Loneliness and Its Determinants among the Elderly in an Urban Field Practice Area of a Medical College in North India: A Community-Based Cross-sectional Study

Abstract

Background and Objective:

Mental health disorders affect over 20% of the elderly, with depression being one of the most prevalent conditions. Loneliness plays a significant role in contributing to depression, leading to negative effects on physical, cognitive, and emotional well-being. It also reduces overall quality of life and increases the risk of mortality. Despite its impact, research on gero-psychiatry remains limited, often focusing on clinical settings rather than community-based studies. This study aimed to assess the prevalence of loneliness among older adults and explore the associated factors. The present study was conducted as part of a larger study assessing depression among the elderly. However, the current manuscript specifically focuses on loneliness as an independent entity of public health importance.

Materials and Methods:

A community-based cross-sectional study was conducted in Mehrauli, a field practice area for Lady Hardinge Medical College, from January 2023 to February 2024. A total of 338 elderly individuals participated. Loneliness was measured using the UCLA scale and functional ability was evaluated using the Katz Index.

Results:

Most participants (74.6%) were aged 60–69 years (mean age 66.28 ± 7.12), and 53.3% were women. The overall prevalence of loneliness was 52.4%. Significant associations were found between loneliness and factors such as age, gender, marital status, education, family structure, socioeconomic status, chronic illness, and functional impairment (P < 0.05).

Conclusion:

The high prevalence of loneliness among urban elderly highlights the need for timely interventions. Policies promoting social security, mental health awareness, and community engagement can help reduce its impact. NGOs and government initiatives should work together to support elderly well-being.

INTRODUCTION

Aging is an inevitable biological process that brings multiple physical, psychological, and social transitions that influence health and overall quality of life. With advancing age, individuals often experience chronic illnesses, loss of independence, bereavement, and diminishing social networks. These challenges can lead to emotional distress and a sense of disconnection from others. Loneliness, defined as the subjective feeling of being socially isolated or lacking meaningful companionship, has emerged as an important public health issue in the elderly.[1] It represents not only a psychological concern but also a determinant of several health outcomes that affect both morbidity and mortality.[1]

The global demographic shift toward an aging population has amplified the importance of geriatric mental health. According to the World Health Organization, the proportion of people aged 60 years and above is projected to nearly double—from 12% in 2015 to 22% by 2050—most of whom will reside in low- and middle-income countries.[2] In India, older adults constituted about 8% of the population in 2012, and this proportion is expected to increase to nearly 20% by 2050.[3] This rapid demographic transition is accompanied by social and economic challenges, including changes in living arrangements, limited access to healthcare, and reduced social interaction—all of which contribute to the growing problem of loneliness among the elderly.

Evidence indicates that loneliness in older adults is associated with a wide range of adverse health effects. Studies have demonstrated its strong links with mental disorders such as depression and anxiety, as well as with physical conditions including cardiovascular diseases, cognitive decline, and functional disability.[4,5,6] Persistent loneliness may even carry health risks comparable to those of established behavioral risk factors like smoking or obesity.[7] It can also lead to poor self-care, lower adherence to medical treatment, and increased dependence on healthcare services, placing additional strain on families and healthcare systems.

Traditionally, Indian society has been characterized by close-knit joint family structures and strong intergenerational bonds that offered emotional and social support to older members. However, urbanization, industrial growth, migration for employment, and the growing prevalence of nuclear families have altered these patterns.[8,9] As a result, many elderly individuals, particularly in urban areas, now live alone or experience reduced family contact. Factors such as widowhood, declining physical health, and restricted mobility further intensify the sense of isolation.[10,11] The erosion of traditional support systems and limited community engagement make loneliness an emerging yet under-recognized concern in India’s aging population.

Despite its significance, loneliness among older adults remains relatively underexplored in Indian research. Most available studies have focused on related issues such as depression or cognitive decline, while loneliness has often been treated as a secondary finding.[12] Moreover, there is a paucity of community-based data, particularly from urban settings where sociodemographic transitions are rapid. Identifying the magnitude of loneliness and the factors influencing it is therefore essential to design effective, culturally relevant interventions that address the emotional and social needs of older persons.

The findings of this study are expected to provide insights that can guide community-level mental health promotion, strengthen family and social support systems, and inform policies aimed at fostering healthy and active aging in India.

Objective: To estimate the prevalence of loneliness among the elderly in an Urban Field Practice Area of a Medical College in North India and to study the various factors associated with loneliness.

