- Email: [email protected]
Serum Dehydroepiandrosterone Sulfate and Incident Depression in the Elderly: The Pro.V.A. Study
Serum Dehydroepiandrosterone Sulfate and Incident Depression in the Elderly: The Pro.V.A. Study Nicola Veronese, M.D., Marina De Rui, M.D., Francesco Bolzetta, M.D., Sabina Zambon, M.D., Maria Chiara Corti, M.D., Giovannella Baggio, M.D., Elena Debora Toffanello, M.D., Gaetano Crepaldi, M.D., Egle Perissinotto, Sc.D., Enzo Manzato, M.D., Giuseppe Sergi, M.D.
Objective: Dehydroepiandrosterone sulfate (DHEAS) appears to have a protective effect against depression, but contrasting findings are available. Therefore, we investigated whether high serum DHEAS levels were associated with any protective effect on incident depression and incident severe depression in a representative group of elderly men and women. Methods: In a population-based cohort longitudinal study in the general community, 789 older participants without depression and cognitive impairment at the baseline were included, among 3,099 screened subjects. Serum DHEAS levels were determined based on blood samples; incident depression and severe depression were diagnosed by means of the Geriatric Depression Scale (GDS) and confirmed by geriatricians skilled in psychogeriatric medicine. Results: No baseline differences were found in GDS across age- and gender-specific tertiles of serum DHEAS. Over 4.4 years of follow-up, 137 new cases of depression were recorded. Of them, 35 among men and 64 in women were cases of incident severe depression. Cox’s regression analysis, adjusted for potential confounders, revealed that higher DHEAS levels were associated with reduced risk of incident depression irrespective of gender (HR: 0.30; 95% CI: 0.09e0.96; Wald c2 ¼ 4.09; df ¼ 1; p ¼ 0.04; women: HR: 0.31; 95% CI: 0.14e0.69; Wald c2 ¼ 8.37; df ¼ 1; p ¼ 0.004) and of severe incident depression only in men (HR: 0.25; 95% CI: 0.06e0.99; Wald c2 ¼ 4.05; df ¼ 1; p ¼ 0.04). Conclusion: Higher serum DHEAS levels were found to be significantly protective for the onset of depression irrespective of gender, whereas only in men was this association found also for incident severe depression. (Am J Geriatr Psychiatry 2015; -:-e-) Key Words: Serum DHEAS, depression, elderly
Received July 3, 2014; revised October 14, 2014; accepted October 30, 2014. From the Department of Medicine DIMED, Geriatrics Division (NV, MDR, FB, EDT, EM, GS), Department of Medical and Surgical Sciences (SZ), and Department of Cardiac, Thoracic and Vascular Sciences, Unit of Biostatistics, Epidemiology and Public Health (EP), University of Padova, Padova, Italy; National Research Council, Aging Branch (SZ, GC, EM), Institute of Neuroscience, Padova, Italy; Azienda Unità Locale Socio Sanitaria (ULSS) 16 (MC), Padova, Italy; and Internal Medicine Division (GB), Azienda Ospedaliera, Padova, Italy. Send correspondence and reprint request to Nicola Veronese, M.D., Department of Medicine - DIMED, Geriatrics Division, University of Padova, Via Giustiniani, 2 35128 Padova, Italy. e-mail: [email protected] Ó 2015 American Association for Geriatric Psychiatry http://dx.doi.org/10.1016/j.jagp.2014.10.009
Am J Geriatr Psychiatry
-:-, -
2015
1
DHEAS and Depression in the Elderly
INTRODUCTION
METHODS
Depression is widespread among elderly people, with a prevalence from 0.9% to 9.4% in communitydwelling people and higher in other settings like hospitals, nursing homes, and institutions.1 The presence of depression, particularly if severe, is strongly associated with disability, increased mortality, and poorer outcomes from physical illness.2 Research has shown that among possible factors affecting mood in the elderly, serum dehydroepiandrosterone sulfate (DHEAS) could be relevant. DHEAS is a sex steroid hormone mainly secreted by the adrenal glands but also in small amounts by the brain, where receptors for this hormone have selectively been found.3e5 DHEAS antidepressant and anxiolytic effects seem to be related to the agonistic activity on the s-1 receptor,6 the inhibition of g-aminobutyric acid A receptor directly and by decreasing pregnenolonesulfate concentrations in the brain,7 the activation of N-methyl-D-aspartate receptor,7e11 and the upregulation of brain-derived neurotrophic factor levels.12 Contrasting results have been reported in crosssectional studies about the association between serum DHEAS levels and depression, particularly in females.13e17 At the same time, only a few longitudinal studies have investigated the protective association of high serum DHEAS levels for depressive symptoms in older adults. Two of these studies found a significant association only in men, whereas another two failed to find any relationship.5,18e20 Finally, to the best of our knowledge, no study about DHEAS and incident severe depression are available, but this topic may be relevant because older people with severe depression have the highest suicide rate of any age group and a poorer prognosis of survival compared with younger subjects.21 The aim of the present study was thus to examine whether high serum DHEAS levels were associated with any protective effect on incident depression and incident severe depression and the possible role of gender in a representative group of elderly men and women over a lengthy follow-up (4.4 years).
Data Source and Subjects
2
Data for this analysis came from the Progetto Veneto Anziani (Pro.V.A.), an observational cohort study on the Italian population aged 65 years. Our study population included 3,099 age- and sex-stratified white participants (1,854 women and 1,245 men) randomly selected between 1995 and 1997 using a multistage stratified method. Sampling procedures and data collection methods are described elsewhere.22 Trained physicians and nurses examined participants at various clinics. The present study concerns the information collected on the incidence of depression and severe depression over a mean of 4.4 years (1.2 standard deviation) of follow-up. Among the 3,099 subjects considered, 601 participants were excluded because the Mini-Mental State Exam (MMSE) was below 24 and thus the Geriatric Depression Scale (GDS) may have been unreliable.23 Moreover, 500 participants were excluded because they were already found to be depressed at baseline, and 118 were taking antidepressant medications. Finally, 696 participants did not have DHEAS measurement available at baseline. The final sample thus consisted of 789 participants (247 died and 148 were lost at follow-up) (Fig. 1). Participants without DHEAS measurement at baseline did not differ from the participants included in the final analysis in age (72.9 6.4 versus 72.8 6.1 years, p ¼ 0.79) or MMSE score (27.34 1.76 versus 27.10 1.74, p ¼ 0.09). However, they showed a significantly lower GDS score (6.15 2.39 versus 6.79 2.00 points, p <0.0001). The ethical committees of Padua University and the Veneto Region’s Local Health Units (USSL) n. 15 and n. 18 approved our study protocol. Participants gave written informed consent to the study. Social and Medical Conditions Participants were examined at the city hospitals by trained physicians and nurses. Information was collected during a face-to-face individual interview on their social and medical conditions. Smoking habits were dichotomized as “current smoker” versus “never/former smoker” (if a subject had given
Am J Geriatr Psychiatry
-:-, -
2015
Veronese et al.
FIGURE 1. Flow chart of how the study population was selected: the Pro.V.A. study.
