a significant association (OR 2.26, 95% CI 1.12-4.57) between pre-frailty and depression over a 4-year follow up. A possible explanation for these differences could be that Feng et al. (8) did not include MMSE score among the exclusion criteria whereas Mikizako et al. (9) excluded subjects with a MMSE score <18/30 as well as we excluded all participants with a MMSE score < 24/30. Including people with a possible diagnosis of dementia poses diagnostic uncertainties since that in older adults depressive symptoms may mimicry cognitive deficits (24). Moreover, scales for the evaluation of depression in the elderly as the Geriatric Depression Scale do not maintain their validity in populations that contain patients with dementia (20).
As regards the change in frailty status, we observed that the majority of participants remained stable whereas 41% experienced a change from baseline frailty status. Among subjects that changed their status, about a third of them experienced an improvement, and the remaining two thirds deteriorated. The proportion of worsening subjects is higher compared to that reported by Lee et al. (4) but they included also frail subjects, who clearly cannot worsen further.
Regarding the relation between changes of frailty status and incidence of depression, to our knowledge, in the literature studies are lacking. Not necessarily accumulation of one or more frailty criteria meant that the subject became frail. Our findings seem to suggest that pre-frail elderly subjects are at higher risk for depression when they increase their level of frailty. In fact,
Page 14 of 23
subjects who worsened in terms of frailty, at baseline had a greater physical comorbidity, as evidenced by the higher prevalence of diabetes, cardiovascular diseases, osteoarthritis and sensory deficits. However, change in frailty status remained a predictor of incident depression even after adjustment for potential confounders, including comorbidities. In this sense, new onset depression could be a consequence of a negative modification of subject’s basal condition to which he/she had previously adapted. Subjective perception of physical worsening may suggest getting closer to death and impairment in motor functions favor social isolation. However, this finding does not exclude that accumulation and progression of chronic comorbidities may add an important burden to the onset of depression.
Considering that multimorbidity contributes and predisposes to the frailty phenotype, the practical implication of this finding could be that trying to improve or just preserving patients clinical and functional status may reduce the long-term risk of developing depression. And recalling the vicious circle in which comorbidity precipitates depression and depression predisposes to comorbidity and hence to disability, frailty and death, delaying or preventing worsening of frailty status, and treating depression where present, may reduce also health care costs and mortality.
The present study has a number of limitations. First we were unable to assess changes in the number of frailty items at shorter intervals than the 4.4-year follow up. Moreover we did not have any information if during their life these participants suffered from any form of depression. We were able to ascertain depression only at the time of evaluation, thus, it is possible that some cases of incident depression were relapses of a previously diagnosed depression. Second, unintentional weight loss was self-reported so it may be not so accurate. However, this criterion
Page 15 of 23
was the same as proposed by Fried et al. (3) in their original work. Another limit of the present study lies in that daily physical activity level was self-reported by participants during face-to face interviews, so the chances of an under- or over-reporting bias should be taken into account. Moreover, we had a number of people excluded at baseline which may lead to an underestimation of the incidence of depression in the general population. Third, both depression and frailty are conditions more frequent in older women than in men. Unfortunately, due to the limited sample size of our study, we were not able to stratify our analyses for gender. Future studies are so needed to detect specific gender differences. Finally, we used a slightly modified version of frailty index proposed by Fried et al. (3), and adapting the frailty criteria can influence the quality of the composite score and potentially introduce bias (25).
In conclusion our study confirmed that the dynamism of frailty pathway entails, at least in the early stages (i.e. pre-frailty), the possibility to reverse to a lower level of impairment. Our data did not offer evidence that pre-frailty per se predisposes to the onset of depression, but negative changes (i.e. worsening) in frailty level almost doubles the risk of incident depression, irrespective from the initial level of impairment. These findings point to the importance of improving or at least trying to preserve elderly people’s functionality in order to reduce comorbidities accumulation and functional deficits that would drag the subjects in an escalation of disease vulnerability, depression, worsening of frailty status and death. Future randomized clinical trials are needed to understand the clinical implication of treating frailty or pre-frailty in the prevention of new onset depression in the elderly.