MATERIALS AND METHODS

Study design: This was a quantitative analytical cross-sectional study conducted among the elderly residing in the Mehrauli area of South Delhi. Study Period: It was conducted from January 2023 to February 2024. Study participants: Elderly people aged 60 or more were included as subjects for the study. If there were multiple elderly individuals (≥60 years) in a single household, both were included in the study. If a household did not have any elderly individuals, the next house was selected. The purpose of the study was explained, and written informed consent was obtained from the participants. Inclusion criteria: All elderly individuals who gave consent, of either gender, aged ≥60 years, residing in the Mehrauli area of South Delhi for at least 1 year. Exclusion criteria: Participants’ cognitive ability to comprehend and respond was assessed during the initial interaction using simple orientation and communication checks (e.g., ability to understand questions and provide coherent responses). Those unable to do so were excluded.

Consent: Informed consent was obtained from the study subjects in the language that they understood. Privacy, modesty, and confidentiality were ensured. Methodology: Participants who did not consent during the initial visit were revisited up to two additional times to seek participation. Those who refused to participate or could not be reached after repeated visits, including households that remained locked during two separate attempts, were categorized as non-respondents.

Once written informed consent was obtained, a brief general health assessment was performed to facilitate participant engagement. Subsequently, information was collected using a pre-tested structured interview schedule to record sociodemographic and socioeconomic details.

Ethical approval was obtained from the Institutional Ethics Committee of Lady Hardinge Medical College, New Delhi.

Sampling method and sample size: This study is a part of a bigger study. Hence, the sample size was calculated according to the primary objective, which is to estimate the prevalence of depression among the geriatric population in the Mehrauli area of South Delhi. Sample size was calculated assuming the prevalence of depression as 40%[4] at 95% level of significance and a 15% relative error, 10% non-response rate, and a 1.2 design effect; the final sample size was 338. To assess its suitability for the loneliness outcome, a post-hoc evaluation was conducted. With an observed loneliness prevalence of 52.4% and a total sample of 338 participants, the standard error was calculated as 0.027, resulting in a 95% confidence interval ranging from 47.1% to 57.7%, indicating a precision of approximately ± 5.3%. These findings suggest that the sample size was adequate for estimating loneliness prevalence with reasonable precision. In addition, the number of observed events was sufficient to allow reliable multivariable logistic regression analysis.

Although formal cluster sampling was not used, a design effect of 1.2 was applied to account for potential intra-household correlation arising from the inclusion of more than one elderly participant per household. In the absence of prior estimates, a conservative design effect was assumed, and the intracluster correlation coefficient (ICC) was not calculated.

There are a total of eight wards in the Mehrauli area, out of which three wards (Ward number 1, 2 and 3) were selected by the simple random sampling method. Sampling units were households, and the study units were geriatric members of the family. The total number of individuals was 9700, 33000 and 11300 in ward number 1, 2, and 3, respectively. In each ward, the systematic random sampling method was used to select households. If more than 1 elderly person was present in a single household, both were selected.

Study instrument: A self-designed, pre-tested interview schedule containing identification details, sociodemographic profile and comorbidity status based on self-reporting or available medical records. Standard validated study tools used were: Modified Kuppuswamy scale (2022) for socioeconomic status[5]; Katz Index to assess functional abilities and identify any difficulties in performing activities of daily living[6]; To estimate the prevalence of loneliness a validated Hindi translation of the UCLA-3 item scale (Cut-off scores: 3–5 = not lonely; 6–9 = lonely) was used which include item no. 4, 10 and 16 from the UCLA-20 item scale; a shorter questionnaire comprising three items has been evaluated for its effectiveness with older adults.[13,14]

Statistical analysis: Data analysis was performed using IBM SPSS Statistics for Windows, Version 25.0 (IBM Corp., Armonk, NY, USA). Only complete cases were included in the analysis; observations with missing data were excluded (complete-case analysis). Descriptive statistics were used to present the baseline characteristics of the study participants. The Chi-square test assessed the association between loneliness (outcome variable) and various characteristics of the study participants. In the bivariate analysis, crude odds ratios (ORs) with 95% confidence intervals (CIs) were calculated. Variables with a P value below 0.20 were subsequently included in a multivariable binary logistic regression model to adjust for possible confounding factors and determine independent predictors of loneliness. Multicollinearity among independent variables included in the final logistic regression model was assessed using variance inflation factor (VIF). All variables had low VIF values (range: 1.07–1.33), indicating absence of significant multicollinearity. The enter method was applied for model fitting. Adjusted odds ratios (AORs) along with their 95% CIs were reported, and a P value less than 0.05 was considered statistically significant.

RESULTS

Most participants (74.6%) were aged 60–69 years, with a mean age of 66.28 ± 7.12 years (range: 60–100 years). Females constituted 53.3% of the study population. The majority identified as Hindu (88.2%), followed by Muslims (6.2%), Christians (3.2%), and Sikhs (2.4%). Most participants were married (79.3%), while 19.3% were widowed or divorced and 0.9% were unmarried. Nearly two-thirds (64.2%) belonged to joint families [Table 1].