measured with hexokinase glucose-6-phosphate dehydrogenase (Dimension Vista System; Siemens, Milan, Italy). Total cholesterol was calculated using the enzymatic method. The diagnosis of vision impairment was made with the aid of the visual acuity test and that of hearing impairment with audiometry and confirmed by personnel experts in ophthalmology and otorhinolaryngology, respectively. Body weight and height were measured by trained physicians, and the body mass index (kg/m2) was calculated. Multidimensional Evaluation
up smoking at least a year earlier). Regular physical activity was defined as at least 4 hours a week in the previous month of at least moderate physical activity (brisk walking, cycling, gardening, dancing, or physical exercising). Alcohol drinking was categorized as “yes” or “no” during the last month. Educational level (the total number of years of school attended) was dichotomized as less than 5 versus 5 or mreo years of schooling. Monthly income was indicated as family income less than or equal to or more than 500 V. Any diseases were assessed by board-certified physicians, who examined all clinical details collected for each participant in the study, including their clinical history, symptoms self-reported by means of standardized questionnaires, medical and hospital records, blood tests, and physical examinations. A history of major diseases that could lead to depression was recorded for cardiovascular diseases (congestive heart failure, angina and myocardial infarction, stroke, and peripheral artery disease) and diabetes. Also, blood parameters that previous research has shown to be related to depression (fasting plasma glucose and cholesterol) were included in this work.20 Fasting plasma glucose was
Am J Geriatr Psychiatry
-:-, -
2015
A multidimensional evaluation was made by expert physicians in geriatric medicine both at baseline and follow-up. Functional status was assessed using an activities of daily living score.24 The Short Physical Performance Battery (SPPB)25 was derived from three objective physical function tests. Each test was scored from 0 (inability to complete the test) to 4 (highest level of performance). The scores from all three tests were summarized in a composite score of 0 to 12, higher scores reflecting better physical function. The three tests were as follows: 1. Tandem test: Participants were asked to maintain balance progressively side-by-side, in semitandem, and in full-tandem position. 2. Gait speed: The best performance achieved in two walks at usual pace along a 4-m corridor was recorded in meters/second. Participants were allowed to use canes or walkers. 3. Chair stand time: Participants were asked to stand up and sit down five times as quickly as possible, with their hands folded across their chest. The time taken to complete the test, in seconds, was recorded. Cognitive function was assessed by administering the 30-item MMSE. Frailty was assessed with the evaluation of five measurable items. Unintentional weight loss was defined as self-report of unintentional weight loss above 5 kg over the past year. Weakness and slow gait speed were defined, respectively, using the best value of handgrip strength and the best timed walk over 4-m at usual pace stratified according to sex and body mass index cut-offs as proposed by Fried et al.26 Exhaustion was defined using the question, “Do you feel full of energy?,” in the GDS
3
DHEAS and Depression in the Elderly questionnaire; low-energy expenditure was defined as the sum of leisure-time activities assessed in an ideal week of the previous month with a cut-off value below 383 kcal/wk in men and 270 kcal/wk in women.26,27 Participants were classified as frail if they met three or more of the five criteria. This variable was categorized as “frail” versus “not frail.” DHEAS Measurement Serum levels of DHEAS were determined by means of a chemiluminescent enzyme immunoassay with an interassay coefficient of variation of 10.9%. This method seems to be highly comparable with standard radioimmunoassay.20 DHEAS rather than DHEA was measured because it shows little circadian, monthly, and seasonal variations, making it more reliable as an epidemiologic marker.28 Because of the linear decrease with aging and the significant differences between genders, DHEAS tertiles were calculated using gender-specific tertiles according to the median age of our sample (75 years). In those younger than 75 years, the two cut-offs for dividing the sample in tertiles were 1.90 and 3.40 mmol/L in men and 1.00 and 2.00 mmol/L in women, whereas the correspondent values in those over 75 years were 1.60 and 2.80 and 0.80 and 1.60 mmol/L, respectively. Definition of Depression Any presence of depressive symptoms was assessed at baseline and at follow-up with the GDS, a 30-item self-reporting tool for identifying depression extensively validated for use in the elderly.29 GDS scores range from 0 to 30, with a score of at least 11 diagnostic for depression and a score of at least 20 plus limitation in activities of daily living due to depressive symptoms indicating severe depression.29 The diagnosis of depression and severe depression were 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.22 Statistical Analysis Participants’ characteristics were summarized using means (standard deviations) for continuous
4
variables and counts or percentages for categorical variables. For continuous variables, normal distributions were tested using the Shapiro-Wilk test. Because previous research has shown that serum DHEAS levels significantly differ between genders and there is a significant sex by DHEAS level interaction for the onset of depression,3,13,18,20 we tested if a similar interaction exists in our study using a Cox’s regression analysis. Because this interaction was significant, all data analyses were made using gender-specific tertiles of baseline DHEAS. The independent sample t test was used for comparing serum DHEAS values between men and women, without adjustment for age. Age-adjusted p values for trends were calculated, checking the differences between the means of the covariates by DHEAS group using analysis of variance. Differences in categorical variables were examined using the c2 test, adjusted for age using a logistic regression. Cox’s proportional hazard models were used to assess the association between gender-specific tertiles of DHEAS and incident and incident severe depression. Known factors associated with DHEAS and/or depression were considered for inclusion in the analysis, and consequently two models were built. Model 1 includes social conditions and items of the multidimensional evaluation; model 2 includes the covariates in model 1 and some medical conditions or blood parameters that previous research has shown to be associated with the development of depression. To explore whether a variable should be included as a predictor in the final survival model, the log-rank test of equality across strata was performed for all categorical variables and Cox’s univariate proportional hazards regression for all continuous variables. The predictors included in the final model were all the variables reaching p <0.20 in the univariate analyses. Collinearity among covariates was quantified by the variance inflation factor. Hazard ratios (HRs) and 95% confidence intervals (CIs) were used to compare rates of incident depression across DHEAS tertiles, taking those with lower DHEAS levels as reference. Similar analysis was made for incident severe depression. All analyses were performed using SPSS 21.0 for Windows (SPSS Inc., Chicago, IL). All statistical tests were two-tailed, and statistical significance was assumed for p <0.05.
Am J Geriatr Psychiatry
-:-, -
2015
Veronese et al.
TABLE 1. Baseline Characteristics According to Baseline DHEAS Tertiles: the Pro.V.A. Study Men
Age, years BMI, kg/m2 Social conditions Current smokers, % Physical activity 4 hours/weeks, % Alcohol drinking, % Education >5 years, % Monthly income 500 V, % Living alone, % Multidimensional evaluation GDS score (points) MMSE score (points) ADL score (points) SPPB total score (points) Medical conditions CVD, % Diabetes, % Vision impairment, % Hearing impairment, % Biohumoral tests Total cholesterol (mg/dL) FPG (mg/dL)
Tertile 1 (N [ 125)
Tertile 2 (N [ 115)
Tertile 3 (N [ 117)
pa
74.88 (7.36) 26.76 (3.82)
72.86 (6.23) 27.58 (3.42)
71.82 (5.33) 27.80 (4.16)
<0.0001b 0.24
8.8 40.8 87.2 25.0 53.6 8.1
13.9 47.0 91.3 25.4 43.5 4.4
23.1 36.8 90.6 21.7 46.2 7.7
0.007 0.37 0.30 0.28 0.44 0.59
6.58 (1.94) 26.06 (3.25) 5.73 (0.72) 10.18 (1.97)
6.48 (1.97) 25.91 (3.53) 5.78 (0.74) 10.52 (1.83)
6.69 (1.96) 26.09 (3.31) 5.77 (0.61) 10.75 (1.82)
0.65 0.42 0.97 0.67
19.2 19.2 36.5 75.2
20.0 14.8 36.0 81.7
14.5 7.7 22.2 80.3
0.69 0.01 0.27 0.13
223.58 (38.48) 106.98 (30.77)
222.22 (41.69) 108.96 (31.62)
231.71 (38.71) 103.54 (20.97)
0.34 0.30
Women
Age, years BMI, kg/m2 Social conditions Current smokers, % Physical activity 4 h/wk, % Alcohol drinking, % Education >5 years, % Monthly income 500 V, % Living alone, % Multidimensional evaluation GDS score (points) MMSE score (points) ADL score (points) SPPB total score (points) Medical conditions CVD, % Diabetes, % Vision impairment, % Hearing impairment, % Biohumoral tests Total cholesterol, mg/dL FPG, mg/dL
Tertile 1 (N [ 156)
Tertile 2 (N [ 133)
Tertile 3 (N [ 143)
pa
74.17 (6.18) 28.48 (5.00)
72.41 (5.54) 28.74 (5.23)
71.93 (5.72) 28.68 (4.74)
<0.0001b 0.95
3.8 25.0 56.4 15.6 76.6 22.6
7.1 25.6 59.4 10.6 70.8 21.2
8.4 18.2 68.5 11.3 67.2 21.8
0.94 0.07 0.02 0.13 0.10 0.92
6.90 (1.98) 25.51 (3.54) 5.60 (0.74) 9.08 (2.44)
6.78 (2.08) 25.41 (3.14) 5.57 (0.67) 9.45 (2.16)
6.87 (2.18) 25.47 (3.48) 5.65 (0.61) 9.80 (2.11)
0.91 0.44 0.64 0.32
10.9 12.8 30.8 53.2
6.8 15.0 33.8 62.4
9.1 16.8 23.1 53.1
0.93 0.33 0.51 0.19
246.66 (40.33) 103.17 (26.10)
245.61 (37.28) 106.83 (31.22)
249.93 (46.17) 108.29 (31.51)
0.72 0.39
Notes: Values are means with standard deviations in parentheses or percentages, as indicated. DHEAS tertiles were defined using a stratification for median of age (75 years) and gender. In those younger than 75 years, cut-offs were 1.90 and 3.40 in men and 1.00 and 2.00 mmol/L in women; the correspondent values in those over 75 years were 1.60 and 2.80 and 0.80 and 1.60 mmol/L, respectively. Tertile 1 indicates the lowest tertile and Tertile 3 the highest. BMI: body mass index; ADL: activity of daily living; CVD: cardiovascular diseases; FPG: fasting plasma glucose. a Unless otherwise specified, p values are adjusted for age using a general linear model or logistic regression, as appropriate. The c2 test with df ¼ 2. df for F-tests: in men df ¼ 2, 354 (among groups) and 2, 356 (within groups); in women df ¼ 2, 429 (among groups) and 2, 431 (within groups), respectively. b Univariate analysis of variance, not adjusted for age.
