It was a cross-sectional descriptive study, conducted from May 2012 to December 2013 among elderly people (age ≥ 60) [3] who were enrolled in old age homes (OAHs) and from general population living in Lucknow city. Two sets of elderly population were covered with inclusion criteria of age ≥60 years, participants residing in community or OAHs for ≥6 months (permanent or temporary resident). Mini mental state examination (MMSE) was the screening test performed on each participant to assess cognitive impairment and those who scored ≥ 20 on MMSE were taken to be part of study [4–6]. The purpose of MMSE was to show that the Geriatric Depression Scale (GDS) which has been used among participants is not valid for cognitively impaired subjects (MMSE <20) [7].

Participants with serious morbidities as mental illness, cancer, advanced heart disease or chronic renal disease on the basis of their current medical reports, were excluded from the study. Old age homes were of two types: 1) private type in which inmates paid some amount per month & 2) public type which was free of cost. Out of a total 5 OAHs, two were operated by government sponsored NGO’s, two were private & one was government operated. All of these were situated in urban area of Lucknow. Elderly people living in OAHs were contacted after taking permission from In-charges of old age homes and verbal consent from participants. All of them gave consent to participate in study. Of these 6 were excluded from the study because they were unable to get a score of ≥ 20 on MMSE. Thus a total 101 subjects were included in the study. Using prevalence of psychosocial problem in elderly people as 42% [8] sample size from community was 141. Multistage sampling technique was used to obtain required number of participants from the community.

In 1^st stage, out of 9 Legislative Constituencies (LC) of Lucknow district, 5 LC were included in the study which represented the urban population. In 2^nd stage a list of polling stations was obtained from the website of Chief Electoral Officer, Uttar Pradesh [9] for each of the Legislative Constituencies. Out of these 28 polling stations were selected from each LC & 29 polling stations from largest LC by the method of simple random sampling.

In 3^rd stage from each selected polling station, a list of all the voters was taken from the website of Chief Electoral Officer, Uttar Pradesh, India. A list of voters aged ≥60 years was separated & 5 elderly people were selected from each polling station by simple random sampling technique. Of five selected participants first was approached & if the individual gave consent and scored ≥ 20 on MMSE, then interview was further continued, if not the next individual was approached and so on. If first individual gave consent and scored ≥ 20 on MMSE, rest were not contacted from that poll station. Thus from each polling station, only one person was selected. A total of 204 subjects were approached in community, of which 49 refused to participate (40 refusals were by their by care takers due to their personal reasons and only 9 refusals were by participants). Fourteen subjects were excluded as they were unable to get score ≥ 20 on MMSE. Thus a total of 141 subjects were finally included as study participants.

Tool which was used to screen out Depression among participants was Geriatric Depression Scale – Hindi Version (GDS-H) [10]. Participants with any systemic chronic/serious disease were considered as diagnosed cases if they showed prescription of government health facility (at least secondary level) or specialist private practitioner or investigation reports with drugs that were being used regularly along with clinical sign and symptoms.

Dependent variable: Depression.

Independent variables: Age, Number of children, Religion, Caste, Female sex, marital & living status, Education, Financial Dependency, Systemic Morbidities, Living Status, Feeling of Neglect Feeling of Loneliness.

The schedule was pretested on elderly people of ≥ 60 years living in community. Necessary modifications were made in the schedule to overcome the difficulties.

The level of significance was set at p < 0.05. Appropriate statistical software was used for analysis. The study was approved by the ethical review board of King George Medical University prior to study.

Demography of elderly people was studied in terms of Age group, Sex, Religious group, Marital status, Educational profile, Children and Morbidity Status for both the groups. In Community, the elderly population comprised 58.2%, 32.6% and 9.2% whereas in old age homes the elderly were 28.7%, 37.6% and 33.7% in age groups 60 to <70 years, 70 to <80 years in ≥80 years respectively. Majority in both the groups were male (70.9% in Community and 53.5% in OAHs). Majority of elderly people in both the groups belonged to Hindu Religious group i.e., 80.9% in Community and 99% in OAHs, 78.7% elderly people in the community and 31.7% in OAHs were married, 18.4% in Community and 54.5% in OAHs were widow/widower, 1.4% in Community and 7.9% in OAHs were divorced/separated and the rest 1.4% in Community and 5.9% in OAHs were unmarried. In Community 16.3% elderly people were illiterate followed by 28.4 % 10^th pass, 24.8% Graduate/Diploma and 21.3 % Postgraduate & above while in OAHs 21.8 were illiterate, 14.9% primary pass, 11.9% 10^th pass, 28.7% Graduate/Diploma and 22.8% Postgraduate & above. In Community, 95.7% individuals and 57% in OAHs had at least one male child and the rest in both the groups had either no child or only female child. 58.9% of elderly people in the community have at least one type of diagnosed systemic morbidity while it was 81.2% in OAHs.