Table 1.

Sociodemographic characteristics of the study population (n=338)

Characteristic		Male (158)		Female (180)		Total (n=338)
		n (%)		n (%)		n (%)
Age (in completed years)
60–69		109 (69.0)		143 (79.4)		252 (74.6)
70–79		40 (25.3)		26 (14.4)		66 (19.5)
80 and above		9 (5.7)		11 (6.2)		20 (5.9)
Religion
Hindu		136 (86.1)		162 (90.0)		298 (88.2)
Muslim		13 (8.2)		8 (4.4)		21 (6.2)
Sikh		4 (2.5)		4 (2.2)		8 (2.4)
Christian		5 (3.2)		6 (3.3)		11 (3.2)
Marital Status
Married		148 (93.7)		120 (66.7)		268 (79.3)
Unmarried		2 (1.3)		1 (0.6)		3 (0.9)
Widow/Divorced		8 (5.0)		59 (32.7)		67 (19.8)
Type of family
Nuclear family		64 (40.5)		57 (31.7)		121 (35.8)
Joint family		94 (59.5)		123 (68.3)		217 (64.2)

Approximately one-third of participants were illiterate (32.2%). A large proportion (79.3%) had a monthly family income below ₹46,089. About 62.4% were not currently employed. In terms of socioeconomic status, 58.3% were classified as upper lower class, followed by lower (23.7%), lower middle (13.6%), and upper middle (4.4%). Notably, there were no participants classified in the upper socioeconomic status [Table 2].

Table 2.

Socioeconomic characteristics of the study population (modified Kuppuswamy scale 2022) (n=338)

Characteristic		Male (158)		Female (180)		Total (n=338)
		n (%)		n (%)		n (%)
Literacy Status
Profession/Honors		2 (1.3)		0 (0.0)		2 (0.6)
Post-graduate/graduate		29 (18.3)		11 (6.1)		40 (11.8)
Intermediate/diploma		18 (11.3)		13 (7.2)		31 (9.2)
High school		27 (17.1)		15 (8.3)		42 (12.5)
Middle school		24 (15.2)		22 (12.2)		46 (13.6)
Primary school		35 (22.2)		33 (18.4)		68 (20.1)
Illiterate		23 (14.6)		86 (47.8)		109 (32.2)
Occupation
Unemployed/retired/homemaker		114 (72.2)		162 (90.0)		276 (81.7)
Elementary occupation		10 (6.3)		8 (4.4)		18 (5.3)
Plant and machine operator		5 (3.2)		1 (0.5)		6 (1.8)
Craft and trade worker		2 (1.3)		0 (0.0)		2 (0.6)
Skilled agriculture worker		1 (0.6)		3 (1.7)		4 (1.2)
Shop and market sales worker		24 (15.2)		5 (2.8)		29 (8.6)
Clerk		1 (0.6)		0 (0.0)		1 (0.3)
Technician		0 (0.0)		0 (0.0)		0 (0.0)
Professionals		0 (0.0)		1 (0.5)		1 (0.3)
Legislator/senior official/manager		1 (0.6)		0 (0.0)		1 (0.3)
Socioeconomic status
II (Upper middle class)		11 (7.0)		4 (2.2)		15 (4.4)
III (Lower middle class)		35 (22.2)		11 (6.1)		46 (13.6)
IV (Upper lower class)		93 (58.8)		104 (57.8)		197 (58.3)
V (Lower class)		19 (12.0)		61 (33.9)		80 (23.7)

Out of the 338 individuals assessed using the UCLA-3 scale, 52.4% were identified as experiencing loneliness, while 47.6% were not. This shows that more than half of the participants faced loneliness to some extent [Figure 1].

Figure 1.

Prevalence of loneliness by key predictors

The prevalence of loneliness was significantly higher among participants aged 80 years and above compared to those aged 60–79 years (OR = 3.90, P = 0.011). Females reported more loneliness than males (OR = 3.94, P = 0.001). Marital status showed a strong association, with unmarried, widowed, or divorced individuals being more likely to experience loneliness compared to married participants (OR = 10.31, P = 0.001). Similarly, those living in nuclear families (OR = 2.28, P = 0.001), having literacy status below primary level (OR = 2.58, P = 0.001), and belonging to below lower middle socioeconomic strata (OR = 2.07, P = 0.011) were more prone to loneliness. Presence of chronic illness (OR = 3.31, P = 0.001) and impaired activities of daily living (OR = 6.76, P = 0.001) were also significantly associated with higher loneliness. No statistically significant association was observed with religion [Table 3].

Table 3.