Am J Geriatr Psychiatry
-:-, -
2015
5
DHEAS and Depression in the Elderly
RESULTS The sample consisted of 789 community-dwelling subjects over age 65 years (357 men and 432 women) without depression and cognitive impairment at baseline. Mean ages were 73.2 6.5 for men and 72.9 5.1 years for women. Mean DHEAS levels in men were significantly higher than in women (independent samples t test; 2.93 1.96 mmol/L in men versus 1.66 1.26 mmol/L in women; df ¼ 787; t ¼ 11.003; p <0.0001). Moreover, a Cox’s regression analysis showed a significant interaction sex by baseline serum DHEAS levels for incident depression (HR: 0.89; 95% CI: 0.84e0.94; Wald c2 ¼ 15.60; df ¼ 1; p <0.0001). Table 1 shows the baseline characteristics of participants divided according to gender and baseline serum DHEAS levels. In men, there was a significant linear trend in percentage of current smokers but an inverse trend for the presence of diabetes. In women, only the presence of alcohol drinkers was significantly different across the tertiles, being higher in those with higher serum DHEAS levels (Table 1). No differences were evident for other variables investigated, including baseline GDS score (Table 1). Over a mean period of 4.4 years, 137 new cases of depression were recorded, with an incidence of 13.7% in men and 20.3% in women. Of these, 35 among men and 64 in women were cases of incident severe depression. As shown in Table 2, taking those with lower DHEAS as reference, Cox’s regression analysis shows that in the sample as a whole, participants with higher serum DHEAS levels had a significantly decreased risk of being depressed during the follow-up period, whereas no association was evident for incident severe depression. After dividing for gender, male participants with higher serum DHEAS levels had a significantly decreased risk of having both incident (HR: 0.30; 95% CI: 0.09e0.96; Wald c2 ¼ 4.09; df ¼ 1; p ¼ 0.04) and incident severe depression (HR: 0.25; 95% CI: 0.06e0.99; Wald c2 ¼ 4.05; df ¼ 1; p ¼ 0.04), whereas in women the association between high DHEAS levels and depression was present only for incident depression (HR: 0.31; 95% CI: 0.14e0.69; Wald c2 ¼ 8.37; df ¼ 1; p ¼ 0.004) (Table 2). Modeling serum DHEAS as a continuous variable, an increase of 1 mmol/L led to a significant decrease in having both
6
incident and incident severe depression in male participants, whereas in female participants these associations were not evident (Table 2). Among the other covariates considered, SPPB was a significant protective factor for incident depression independently from gender (HR: 0.73; 95% CI: 0.58e0.91; Wald c2 ¼ 7.70; df ¼ 1; p ¼ 0.001 in men and HR: 0.87; 95% CI: 0.75e0.99; Wald c2 ¼ 4.90; df ¼ 1; p ¼ 0.03 in women), whereas hearing impairment was a risk factor for incident depression (HR: 3.37; 95% CI: 1.07e12.70; Wald c2 ¼ 4.87; df ¼ 1; p ¼ 0.03 in men and HR: 2.71; 95% CI: 1.41e5.20; Wald c2 ¼ 8.91; df ¼ 1; p ¼ 0.003 in women). On the contrary, the presence of frailty produced a significant risk factor only in women (HR: 1.33; 95% CI: 1.04e1.62; Wald c2 ¼ 8.02; df ¼ 1; p ¼ 0.004).
DISCUSSION Our large population-based prospective study found that higher serum DHEAS levels reduced the risk of incident depression over a 4.4-year follow-up in both genders, whereas the association with severe depression was found only in men. The few existing studies about this topic gave contrasting results. Two studies, in fact, did not find any association between serum DHEAS levels and risk of depression,5,30 whereas another two studies found an association only in men.18,20 The differences between our study and others with negative findings may be explained by methodologic issues. First, we used the 30-item GDS for detection of depression, and the diagnosis was confirmed by geriatricians skilled in psychogeriatric medicine, whereas in other studies shorter forms of the GDS or other tools (like the Center for Epidemiologic Studies Depression Scale) were used alone. Second, we did not include subjects with dementia or cognitive impairment. Participants with cognitive impairment at baseline, at higher risk for dementia than depression, could create a bias in assessing the relationship between serum DHEAS levels and depression. Moreover, the reliability of tools assessing depressed subjects with dementia or cognitive impairment is still a matter of discussion.23 Also, Haren et al.31 found that serum DHEAS is significantly associated with depression only in subjects without cognitive impairment, confirming that
Am J Geriatr Psychiatry
-:-, -
2015
Am J Geriatr Psychiatry
TABLE 2. Association Between Baseline DHEAS Tertiles, Incident, and Severe Depression According to Gender: the Pro.V.A. Study
-:-, -
2015
Model 2c HR (95% CI)
No. of Participants
53 51 33
281 248 260
1 [reference] 0.75 (0.47e1.20) 1.45 0.40 (0.23e0.70) 10.34 0.88 (0.78e0.99) 4.64
0.23 0.001 0.03
1 [reference] 0.78 (0.48e1.25) 1.08 0.35 (0.20e0.63) 12.32 0.88 (0.78e0.99) 4.42
0.30 <0.0001 0.04
1 [reference] 0.81 (0.48e1.35) 0.68 0.35 (0.19e0.65) 11.39 0.87 (0.77e0.999) 3.87
0.41 0.001 0.049
41 33 25
281 248 260
1 [reference] 1.13 (0.65e1.98) 0.19 0.63 (0.33e1.19) 2.05 0.92 (0.80e1.07) 1.13
0.67 0.15 0.29
1 [reference] 1.12 (0.63e1.99) 0.16 0.53 (0.27e1.04) 3.40 0.92 (0.79e1.06) 1.32
0.69 0.07 0.25
1 [reference] 1.22 (0.66e2.25) 0.39 0.56 (0.28e1.13) 2.63 0.92 (0.78e1.09) 0.99
0.53 0.10 0.32
20 19 10
125 115 117
1 [reference] 0.81 (0.36e1.82) 0.27 0.34 (0.13e0.86) 5.23 0.86 (0.74e1.00) 3.68
0.60 0.02 0.06
1 [reference] 0.57 (0.23e1.47) 1.35 0.24 (0.08e0.71) 6.70 0.83 (0.70e0.98) 5.05
0.57 0.01 0.03
1 [reference] 0.75 (0.27e2.15) 0.28 0.30 (0.09e0.96) 4.09 0.77 (0.63e0.94) 6.39
0.60 0.04 0.01
16 12 7
125 115 117
1 [reference] 1.03 (0.40e2.66) 0.05 0.38 (0.13e0.96) 2.89 0.86 (0.72e1.03) 2.61
0.94 0.02 0.11
1 [reference] 0.73 (0.24e2.12) 0.31 0.27 (0.08e0.96) 4.07 0.79 (0.65e0.97) 5.03
0.58 0.04 0.03
1 [reference] 0.81 (0.23e2.81) 0.12 0.25 (0.06e0.99) 4.05 0.71 (0.55e0.91) 7.35
0.74 0.04 0.007
33 32 23
156 133 143
1 [reference] 0.72 (0.40e1.28) 1.26 0.33 (0.13e0.86) 5.55 0.89 (0.74e1.09) 1.26
0.26 0.02 0.26
1 [reference] 0.82 (0.46e1.49) 0.41 0.30 (0.14e0.64) 9.63 0.94 (0.77e1.14) 0.41
0.52 0.002 0.52
1 [reference] 0.83 (0.44e1.58) 0.32 0.31 (0.14e0.69) 8.37 0.99 (0.80e1.22) 0.008
0.57 0.004 0.93
25 21 18
156 133 143
1 [reference] 1.22 (0.61e2.47) 0.32 0.78 (0.35e1.73) 0.37 1.03 (0.80e1.33) 0.06
0.58 0.54 0.81
1 [reference] 1.32 (0.64e2.70) 0.56 0.54 (0.23e1.27) 2.00 1.07 (0.82e1.39) 0.24
0.45 0.16 0.62
1 [reference] 1.50 (0.69e3.27) 1.03 0.59 (0.24e1.45) 1.31 1.16 (0.87e1.53) 1.02
0.31 0.25 0.31
Wald c2
pa
Wald c2
pa
Wald c2
pa
7
Notes: Unless otherwise specified, values are HR with 95% CI in parentheses. DHEAS tertiles were defined using a stratification for median of age (75 years) and gender. In those younger than 75 years, the cut-offs were 1.90 and 3.40 in men and 1.00 and 2.00 mmol/L in women; the correspondent values in those over 75 years were 1.60 and 2.80 and 0.80 and 1.60 mmol/L. Tertile 1 indicates the lowest tertile and Tertile 3 the highest. a Each p value was based on a Wald c2 test with df ¼ 1. b Model 1: adjusted for age and body mass index, activities of daily living score, MMSE score, educational level, smoking status, monthly income, physical activity level, hearing and vision impairment, living alone, SPPB, and presence of frailty. In the sample as whole, model 1 also comprehends gender as a covariate. c Model 2: adjusted for covariates in model 2 and baseline GDS score, presence of cardiovascular diseases, diabetes, and serum levels of fasting plasma glucose and total cholesterol.
Veronese et al.
Entire sample Incident depression Tertile 1 Tertile 2 Tertile 3 DHEAS (For increase of 1 mmol/L) Incident severe depression Tertile 1 Tertile 2 Tertile 3 DHEAS (For increase of 1 mmol/L) Men Incident depression Tertile 1 Tertile 2 Tertile 3 DHEAS (For increase of 1 mmol/L) Incident severe depression Tertile 1 Tertile 2 Tertile 3 DHEAS (For increase of 1 mmol/L) Women Incident depression Tertile 1 Tertile 2 Tertile 3 DHEAS (For increase of 1 mmol/L) Incident severe depression Tertile 1 Tertile 2 Tertile 3 DHEAS (For increase of 1 mmol/L)
Age-adjusted HR (95% CI)
Model 1b HR (95% CI)
No. of Events
DHEAS and Depression in the Elderly cognitive impairment is an important bias to consider in the assessment of depression in older people. The association between serum DHEAS and depression could be conditioned by gender differences as shown also by our study when considering severe depression. The severity of depression may be particularly important considering that severe depression, besides being the strongest predictor of the course of suicidal ideation, is also a strong risk factor for disability and mortality in older people.32 Our study suggests that in men, elevated serum DHEAS levels are necessary for avoiding more severe forms of depression. On the contrary, in women no association was observed between serum DHEAS levels and incidence of severe depression. The same figure was observed also when serum DHEAS was modeled as a continuous variable. Some hypotheses can be formulated to explain this result, borrowing some findings from animal models. First, in rats, male and female brain areas have shown different sensitivity to g-aminobutyric acid A receptor ligands, including DHEAS.33 Moreover, a threshold effect of DHEAS could also be involved. Supplementation with DHEA was able to increase serotonin brain levels in rats only for the highest doses, whereas no effects were observed for the other doses of treatment.34 It may be possible therefore that the association between serum DHEAS and severe depression was seen only in men who have higher DHEAS levels and that DHEAS levels even higher than those of women in the third tertile are necessary to reduce the risk for severe depression. Other than hearing impairment, an important and well-known risk factor for depression,35 frailty is another important risk factor for depression present only in women according to recent research.36 On the contrary, high SPPB scores were protective for the onset of depression, confirming that higher performance levels are associated with a lower risk of depression.37 Our findings suggest that serum
DHEAS concentrations have a protective role in the elderly, but its role could be attenuated in some conditions such as in sedentary lifestyle and frailty, and so further specific research about these populations is necessary. Several limitations of this study need to be considered. First, some subjects potentially eligible for our study did not have DHEAS measurements and had significant lower GDS scores at baseline. Second, DHEAS was measured only at baseline, and it is known that this hormone has unpredictable changes.38 Moreover, we did not estimate cortisol levels because it is known that the cortisol-to-DHEAS ratio is an independent predictor of depression.39 Finally, even if we used a reliable method for diagnosis of depression, we did not use the gold standard method, that is, a structured clinical interview combined with a clinical interview by an experienced mental health professional; therefore, other possible conditions similar to depression (e.g., thyroid diseases) were not fully investigated. In conclusion, higher serum DHEAS levels were found to be significantly protective for the onset of depression irrespectively of gender. However, only in men was this association also found for incident severe depression. The authors are grateful to all interviewers, nurses, and physicians who took part in the study. The authors also acknowledge Emily Orr for her competency in English. The Pro.V.A. study was funded by the Fondazione Cassa di Risparmio di Padova e Rovigo, University of Padova; Azienda Unità Locale Socio Sanitaria 15 and 18 of the Veneto Region, and by a grant from the Veneto Regional Authority (Ricerca Sanitaria Finalizzata n.156/03). None of the authors has any financial arrangements, organizational affiliations, or other relationships that might give rise to any conflict of interest regarding the subject matter of this article.