ACKNOWLEDGEMENTS
Page 16 of 23
This work was supported by the Fondazione Cassa di Risparmio di Padova e Rovigo; University of Padova; the Azienda Unità Locale Socio Sanitaria 15 and 18 of the Veneto Region; a grant from the Veneto Regional Authority (Ricerca Sanitaria Finalizzata n.156/03), and a grant from the University of Padova (Population aging - economics, health, retirement and the welfare state - POPA_EHR).
Page 17 of 23
REFERENCES
1. Cigolle CT, Ofstedal MB, Tian Z, et al. Comparing models of frailty: the Health and Retirement Study. J Am Geriatr Soc 2009;57:830-9. 2. Buigues C, Padilla-Sánchez C, Garrido JF, et al. The relationship between depression and frailty syndrome: a systematic review. Aging Ment Health 2015;19:762-72. 3. Fried LP, Tangen CM, Walston J, et al.; Cardiovascular Health Study Collaborative Research Group. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci 2001;56:M146-56. 4. Lee JS, Auyeung TW, Leung J, et al. Transitions in frailty states among communityliving older adults and their associated factors. J Am Med Dir Assoc 2014;15(4):281-6. 5. Fugate Woods N, LaCroix AZ, Gray SL, et al. Frailty: Emergence and consequences in women aged 65 and older in the Women’s Health Initiative Observational Study. J Am Geriatr Soc 2005;53:1321e1330. 6. Sergi G, Veronese N, Fontana L, et al. Pre-frailty and risk of cardiovascular disease in elderly men and women: the Pro.V.A. study. J Am Coll Cardiol 2015;65:976-83. 7. Veronese N, Stubbs B, Fontana L, et al. Frailty Is Associated with an Increased Risk of Incident Type 2 Diabetes in the Elderly. J Am Med Dir Assoc 2016;17:902-7. 8. Feng L, Nyunt MS, Feng L, et al. Frailty predicts new and persistent depressive symptoms among community-dwelling older adults: findings from Singapore longitudinal aging study. J Am Med Dir Assoc 2014;15:76.e7-76.e12.
Page 18 of 23
9. Makizako H, Shimada H, Doi T, Yoshida D, et al. Physical frailty predicts incident depressive symptoms in elderly people: prospective findings from the Obu Study of Health Promotion for the Elderly. J Am Med Dir Assoc 2015;16:194-9. 10. Rovner BW, German PS, Brant LJ, et al. Depression and mortality in nursing homes. JAMA 1991;265:993-6. Erratum in: JAMA 1991;265:2672. 11. Penninx BW, Guralnik JM, Ferrucci L, et al. Depressive symptoms and physical decline in community-dwelling older persons.JAMA 1998;279:1720-6. 12. Whooley MA, Kip KE, Cauley JA, et al. Depression, falls, and risk of fracture in older women. Study of Osteoporotic Fractures Research Group. Arch Intern Med 1999;159:484-90. 13. Bruce ML. Depression and disability in late life: directions for future research. Am J Geriatr Psychiatry 2001;9:102-12. 14. Andersen K, Lolk A, Kragh-Sørensen P, et al. Depression and the risk of Alzheimer disease. Epidemiology2005;16:233-8. 15. Gill TM, Gahbauer EA, Allore HG, et al. Transitions between frailty states among communityliving older persons. Arch Intern Med. 2006; 166:418–423. 16. Corti MC, Guralnik JM, Sartori L, et al. The effect of cardiovascular and osteoarticular diseases on disability in older Italian men and women: rationale, design, and sample characteristics of the Progetto Veneto Anziani (PRO.V.A.) study. J Am Geriatr Soc 2002;50:1535-40. 17. http://www.istat.it/it/archivio/30440. Accessed April2, 2016. 18. Katz S, Ford A, Moskowitz R. The index of ADL: a standardized measure of biological and psychosocial function. JAMA 1963;185:914-9.