In community most frequent morbidity was hypertension (17.7%) followed by eye related morbidity (14.9%) while 41.1 % elderly had no diagnosed morbidity. In OAH out of total the musculoskeletal morbidity (33.7%) was most frequent followed by eye related morbidity (24.8%) and 18.8 % had no diagnosed morbidity [Table/Fig-1].

Out of 242 study subjects 34.0 % were diagnosed as depressed in age group ≥ 80 years. Depression among elderly females was 28.4 % while among males it was 16.2 %. Elderly people those who were single (living without partner) 30.3% were depressed, on the other hand it was 14.0% among elderly who were married & living with partner. Elderly people with no child or only female child were found to be more depressed (32.0%). Depression was 27.7% among elderly people residing in OAHs while it was 15.6% those residing at their own homes. Among financially independent elderly 10.0 % were depressed while it was 32.3% among those who were totally dependent. Relation of depression with age group, sex, marital & living status, children, residential status, financial dependency and with feeling of loneliness was statistically significant [Table/Fig-2].

On multivariate analysis education and financial dependency were found to be primary predictors for depression [Table/Fig-3].

Among the elderly population residing in the community it was found that the number of male respondents was much higher than female, the difference mainly being because males were more eager to participate in the study as compared to females. Being a male dominating society there were more participation of males than females. The reason for total absence of Sikhs and only a minor population of Muslims in the OAHs may be due to joint family system still being present in these religions. Similar findings were also observed in studies conducted in Lucknow [17,18].

Among elderly living in community 58.9% had one or more morbidities and almost similar findings were observed in a study conducted in Andhra Pradesh [19]. On the other hand in OAHs, 81.2% of elderly people had one or more morbidities which was higher than subjects residing in community.

Many studies conducted in different part of India like in Chandigarh [20], in urban area of Udaipur [8], among urban elderly of Bengaluru [21] and in OAHs of Calcutta [22], showed higher prevalence of different morbidities as compared to present study which may be attributed to the fact that only those diagnosed earlier were considered as case in the present study.

Prevalence of depression among elderly people residing in community was 15.6 % and in OAHs was 27.7 %. Studies in Indian setting showed that prevalence of depression among elderly in community ranges from 11.6% to 31.1% [23–25]. While most of the studies conducted in other part of world showed community prevalence ranging from 17.3% to 22.07% [26–28]. A review based on 122 studies in the Western countries reported the prevalence of major depression between the range of 0.9 % to 9.4 % in private households and from 14 % to 42 % in institutional settings [2,29], whereas a study in United Kingdom on residents in OAHs found 40% prevalence [30]. The prevalence rates have varied in different studies and these variation were most probably due to clinical criteria used to measure depression, the cultural settings and the population diversity [31]. There is evidence indicating higher rates of depression in OAHs compared to the community [27,32].

The family has been consistently reported to be a source of strength and security for the elderly people in providing them with social, financial and emotional support [33] which is associated with their well-being. Therefore it is obvious that study subjects who were residents of OAHs reported higher prevalence of depression due to their feeling of isolation (loneliness) from their families [34] and can also be reason for higher rate of morbidities in OAHs as compare to community.

Result showed that female respondents are more likely to be depressed and elderly married & living with partner have reduced risk of being depressed and similar findings were found in other studies [17,35–37].

Present study also showed that education and financial dependency were primary predictors whereas age group, sex, marital & living status, children, residential status, feeling of loneliness were found to be secondary predictors of depression. Studies conducted in different part of world also showed similar finding [38–40].

Depression was more common among elderly living in OAHs (27.7%) as compare to those living in community (15.6%). Similarly it was more among those who were single (living without partner) than those who were married and living with their partner. Elderly females were more depressed than elderly males. Hypertension, musculoskeletal morbidities & eye related morbidities were most frequent diagnosed morbidities among elderly people. Education and Financial Dependency were found to be primary predictors of depression.

Among all important predictors found in present study only financial dependency, education, are only modifiable and primary predictors for depression. So based on this following recommendations should be emphasized - Revision old age pension based on inflation, Government sponsored or public-private partnership based OAHs/ day care centre/recreational centre with better infrastructure and facilities should be established at block/sub-district level especially for economically vulnerable sections of society with facilities like health check-up (physical as well as mental), diagnosis and management for different morbidities. Recreational activities/games to improve literacy, mental activities and social relationships to overcome their loneliness.

I hereby declare that some part of article titled “Screening Depression among Urban Elderly in a City of Southeast Asia” has also been presented as poster under title “Study of Depression in Geriatric Population: Old Age Home and Community in Lucknow India” in 20th IEA World Congress of Epidemiology, Alaska, USA on Sunday, 17 August 2014 by Prof. Uday Mohan, Department of Community Medicine & Public Health, K G Medical University, UP, Lucknow, India who was the part of original study as one of the co-author.