References 1. Djernes JK: Prevalence and predictors of depression in populations of elderly: a review. Acta Psychiatr Scand 2006; 113: 372e387 2. Rodda J, Walker Z, Carter J: Depression in older adults. BMJ 2011; 343:d5219
8
3. Orentreich N, Brind JL, Vogelman JH, et al: Long-term longitudinal measurements of plasma dehydroepiandrosterone sulfate in normal men. J Clin Endocrinol Metab 1992; 75:1002e1004 4. Guazzo EP, Kirkpatrick PJ, Goodyer IM, et al: Cortisol, dehydroepiandrosterone (DHEA), and DHEA sulfate in the cerebrospinal
Am J Geriatr Psychiatry
-:-, -
2015
Veronese et al.
5.
6.
7.
8. 9. 10.
11.
12.
13.
14.
15.
16.
17. 18. 19.
20.
21. 22.
fluid of man: relation to blood levels and the effects of age. J Clin Endocrinol Metab 1996; 81:3951e3960 Mazat L, Lafont S, Berr C, et al: Prospective measurements of dehydroepiandrosterone sulfate in a cohort of elderly subjects: relationship to gender, subjective health, smoking habits, and 10year mortality. Proc Natl Acad Sci USA 2001; 98:8145e8150 Urani A, Roman FJ, Phan VL, et al: The antidepressant-like effect induced by sigma(1)-receptor agonists and neuroactive steroids in mice submitted to the forced swimming test. J Pharmacol Exp Ther 2001; 298:1269e1279 Majewska MD, Demirgören S, Spivak CE, et al: The neurosteroid dehydroepiandrosterone sulfate is an allosteric antagonist of the GABAA receptor. Brain Res 1990; 526:143e146 van Broekhoven F, Verkes RJ: Neurosteroids in depression: a review. Psychopharmacology 2003; 165:97e110 Kulkarni SK, Dhir A: s-1 receptors in major depression and anxiety. Expert Rev Neurother 2009; 9:1021e1034 Kinouchi K, Maeda S, Saito K, et al: Effects of d- and l-pentazocine on the release and uptake of norepinephrine in rat brain cortex. Res Commun Chem Pathol Pharmacol 1989; 63:201e213 Monnet FP, Mahe V, Robel P, et al: Neurosteroids, via s receptors, modulate the [3H]norepinephrine release evoked by N-methyl-Daspartate in the rat hippocampus. Proc Natl Acad Sci USA 1995; 92:3774e3778 Naert G, Maurice T, Tapia-Arancibia L, et al: Neuroactive steroids modulate HPA axis activity and cerebral brain-derived neurotrophic factor (BDNF) protein levels in adult male rats. Psychoneuroendocrinology 2007; 32:1062e1078 Berr C, Lafont S, Debuire B, et al: Relationships of dehydroepiandrosterone sulfate in the elderly with functional, psychological, and mental status, and short-term mortality: a French communitybased study. Proc Natl Acad Sci USA 1996; 93:13410e13415 Barrett-Connor E, von Muhlen D, Laughlin GA, et al: Endogenous levels of dehydroepiandrosterone sulfate, but not other sex hormones, are associated with depressed mood in older women: the Rancho Bernardo study. J Am Geriatr Soc 1999; 47:685e691 Morsink LF, Vogelzangs N, Nicklas BJ, et al: Associations between sex steroid hormone levels and depressive symptoms in elderly men and women: results from the Health ABC study. Psychoneuroendocrinology 2007; 32:874e883 T’Sjoen GG, De Vos S, Goemaere S, et al: Sex steroid level, androgen receptor polymorphism, and depressive symptoms in healthy elderly men. J Am Geriatr Soc 2005; 53:636e642 Erdincler D, Bugay G, Ertan T, et al: Depression and sex hormones in elderly women. Arch Gerontol Geriatr 2004; 39:239e244 Goldman N, Glei DA: Sex differences in the relationship between DHEAS and health. Exp Gerontol 2007; 42:979e987 Yaffe K, Ettinger B, Pressman A, et al: Neuropsychiatric function and dehydroepiandrosterone sulfate in elderly women: a prospective study. Biol Psychiatry 1998; 43:694e700 Michikawa T, Nishiwaki Y, Nakano M, et al: Higher serum dehydroepiandrosterone sulfate levels are protectively associated with depressive symptoms in men, but not in women: a community-based cohort study of older Japanese. Am J Geriatr Psychiatry 2013; 21:1154e1163 Manthorpe J, Iliffe S: Suicide in later life: public health and practitioner perspectives. Int J Geriatr Psychiatry 2010; 25:1230e1238 Corti MC, Guralnik JM, Sartori L, et al: The effect of cardiovascular and osteoarticular diseases on disability in older Italian men
Am J Geriatr Psychiatry
-:-, -
2015
23.
24. 25.
26.
27.
28. 29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
and women: rationale, design, and sample characteristics of the Progetto Veneto Anziani (PRO.V.A.) study. J Am Geriatr Soc 2002; 50:1535e1540 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 Katz S, Downs TD, Cash HR, et al: Progress in development of the index of ADL. Gerontologist 1970; 10:20e30 Guralnik JM, Simonsick EM, Ferrucci L, et al: A short physical performance battery assessing lower extremity function: association with self-reported disability and prediction of mortality and nursing home admission. J Gerontol 1994; 49:M85eM94 Fried LP, Tangen CM, Walston J, et al: Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci 2001; 56: M146eM156 Solfrizzi V, Scafato E, Frisardi V, et al: Frailty syndrome and the risk of vascular dementia: The Italian longitudinal study on aging. Alzheim Dement 2013; 9:113e122 Kroboth PD, Salek FS, Pittenger AL, et al: DHEA and DHEA-S: a review. J Clin Pharmacol 1999; 39:327e348 Yesavage JA, Brink TL, Rose TL, et al: Development and validation of a geriatric depression screening scale: a preliminary report. J Psychiatr Res 1982e83; 17:37e49 Fukai S, Akishita M, Yamada S, et al: Association of plasma sex hormone levels with functional decline in elderly men and women. Geriatr Gerontol Int 2009; 9:282e289 Haren MT, Malmstrom TK, Banks WA, et al: Lower serum DHEAS levels are associated with a higher degree of physical disability and depressive symptoms in middle-aged to older African American women. Maturitas 2007; 57:347e360 Alexopoulos GS, Bruce ML, Hull J, et al: Clinical determinants of suicidal ideation and behavior in geriatric depression. Arch Gen Psychiatry 1999; 56:1048e1053 Nett ST, Jorge-Rivera JC, Myers M, et al: Properties and sexspecific differences of GABAA receptors in neurons expressing gamma1 subunit mRNA in the preoptic area of the rat. J Neurophysiol 1999; 81:192e203 Abadie JM, Wright B, Correa G, et al: Effect of dehydroepiandrosterone on neurotransmitter levels and appetite regulation of the obese Zucker rat. The Obesity Research Program. Diabetes 1993; 42:662e669 Li CM, Zhang X, Hoffman HJ, et al: Hearing impairment associated with depression in US adults, National Health and Nutrition Examination Survey 2005e2010. JAMA Otolaryngol Head Neck Surg 2014; 140:293e302 Lohman M, Dumenci L, Mezuk B: Sex differences in the construct overlap of frailty and depression: evidence from the Health and Retirement Study. J Am Geriatr Soc 2014; 62:500e505 Toffanello ED, Sergi G, Veronese N, et al: Serum 25-hydroxyvitamin D and the onset of late-life depressive mood in older men and women: the Pro.V.A. study. J Gerontol A Biol Sci Med Sci 2014; 69:1554e1561 Cappola AR, O’Meara ES, Guo W, et al: Trajectories of dehydroepiandrosterone sulfate predict mortality in older adults: the cardiovascular health study. J Gerontol A Biol Sci Med Sci 2009; 64:1268e1274 Goodyer IM, Park RJ, Netherton CM, et al: Possible role of cortisol and dehydroepiandrosterone in human development and psychopathology. Br J Psychiatry 2001; 179:243e249