Page 19 of 23
19. Folstein MF, Folstein SE, McHugh PR. "Mini-mental state". A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189-98. 20. Burke WJ, Houston MJ, Boust SJ, et al. Use of the Geriatric Depression Scale in dementia of the Alzheimer type. J Am Geriatr Soc 1989; 37:856e860. 21. Yesavage JA, Brink TL, Rose TL, et al. Development and validation of a geriatric depression screening scale: a preliminary report. J Psychiatr Res 19821983;17:37-49. 22. Gary R. Evaluation of frailty in older adults with cardiovascular disease: incorporating physical performance measures. J Cardiovasc Nurs 2012;27:120-31. 23. Collard RM, Arts M, Comijs HC, et al. The role of frailty in the association between depression and somatic comorbidity: results from baseline data of an ongoing prospective cohort study. Int J Nurs Stud 2015;52:188-96. 24. Leyhe T, Reynolds CF 3rd, Melcher T, et al. A common challenge in older adults: Classification, overlap, and therapy of depression and dementia. Alzheimers Dement 2016; [Epub ahead of print] doi: 10.1016/j.jalz.2016.08.007. 25. Theou O, Cann L, Blodgett J, et al. Modifications to the frailty phenotype criteria: Systematic review of the current literature and investigation of 262 frailty phenotypes in the survey of health, ageing, and retirement in Europe. Ageing Res Rev 2015;21:78e94.
Page 20 of 23
FIGURE LEGENDS Figure 1. Flow chart of patients’ selection process.
Figure 2. Association between change in frailty status and risk of incident depression afteradjusting for sex and age (Figure 2a) and also for other potential confounders (Figure 2b).
Figure 2 footnotes: in Figure 2b, HR are adjusted for sex, age, monthly income, current smoking, GDS score, MMSE score, diabetes, cardiovascular diseases, osteoarthritis, and visual and hearing impairment.
Page 21 of 23
Table 1. Baseline characteristics of the sample divided by changes in frailty status from baseline to follow up. Patients’ characteristics
Group 1 - Improved
Group 2 - Stable
Group 3 - Worsened
p-value
(n = 113)
(n = 527)
(n = 251)
for overall comparison*†^
Age (years)
70.9±5.3
71.2±5.2
73.5±6.1
<0.0001**
Women (%)
69.0
49.9
53.8
0.001
Educational level > 5 years (%)
79.6
75.0
80.5
0.34
Monthly income ≥ 800 € (%)
61.1
52.0
57.4
0.18
Physical activity ≥4 h/week (%)
27.4
34.5
33.3
0.35
Current smoking (%)
6.2
12.0
12.0
0.19
Alcohol drinking (%)
72.6
74.6
72.5
0.79
Disability (%)
25.7
21.1
25.1
0.34
BMI (Kg/m2)
28.2±2
27.9±4.2
28.0±4.4
0.44
18.6
13.1
14.5
0.31
GDS score
6.7±2.6
6.3±2.1
6.6±2.1
0.06
MMSE score
26.9±2.0
27.4±1.8
27.2±1.8
0.04
General and anthropometric characteristics
Living alone (%) Cognitive functions
Comorbidities (%)
Page 22 of 23
Diabetes
11.5
10.4
17.9
0.02
CVD
13.3
9.7
18.3
0.004
Hip fracture
2.7
2.1
2.0
0.92
Hand osteoarthritis
13.3
11.4
15.9
0.21
Limbs osteoarthritis
19.5
16.1
23.1
0.06
Cancer
6.2
5.1
6.4
0.75
Visual impairment
28.3
24.9
36.7
0.003
Hearing impairment
56.6
63.0
73.3
0.002
Numbers are mean values ± standard deviations or percentages (%), as appropriate. GDS: Geriatric Depression Scale; MMSE: Mini-Mental State Examination; BMI: body mass index; CVD: cardiovascular diseases. *Unless otherwise specified, p values are adjusted for age and sex using a general linear model or logistic regression, as appropriate. †
The test with df=2
^df for the F-tests: df=2, 1206. **Univariate ANOVA (ANalysis Of Variance), not adjusted for age.
Page 23 of 23