9
Objective: Dehydroepiandrosterone sulfate (DHEAS) appears to have a protective effect against depression, but contrasting findings are available. Therefore, we investigated whether high serum DHEAS levels were associated with any protective effect on incident depression and incident severe depression in a representative group of elderly men and women. Methods: In a population-based cohort longitudinal study in the general community, 789 older participants without depression and cognitive impairment at the baseline were included, among 3,099 screened subjects. Serum DHEAS levels were determined based on blood samples; incident depression and severe depression were diagnosed by means of the Geriatric Depression Scale (GDS) and confirmed by geriatricians skilled in psychogeriatric medicine. Results: No baseline differences were found in GDS across age- and gender-specific tertiles of serum DHEAS. Over 4.4 years of follow-up, 137 new cases of depression were recorded. Of them, 35 among men and 64 in women were cases of incident severe depression. Cox’s regression analysis, adjusted for potential confounders, revealed that higher DHEAS levels were associated with reduced risk of incident depression irrespective of gender (HR: 0.30; 95% CI: 0.09e0.96; Wald c2 ¼ 4.09; df ¼ 1; p ¼ 0.04; women: HR: 0.31; 95% CI: 0.14e0.69; Wald c2 ¼ 8.37; df ¼ 1; p ¼ 0.004) and of severe incident depression only in men (HR: 0.25; 95% CI: 0.06e0.99; Wald c2 ¼ 4.05; df ¼ 1; p ¼ 0.04). Conclusion: Higher serum DHEAS levels were found to be significantly protective for the onset of depression irrespective of gender, whereas only in men was this association found also for incident severe depression. (Am J Geriatr Psychiatry 2015; -:-e-) Key Words: Serum DHEAS, depression, elderly
Received July 3, 2014; revised October 14, 2014; accepted October 30, 2014. From the Department of Medicine DIMED, Geriatrics Division (NV, MDR, FB, EDT, EM, GS), Department of Medical and Surgical Sciences (SZ), and Department of Cardiac, Thoracic and Vascular Sciences, Unit of Biostatistics, Epidemiology and Public Health (EP), University of Padova, Padova, Italy; National Research Council, Aging Branch (SZ, GC, EM), Institute of Neuroscience, Padova, Italy; Azienda Unità Locale Socio Sanitaria (ULSS) 16 (MC), Padova, Italy; and Internal Medicine Division (GB), Azienda Ospedaliera, Padova, Italy. Send correspondence and reprint request to Nicola Veronese, M.D., Department of Medicine - DIMED, Geriatrics Division, University of Padova, Via Giustiniani, 2 35128 Padova, Italy. e-mail: [email protected] Ó 2015 American Association for Geriatric Psychiatry http://dx.doi.org/10.1016/j.jagp.2014.10.009
Am J Geriatr Psychiatry
-:-, -
2015
1
DHEAS and Depression in the Elderly
INTRODUCTION
METHODS
Depression is widespread among elderly people, with a prevalence from 0.9% to 9.4% in communitydwelling people and higher in other settings like hospitals, nursing homes, and institutions.1 The presence of depression, particularly if severe, is strongly associated with disability, increased mortality, and poorer outcomes from physical illness.2 Research has shown that among possible factors affecting mood in the elderly, serum dehydroepiandrosterone sulfate (DHEAS) could be relevant. DHEAS is a sex steroid hormone mainly secreted by the adrenal glands but also in small amounts by the brain, where receptors for this hormone have selectively been found.3e5 DHEAS antidepressant and anxiolytic effects seem to be related to the agonistic activity on the s-1 receptor,6 the inhibition of g-aminobutyric acid A receptor directly and by decreasing pregnenolonesulfate concentrations in the brain,7 the activation of N-methyl-D-aspartate receptor,7e11 and the upregulation of brain-derived neurotrophic factor levels.12 Contrasting results have been reported in crosssectional studies about the association between serum DHEAS levels and depression, particularly in females.13e17 At the same time, only a few longitudinal studies have investigated the protective association of high serum DHEAS levels for depressive symptoms in older adults. Two of these studies found a significant association only in men, whereas another two failed to find any relationship.5,18e20 Finally, to the best of our knowledge, no study about DHEAS and incident severe depression are available, but this topic may be relevant because older people with severe depression have the highest suicide rate of any age group and a poorer prognosis of survival compared with younger subjects.21 The aim of the present study was thus to examine whether high serum DHEAS levels were associated with any protective effect on incident depression and incident severe depression and the possible role of gender in a representative group of elderly men and women over a lengthy follow-up (4.4 years).
Data Source and Subjects
2
Data for this analysis came from the Progetto Veneto Anziani (Pro.V.A.), an observational cohort study on the Italian population aged 65 years. Our study population included 3,099 age- and sex-stratified white participants (1,854 women and 1,245 men) randomly selected between 1995 and 1997 using a multistage stratified method. Sampling procedures and data collection methods are described elsewhere.22 Trained physicians and nurses examined participants at various clinics. The present study concerns the information collected on the incidence of depression and severe depression over a mean of 4.4 years (1.2 standard deviation) of follow-up. Among the 3,099 subjects considered, 601 participants were excluded because the Mini-Mental State Exam (MMSE) was below 24 and thus the Geriatric Depression Scale (GDS) may have been unreliable.23 Moreover, 500 participants were excluded because they were already found to be depressed at baseline, and 118 were taking antidepressant medications. Finally, 696 participants did not have DHEAS measurement available at baseline. The final sample thus consisted of 789 participants (247 died and 148 were lost at follow-up) (Fig. 1). Participants without DHEAS measurement at baseline did not differ from the participants included in the final analysis in age (72.9 6.4 versus 72.8 6.1 years, p ¼ 0.79) or MMSE score (27.34 1.76 versus 27.10 1.74, p ¼ 0.09). However, they showed a significantly lower GDS score (6.15 2.39 versus 6.79 2.00 points, p <0.0001). The ethical committees of Padua University and the Veneto Region’s Local Health Units (USSL) n. 15 and n. 18 approved our study protocol. Participants gave written informed consent to the study. Social and Medical Conditions Participants were examined at the city hospitals by trained physicians and nurses. Information was collected during a face-to-face individual interview on their social and medical conditions. Smoking habits were dichotomized as “current smoker” versus “never/former smoker” (if a subject had given
Am J Geriatr Psychiatry
-:-, -
2015
Veronese et al.
FIGURE 1. Flow chart of how the study population was selected: the Pro.V.A. study.
measured with hexokinase glucose-6-phosphate dehydrogenase (Dimension Vista System; Siemens, Milan, Italy). Total cholesterol was calculated using the enzymatic method. The diagnosis of vision impairment was made with the aid of the visual acuity test and that of hearing impairment with audiometry and confirmed by personnel experts in ophthalmology and otorhinolaryngology, respectively. Body weight and height were measured by trained physicians, and the body mass index (kg/m2) was calculated. Multidimensional Evaluation
up smoking at least a year earlier). Regular physical activity was defined as at least 4 hours a week in the previous month of at least moderate physical activity (brisk walking, cycling, gardening, dancing, or physical exercising). Alcohol drinking was categorized as “yes” or “no” during the last month. Educational level (the total number of years of school attended) was dichotomized as less than 5 versus 5 or mreo years of schooling. Monthly income was indicated as family income less than or equal to or more than 500 V. Any diseases were assessed by board-certified physicians, who examined all clinical details collected for each participant in the study, including their clinical history, symptoms self-reported by means of standardized questionnaires, medical and hospital records, blood tests, and physical examinations. A history of major diseases that could lead to depression was recorded for cardiovascular diseases (congestive heart failure, angina and myocardial infarction, stroke, and peripheral artery disease) and diabetes. Also, blood parameters that previous research has shown to be related to depression (fasting plasma glucose and cholesterol) were included in this work.20 Fasting plasma glucose was
Am J Geriatr Psychiatry
-:-, -
2015
A multidimensional evaluation was made by expert physicians in geriatric medicine both at baseline and follow-up. Functional status was assessed using an activities of daily living score.24 The Short Physical Performance Battery (SPPB)25 was derived from three objective physical function tests. Each test was scored from 0 (inability to complete the test) to 4 (highest level of performance). The scores from all three tests were summarized in a composite score of 0 to 12, higher scores reflecting better physical function. The three tests were as follows: 1. Tandem test: Participants were asked to maintain balance progressively side-by-side, in semitandem, and in full-tandem position. 2. Gait speed: The best performance achieved in two walks at usual pace along a 4-m corridor was recorded in meters/second. Participants were allowed to use canes or walkers. 3. Chair stand time: Participants were asked to stand up and sit down five times as quickly as possible, with their hands folded across their chest. The time taken to complete the test, in seconds, was recorded. Cognitive function was assessed by administering the 30-item MMSE. Frailty was assessed with the evaluation of five measurable items. Unintentional weight loss was defined as self-report of unintentional weight loss above 5 kg over the past year. Weakness and slow gait speed were defined, respectively, using the best value of handgrip strength and the best timed walk over 4-m at usual pace stratified according to sex and body mass index cut-offs as proposed by Fried et al.26 Exhaustion was defined using the question, “Do you feel full of energy?,” in the GDS
3
DHEAS and Depression in the Elderly questionnaire; low-energy expenditure was defined as the sum of leisure-time activities assessed in an ideal week of the previous month with a cut-off value below 383 kcal/wk in men and 270 kcal/wk in women.26,27 Participants were classified as frail if they met three or more of the five criteria. This variable was categorized as “frail” versus “not frail.” DHEAS Measurement Serum levels of DHEAS were determined by means of a chemiluminescent enzyme immunoassay with an interassay coefficient of variation of 10.9%. This method seems to be highly comparable with standard radioimmunoassay.20 DHEAS rather than DHEA was measured because it shows little circadian, monthly, and seasonal variations, making it more reliable as an epidemiologic marker.28 Because of the linear decrease with aging and the significant differences between genders, DHEAS tertiles were calculated using gender-specific tertiles according to the median age of our sample (75 years). In those younger than 75 years, the two cut-offs for dividing the sample in tertiles were 1.90 and 3.40 mmol/L in men and 1.00 and 2.00 mmol/L in women, whereas the correspondent values in those over 75 years were 1.60 and 2.80 and 0.80 and 1.60 mmol/L, respectively. Definition of Depression Any presence of depressive symptoms was assessed at baseline and at follow-up with the GDS, a 30-item self-reporting tool for identifying depression extensively validated for use in the elderly.29 GDS scores range from 0 to 30, with a score of at least 11 diagnostic for depression and a score of at least 20 plus limitation in activities of daily living due to depressive symptoms indicating severe depression.29 The diagnosis of depression and severe depression were 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.22 Statistical Analysis Participants’ characteristics were summarized using means (standard deviations) for continuous
4
variables and counts or percentages for categorical variables. For continuous variables, normal distributions were tested using the Shapiro-Wilk test. Because previous research has shown that serum DHEAS levels significantly differ between genders and there is a significant sex by DHEAS level interaction for the onset of depression,3,13,18,20 we tested if a similar interaction exists in our study using a Cox’s regression analysis. Because this interaction was significant, all data analyses were made using gender-specific tertiles of baseline DHEAS. The independent sample t test was used for comparing serum DHEAS values between men and women, without adjustment for age. Age-adjusted p values for trends were calculated, checking the differences between the means of the covariates by DHEAS group using analysis of variance. Differences in categorical variables were examined using the c2 test, adjusted for age using a logistic regression. Cox’s proportional hazard models were used to assess the association between gender-specific tertiles of DHEAS and incident and incident severe depression. Known factors associated with DHEAS and/or depression were considered for inclusion in the analysis, and consequently two models were built. Model 1 includes social conditions and items of the multidimensional evaluation; model 2 includes the covariates in model 1 and some medical conditions or blood parameters that previous research has shown to be associated with the development of depression. To explore whether a variable should be included as a predictor in the final survival model, the log-rank test of equality across strata was performed for all categorical variables and Cox’s univariate proportional hazards regression for all continuous variables. The predictors included in the final model were all the variables reaching p <0.20 in the univariate analyses. Collinearity among covariates was quantified by the variance inflation factor. Hazard ratios (HRs) and 95% confidence intervals (CIs) were used to compare rates of incident depression across DHEAS tertiles, taking those with lower DHEAS levels as reference. Similar analysis was made for incident severe depression. All analyses were performed using SPSS 21.0 for Windows (SPSS Inc., Chicago, IL). All statistical tests were two-tailed, and statistical significance was assumed for p <0.05.
Am J Geriatr Psychiatry
-:-, -
2015
Veronese et al.
TABLE 1. Baseline Characteristics According to Baseline DHEAS Tertiles: the Pro.V.A. Study Men
Age, years BMI, kg/m2 Social conditions Current smokers, % Physical activity 4 hours/weeks, % Alcohol drinking, % Education >5 years, % Monthly income 500 V, % Living alone, % Multidimensional evaluation GDS score (points) MMSE score (points) ADL score (points) SPPB total score (points) Medical conditions CVD, % Diabetes, % Vision impairment, % Hearing impairment, % Biohumoral tests Total cholesterol (mg/dL) FPG (mg/dL)
Tertile 1 (N [ 125)
Tertile 2 (N [ 115)
Tertile 3 (N [ 117)
pa
74.88 (7.36) 26.76 (3.82)
72.86 (6.23) 27.58 (3.42)
71.82 (5.33) 27.80 (4.16)
<0.0001b 0.24
8.8 40.8 87.2 25.0 53.6 8.1
13.9 47.0 91.3 25.4 43.5 4.4
23.1 36.8 90.6 21.7 46.2 7.7
0.007 0.37 0.30 0.28 0.44 0.59
6.58 (1.94) 26.06 (3.25) 5.73 (0.72) 10.18 (1.97)
6.48 (1.97) 25.91 (3.53) 5.78 (0.74) 10.52 (1.83)
6.69 (1.96) 26.09 (3.31) 5.77 (0.61) 10.75 (1.82)
0.65 0.42 0.97 0.67
19.2 19.2 36.5 75.2
20.0 14.8 36.0 81.7
14.5 7.7 22.2 80.3
0.69 0.01 0.27 0.13
223.58 (38.48) 106.98 (30.77)
222.22 (41.69) 108.96 (31.62)
231.71 (38.71) 103.54 (20.97)
0.34 0.30
Women
Age, years BMI, kg/m2 Social conditions Current smokers, % Physical activity 4 h/wk, % Alcohol drinking, % Education >5 years, % Monthly income 500 V, % Living alone, % Multidimensional evaluation GDS score (points) MMSE score (points) ADL score (points) SPPB total score (points) Medical conditions CVD, % Diabetes, % Vision impairment, % Hearing impairment, % Biohumoral tests Total cholesterol, mg/dL FPG, mg/dL
Tertile 1 (N [ 156)
Tertile 2 (N [ 133)
Tertile 3 (N [ 143)
pa
74.17 (6.18) 28.48 (5.00)
72.41 (5.54) 28.74 (5.23)
71.93 (5.72) 28.68 (4.74)
<0.0001b 0.95
3.8 25.0 56.4 15.6 76.6 22.6
7.1 25.6 59.4 10.6 70.8 21.2
8.4 18.2 68.5 11.3 67.2 21.8
0.94 0.07 0.02 0.13 0.10 0.92
6.90 (1.98) 25.51 (3.54) 5.60 (0.74) 9.08 (2.44)
6.78 (2.08) 25.41 (3.14) 5.57 (0.67) 9.45 (2.16)
6.87 (2.18) 25.47 (3.48) 5.65 (0.61) 9.80 (2.11)
0.91 0.44 0.64 0.32
10.9 12.8 30.8 53.2
6.8 15.0 33.8 62.4
9.1 16.8 23.1 53.1
0.93 0.33 0.51 0.19
246.66 (40.33) 103.17 (26.10)
245.61 (37.28) 106.83 (31.22)
249.93 (46.17) 108.29 (31.51)
0.72 0.39
Notes: Values are means with standard deviations in parentheses or percentages, as indicated. DHEAS tertiles were defined using a stratification for median of age (75 years) and gender. In those younger than 75 years, cut-offs were 1.90 and 3.40 in men and 1.00 and 2.00 mmol/L in women; the correspondent values in those over 75 years were 1.60 and 2.80 and 0.80 and 1.60 mmol/L, respectively. Tertile 1 indicates the lowest tertile and Tertile 3 the highest. BMI: body mass index; ADL: activity of daily living; CVD: cardiovascular diseases; FPG: fasting plasma glucose. a Unless otherwise specified, p values are adjusted for age using a general linear model or logistic regression, as appropriate. The c2 test with df ¼ 2. df for F-tests: in men df ¼ 2, 354 (among groups) and 2, 356 (within groups); in women df ¼ 2, 429 (among groups) and 2, 431 (within groups), respectively. b Univariate analysis of variance, not adjusted for age.
Am J Geriatr Psychiatry
-:-, -
2015
5
DHEAS and Depression in the Elderly
RESULTS The sample consisted of 789 community-dwelling subjects over age 65 years (357 men and 432 women) without depression and cognitive impairment at baseline. Mean ages were 73.2 6.5 for men and 72.9 5.1 years for women. Mean DHEAS levels in men were significantly higher than in women (independent samples t test; 2.93 1.96 mmol/L in men versus 1.66 1.26 mmol/L in women; df ¼ 787; t ¼ 11.003; p <0.0001). Moreover, a Cox’s regression analysis showed a significant interaction sex by baseline serum DHEAS levels for incident depression (HR: 0.89; 95% CI: 0.84e0.94; Wald c2 ¼ 15.60; df ¼ 1; p <0.0001). Table 1 shows the baseline characteristics of participants divided according to gender and baseline serum DHEAS levels. In men, there was a significant linear trend in percentage of current smokers but an inverse trend for the presence of diabetes. In women, only the presence of alcohol drinkers was significantly different across the tertiles, being higher in those with higher serum DHEAS levels (Table 1). No differences were evident for other variables investigated, including baseline GDS score (Table 1). Over a mean period of 4.4 years, 137 new cases of depression were recorded, with an incidence of 13.7% in men and 20.3% in women. Of these, 35 among men and 64 in women were cases of incident severe depression. As shown in Table 2, taking those with lower DHEAS as reference, Cox’s regression analysis shows that in the sample as a whole, participants with higher serum DHEAS levels had a significantly decreased risk of being depressed during the follow-up period, whereas no association was evident for incident severe depression. After dividing for gender, male participants with higher serum DHEAS levels had a significantly decreased risk of having both incident (HR: 0.30; 95% CI: 0.09e0.96; Wald c2 ¼ 4.09; df ¼ 1; p ¼ 0.04) and incident severe depression (HR: 0.25; 95% CI: 0.06e0.99; Wald c2 ¼ 4.05; df ¼ 1; p ¼ 0.04), whereas in women the association between high DHEAS levels and depression was present only for incident depression (HR: 0.31; 95% CI: 0.14e0.69; Wald c2 ¼ 8.37; df ¼ 1; p ¼ 0.004) (Table 2). Modeling serum DHEAS as a continuous variable, an increase of 1 mmol/L led to a significant decrease in having both
6
incident and incident severe depression in male participants, whereas in female participants these associations were not evident (Table 2). Among the other covariates considered, SPPB was a significant protective factor for incident depression independently from gender (HR: 0.73; 95% CI: 0.58e0.91; Wald c2 ¼ 7.70; df ¼ 1; p ¼ 0.001 in men and HR: 0.87; 95% CI: 0.75e0.99; Wald c2 ¼ 4.90; df ¼ 1; p ¼ 0.03 in women), whereas hearing impairment was a risk factor for incident depression (HR: 3.37; 95% CI: 1.07e12.70; Wald c2 ¼ 4.87; df ¼ 1; p ¼ 0.03 in men and HR: 2.71; 95% CI: 1.41e5.20; Wald c2 ¼ 8.91; df ¼ 1; p ¼ 0.003 in women). On the contrary, the presence of frailty produced a significant risk factor only in women (HR: 1.33; 95% CI: 1.04e1.62; Wald c2 ¼ 8.02; df ¼ 1; p ¼ 0.004).
DISCUSSION Our large population-based prospective study found that higher serum DHEAS levels reduced the risk of incident depression over a 4.4-year follow-up in both genders, whereas the association with severe depression was found only in men. The few existing studies about this topic gave contrasting results. Two studies, in fact, did not find any association between serum DHEAS levels and risk of depression,5,30 whereas another two studies found an association only in men.18,20 The differences between our study and others with negative findings may be explained by methodologic issues. First, we used the 30-item GDS for detection of depression, and the diagnosis was confirmed by geriatricians skilled in psychogeriatric medicine, whereas in other studies shorter forms of the GDS or other tools (like the Center for Epidemiologic Studies Depression Scale) were used alone. Second, we did not include subjects with dementia or cognitive impairment. Participants with cognitive impairment at baseline, at higher risk for dementia than depression, could create a bias in assessing the relationship between serum DHEAS levels and depression. Moreover, the reliability of tools assessing depressed subjects with dementia or cognitive impairment is still a matter of discussion.23 Also, Haren et al.31 found that serum DHEAS is significantly associated with depression only in subjects without cognitive impairment, confirming that
Am J Geriatr Psychiatry
-:-, -
2015
Am J Geriatr Psychiatry
TABLE 2. Association Between Baseline DHEAS Tertiles, Incident, and Severe Depression According to Gender: the Pro.V.A. Study
-:-, -
2015
Model 2c HR (95% CI)
No. of Participants
53 51 33
281 248 260
1 [reference] 0.75 (0.47e1.20) 1.45 0.40 (0.23e0.70) 10.34 0.88 (0.78e0.99) 4.64
0.23 0.001 0.03
1 [reference] 0.78 (0.48e1.25) 1.08 0.35 (0.20e0.63) 12.32 0.88 (0.78e0.99) 4.42
0.30 <0.0001 0.04
1 [reference] 0.81 (0.48e1.35) 0.68 0.35 (0.19e0.65) 11.39 0.87 (0.77e0.999) 3.87
0.41 0.001 0.049
41 33 25
281 248 260
1 [reference] 1.13 (0.65e1.98) 0.19 0.63 (0.33e1.19) 2.05 0.92 (0.80e1.07) 1.13
0.67 0.15 0.29
1 [reference] 1.12 (0.63e1.99) 0.16 0.53 (0.27e1.04) 3.40 0.92 (0.79e1.06) 1.32
0.69 0.07 0.25
1 [reference] 1.22 (0.66e2.25) 0.39 0.56 (0.28e1.13) 2.63 0.92 (0.78e1.09) 0.99
0.53 0.10 0.32
20 19 10
125 115 117
1 [reference] 0.81 (0.36e1.82) 0.27 0.34 (0.13e0.86) 5.23 0.86 (0.74e1.00) 3.68
0.60 0.02 0.06
1 [reference] 0.57 (0.23e1.47) 1.35 0.24 (0.08e0.71) 6.70 0.83 (0.70e0.98) 5.05
0.57 0.01 0.03
1 [reference] 0.75 (0.27e2.15) 0.28 0.30 (0.09e0.96) 4.09 0.77 (0.63e0.94) 6.39
0.60 0.04 0.01
16 12 7
125 115 117
1 [reference] 1.03 (0.40e2.66) 0.05 0.38 (0.13e0.96) 2.89 0.86 (0.72e1.03) 2.61
0.94 0.02 0.11
1 [reference] 0.73 (0.24e2.12) 0.31 0.27 (0.08e0.96) 4.07 0.79 (0.65e0.97) 5.03
0.58 0.04 0.03
1 [reference] 0.81 (0.23e2.81) 0.12 0.25 (0.06e0.99) 4.05 0.71 (0.55e0.91) 7.35
0.74 0.04 0.007
33 32 23
156 133 143
1 [reference] 0.72 (0.40e1.28) 1.26 0.33 (0.13e0.86) 5.55 0.89 (0.74e1.09) 1.26
0.26 0.02 0.26
1 [reference] 0.82 (0.46e1.49) 0.41 0.30 (0.14e0.64) 9.63 0.94 (0.77e1.14) 0.41
0.52 0.002 0.52
1 [reference] 0.83 (0.44e1.58) 0.32 0.31 (0.14e0.69) 8.37 0.99 (0.80e1.22) 0.008
0.57 0.004 0.93
25 21 18
156 133 143
1 [reference] 1.22 (0.61e2.47) 0.32 0.78 (0.35e1.73) 0.37 1.03 (0.80e1.33) 0.06
0.58 0.54 0.81
1 [reference] 1.32 (0.64e2.70) 0.56 0.54 (0.23e1.27) 2.00 1.07 (0.82e1.39) 0.24
0.45 0.16 0.62
1 [reference] 1.50 (0.69e3.27) 1.03 0.59 (0.24e1.45) 1.31 1.16 (0.87e1.53) 1.02
0.31 0.25 0.31
Wald c2
pa
Wald c2
pa
Wald c2
pa
7
Notes: Unless otherwise specified, values are HR with 95% CI in parentheses. DHEAS tertiles were defined using a stratification for median of age (75 years) and gender. In those younger than 75 years, the cut-offs were 1.90 and 3.40 in men and 1.00 and 2.00 mmol/L in women; the correspondent values in those over 75 years were 1.60 and 2.80 and 0.80 and 1.60 mmol/L. Tertile 1 indicates the lowest tertile and Tertile 3 the highest. a Each p value was based on a Wald c2 test with df ¼ 1. b Model 1: adjusted for age and body mass index, activities of daily living score, MMSE score, educational level, smoking status, monthly income, physical activity level, hearing and vision impairment, living alone, SPPB, and presence of frailty. In the sample as whole, model 1 also comprehends gender as a covariate. c Model 2: adjusted for covariates in model 2 and baseline GDS score, presence of cardiovascular diseases, diabetes, and serum levels of fasting plasma glucose and total cholesterol.
Veronese et al.
Entire sample Incident depression Tertile 1 Tertile 2 Tertile 3 DHEAS (For increase of 1 mmol/L) Incident severe depression Tertile 1 Tertile 2 Tertile 3 DHEAS (For increase of 1 mmol/L) Men Incident depression Tertile 1 Tertile 2 Tertile 3 DHEAS (For increase of 1 mmol/L) Incident severe depression Tertile 1 Tertile 2 Tertile 3 DHEAS (For increase of 1 mmol/L) Women Incident depression Tertile 1 Tertile 2 Tertile 3 DHEAS (For increase of 1 mmol/L) Incident severe depression Tertile 1 Tertile 2 Tertile 3 DHEAS (For increase of 1 mmol/L)
Age-adjusted HR (95% CI)
Model 1b HR (95% CI)
No. of Events
DHEAS and Depression in the Elderly cognitive impairment is an important bias to consider in the assessment of depression in older people. The association between serum DHEAS and depression could be conditioned by gender differences as shown also by our study when considering severe depression. The severity of depression may be particularly important considering that severe depression, besides being the strongest predictor of the course of suicidal ideation, is also a strong risk factor for disability and mortality in older people.32 Our study suggests that in men, elevated serum DHEAS levels are necessary for avoiding more severe forms of depression. On the contrary, in women no association was observed between serum DHEAS levels and incidence of severe depression. The same figure was observed also when serum DHEAS was modeled as a continuous variable. Some hypotheses can be formulated to explain this result, borrowing some findings from animal models. First, in rats, male and female brain areas have shown different sensitivity to g-aminobutyric acid A receptor ligands, including DHEAS.33 Moreover, a threshold effect of DHEAS could also be involved. Supplementation with DHEA was able to increase serotonin brain levels in rats only for the highest doses, whereas no effects were observed for the other doses of treatment.34 It may be possible therefore that the association between serum DHEAS and severe depression was seen only in men who have higher DHEAS levels and that DHEAS levels even higher than those of women in the third tertile are necessary to reduce the risk for severe depression. Other than hearing impairment, an important and well-known risk factor for depression,35 frailty is another important risk factor for depression present only in women according to recent research.36 On the contrary, high SPPB scores were protective for the onset of depression, confirming that higher performance levels are associated with a lower risk of depression.37 Our findings suggest that serum
DHEAS concentrations have a protective role in the elderly, but its role could be attenuated in some conditions such as in sedentary lifestyle and frailty, and so further specific research about these populations is necessary. Several limitations of this study need to be considered. First, some subjects potentially eligible for our study did not have DHEAS measurements and had significant lower GDS scores at baseline. Second, DHEAS was measured only at baseline, and it is known that this hormone has unpredictable changes.38 Moreover, we did not estimate cortisol levels because it is known that the cortisol-to-DHEAS ratio is an independent predictor of depression.39 Finally, even if we used a reliable method for diagnosis of depression, we did not use the gold standard method, that is, a structured clinical interview combined with a clinical interview by an experienced mental health professional; therefore, other possible conditions similar to depression (e.g., thyroid diseases) were not fully investigated. In conclusion, higher serum DHEAS levels were found to be significantly protective for the onset of depression irrespectively of gender. However, only in men was this association also found for incident severe depression. The authors are grateful to all interviewers, nurses, and physicians who took part in the study. The authors also acknowledge Emily Orr for her competency in English. The Pro.V.A. study was funded by the Fondazione Cassa di Risparmio di Padova e Rovigo, University of Padova; Azienda Unità Locale Socio Sanitaria 15 and 18 of the Veneto Region, and by a grant from the Veneto Regional Authority (Ricerca Sanitaria Finalizzata n.156/03). None of the authors has any financial arrangements, organizational affiliations, or other relationships that might give rise to any conflict of interest regarding the subject matter of this article.
References 1. Djernes JK: Prevalence and predictors of depression in populations of elderly: a review. Acta Psychiatr Scand 2006; 113: 372e387 2. Rodda J, Walker Z, Carter J: Depression in older adults. BMJ 2011; 343:d5219
8
3. Orentreich N, Brind JL, Vogelman JH, et al: Long-term longitudinal measurements of plasma dehydroepiandrosterone sulfate in normal men. J Clin Endocrinol Metab 1992; 75:1002e1004 4. Guazzo EP, Kirkpatrick PJ, Goodyer IM, et al: Cortisol, dehydroepiandrosterone (DHEA), and DHEA sulfate in the cerebrospinal
Am J Geriatr Psychiatry
-:-, -
2015
Veronese et al.
5.
6.
7.
8. 9. 10.
11.
12.
13.
14.
15.
16.
17. 18. 19.
20.
21. 22.
fluid of man: relation to blood levels and the effects of age. J Clin Endocrinol Metab 1996; 81:3951e3960 Mazat L, Lafont S, Berr C, et al: Prospective measurements of dehydroepiandrosterone sulfate in a cohort of elderly subjects: relationship to gender, subjective health, smoking habits, and 10year mortality. Proc Natl Acad Sci USA 2001; 98:8145e8150 Urani A, Roman FJ, Phan VL, et al: The antidepressant-like effect induced by sigma(1)-receptor agonists and neuroactive steroids in mice submitted to the forced swimming test. J Pharmacol Exp Ther 2001; 298:1269e1279 Majewska MD, Demirgören S, Spivak CE, et al: The neurosteroid dehydroepiandrosterone sulfate is an allosteric antagonist of the GABAA receptor. Brain Res 1990; 526:143e146 van Broekhoven F, Verkes RJ: Neurosteroids in depression: a review. Psychopharmacology 2003; 165:97e110 Kulkarni SK, Dhir A: s-1 receptors in major depression and anxiety. Expert Rev Neurother 2009; 9:1021e1034 Kinouchi K, Maeda S, Saito K, et al: Effects of d- and l-pentazocine on the release and uptake of norepinephrine in rat brain cortex. Res Commun Chem Pathol Pharmacol 1989; 63:201e213 Monnet FP, Mahe V, Robel P, et al: Neurosteroids, via s receptors, modulate the [3H]norepinephrine release evoked by N-methyl-Daspartate in the rat hippocampus. Proc Natl Acad Sci USA 1995; 92:3774e3778 Naert G, Maurice T, Tapia-Arancibia L, et al: Neuroactive steroids modulate HPA axis activity and cerebral brain-derived neurotrophic factor (BDNF) protein levels in adult male rats. Psychoneuroendocrinology 2007; 32:1062e1078 Berr C, Lafont S, Debuire B, et al: Relationships of dehydroepiandrosterone sulfate in the elderly with functional, psychological, and mental status, and short-term mortality: a French communitybased study. Proc Natl Acad Sci USA 1996; 93:13410e13415 Barrett-Connor E, von Muhlen D, Laughlin GA, et al: Endogenous levels of dehydroepiandrosterone sulfate, but not other sex hormones, are associated with depressed mood in older women: the Rancho Bernardo study. J Am Geriatr Soc 1999; 47:685e691 Morsink LF, Vogelzangs N, Nicklas BJ, et al: Associations between sex steroid hormone levels and depressive symptoms in elderly men and women: results from the Health ABC study. Psychoneuroendocrinology 2007; 32:874e883 T’Sjoen GG, De Vos S, Goemaere S, et al: Sex steroid level, androgen receptor polymorphism, and depressive symptoms in healthy elderly men. J Am Geriatr Soc 2005; 53:636e642 Erdincler D, Bugay G, Ertan T, et al: Depression and sex hormones in elderly women. Arch Gerontol Geriatr 2004; 39:239e244 Goldman N, Glei DA: Sex differences in the relationship between DHEAS and health. Exp Gerontol 2007; 42:979e987 Yaffe K, Ettinger B, Pressman A, et al: Neuropsychiatric function and dehydroepiandrosterone sulfate in elderly women: a prospective study. Biol Psychiatry 1998; 43:694e700 Michikawa T, Nishiwaki Y, Nakano M, et al: Higher serum dehydroepiandrosterone sulfate levels are protectively associated with depressive symptoms in men, but not in women: a community-based cohort study of older Japanese. Am J Geriatr Psychiatry 2013; 21:1154e1163 Manthorpe J, Iliffe S: Suicide in later life: public health and practitioner perspectives. Int J Geriatr Psychiatry 2010; 25:1230e1238 Corti MC, Guralnik JM, Sartori L, et al: The effect of cardiovascular and osteoarticular diseases on disability in older Italian men
Am J Geriatr Psychiatry
-:-, -
2015
23.
24. 25.
26.
27.
28. 29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
and women: rationale, design, and sample characteristics of the Progetto Veneto Anziani (PRO.V.A.) study. J Am Geriatr Soc 2002; 50:1535e1540 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 Katz S, Downs TD, Cash HR, et al: Progress in development of the index of ADL. Gerontologist 1970; 10:20e30 Guralnik JM, Simonsick EM, Ferrucci L, et al: A short physical performance battery assessing lower extremity function: association with self-reported disability and prediction of mortality and nursing home admission. J Gerontol 1994; 49:M85eM94 Fried LP, Tangen CM, Walston J, et al: Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci 2001; 56: M146eM156 Solfrizzi V, Scafato E, Frisardi V, et al: Frailty syndrome and the risk of vascular dementia: The Italian longitudinal study on aging. Alzheim Dement 2013; 9:113e122 Kroboth PD, Salek FS, Pittenger AL, et al: DHEA and DHEA-S: a review. J Clin Pharmacol 1999; 39:327e348 Yesavage JA, Brink TL, Rose TL, et al: Development and validation of a geriatric depression screening scale: a preliminary report. J Psychiatr Res 1982e83; 17:37e49 Fukai S, Akishita M, Yamada S, et al: Association of plasma sex hormone levels with functional decline in elderly men and women. Geriatr Gerontol Int 2009; 9:282e289 Haren MT, Malmstrom TK, Banks WA, et al: Lower serum DHEAS levels are associated with a higher degree of physical disability and depressive symptoms in middle-aged to older African American women. Maturitas 2007; 57:347e360 Alexopoulos GS, Bruce ML, Hull J, et al: Clinical determinants of suicidal ideation and behavior in geriatric depression. Arch Gen Psychiatry 1999; 56:1048e1053 Nett ST, Jorge-Rivera JC, Myers M, et al: Properties and sexspecific differences of GABAA receptors in neurons expressing gamma1 subunit mRNA in the preoptic area of the rat. J Neurophysiol 1999; 81:192e203 Abadie JM, Wright B, Correa G, et al: Effect of dehydroepiandrosterone on neurotransmitter levels and appetite regulation of the obese Zucker rat. The Obesity Research Program. Diabetes 1993; 42:662e669 Li CM, Zhang X, Hoffman HJ, et al: Hearing impairment associated with depression in US adults, National Health and Nutrition Examination Survey 2005e2010. JAMA Otolaryngol Head Neck Surg 2014; 140:293e302 Lohman M, Dumenci L, Mezuk B: Sex differences in the construct overlap of frailty and depression: evidence from the Health and Retirement Study. J Am Geriatr Soc 2014; 62:500e505 Toffanello ED, Sergi G, Veronese N, et al: Serum 25-hydroxyvitamin D and the onset of late-life depressive mood in older men and women: the Pro.V.A. study. J Gerontol A Biol Sci Med Sci 2014; 69:1554e1561 Cappola AR, O’Meara ES, Guo W, et al: Trajectories of dehydroepiandrosterone sulfate predict mortality in older adults: the cardiovascular health study. J Gerontol A Biol Sci Med Sci 2009; 64:1268e1274 Goodyer IM, Park RJ, Netherton CM, et al: Possible role of cortisol and dehydroepiandrosterone in human development and psychopathology. Br J Psychiatry 2001; 179:243e249
9