Association of loneliness with various characteristics of study population (n=338)

Characteristics				Loneliness n (%)		No Loneliness n (%)		OR (95% CI)		P
Age		60–79*		161 (50.6)		157 (49.4)		3.901 (1.276–11.925)		0.011
		80 and above		16 (80.0)		4 (20.0)
Gender		Male*		55 (34.8)		103 (65.2)		3.939 (2.505–6.194)		0.001
		Female		122 (67.8)		58 (32.2)
Religion		Hindu*		152 (51.0)		146 (49.0)		1.601 (0.812–3.157)		0.172
		Others (Muslim/Christian/Sikh)		25 (62.5)		15 (37.5)
Marital status		Married*		115 (42.9)		153 (57.1)		10.311 (4.750–22.381)		0.001
		Others (unmarred/widow/divorced)		62 (88.6)		8 (11.4)
Family type		Joint*		98 (45.2)		119 (54.8)		2.284 (1.442–3.618)		0.001
		Nuclear		79 (65.3)		42 (34.7)
Literacy status		Above primary*		103 (45.0)		126 (55.0)		2.583 (1.602–4.184)		0.001
		Below primary		74 (67.9)		35 (32.1)
Socioeconomic status		Lower middle and above*		23 (37.7)		38 (62.3)		2.069 (1.170–3.656)		0.011
		Below lower middle		154 (55.6)		123 (44.4)
Chronic illness		Absent*		64 (37.9)		105 (62.1)		3.311 (2.118–5.181)		0.001
		Present		113 (66.9)		56 (33.1)
ADL		Normal*		151 (49.0)		157 (51.0)		6.758 (2.304–19.824)		0.001
		Impaired		26 (86.7)		4 (13.3)

OR: Odds Ratio; CI: Confidence Interval; Reference category indicated by *

Multivariable logistic regression showed that females (AOR = 4.58, P = 0.001), those who were unmarried, widowed, or divorced (AOR = 7.29, P = 0.001), and individuals from nuclear families (AOR = 3.71, P = 0.001) had a significantly greater likelihood of experiencing loneliness. The odds were also higher among participants with chronic illnesses (AOR = 3.25, P = 0.001) and those with impaired activities of daily living (AOR = 6.12, P = 0.005). In contrast, age, religion, literacy status and socioeconomic status did not show any significant association [Table 4]. The goodness-of-fit of the final model was assessed using the Hosmer–Lemeshow test, which indicated an adequate fit (χ² =13.85, df = 8, P = 0.086).

Table 4.

Logistic regression analysis: Factors related to Loneliness among elderly (n=338)

Characteristics				AOR (95% CI)		P
Age		60–79*		0.927 (0.223–3.859)		0.917
		80 and above
Gender		Male*		4.580 (2.469–8.496)		0.001
		Female
Religion		Hindu*		1.942 (0.820–4.599)		0.132
		Others (Muslim/Christian/Sikh)
Marital status		Married*		7.289 (3.150–16.870)		0.001
		Others (unmarried/widow/divorce)
Family type		Joint*		3.708 (2.066–6.652)		0.001
		Nuclear
Literacy status		Above primary*		1.301 (0.684–2.475)		0.423
		Below primary
Socioeconomic status		Lower middle and above*		0.935 (0.446–1.960)		0.858
		Below lower middle
Chronic illness		Absent*		3.254 (1.861–5.692)		0.001
		Present
ADL		Normal*		6.117 (1.721–21.743)		0.005
		Impaired

AOR: Adjusted Odds Ratio; CI: Confidence Interval; Reference category indicated by *

DISCUSSION

In our study, 52.4% of older adults reported experiencing loneliness, as measured by the 3-item UCLA scale.

Similarly, a 2018 study conducted in the field practice area of the Department of Community Medicine, School of Public Health, PGIMER, Chandigarh, found that 55% of respondents experienced loneliness. Specifically, 40.2% of participants felt a lack of companionship, 37.2% felt left out of life, and 34.8% felt socially isolated. Additionally, a study by Grover S et al.,[10] conducted in a rural noncommunicable disease clinic reported that older adults with noncommunicable diseases were found to be at a greater risk of experiencing loneliness. However, as the study setting differs from the present urban field practice area, direct comparisons should be interpreted with caution.

Poor health is frequently linked to loneliness, as it can contribute to social and emotional disengagement. Older adults managing multiple comorbidities may experience increased isolation due to less frequent interaction and communication with family and friends.

A study conducted in Bengaluru, Karnataka, India, reported that 37.6% of participants experienced loneliness, indicated by scores of 50 or above. When analyzed by gender, 29.6% of men and 42.7% of women were found to be lonely. The prevalence of loneliness increased with age, with the highest rates seen in individuals aged 80 years and above. Higher levels of loneliness were also observed among those living in smaller families (85.7%) and among individuals who had lost or were without a spouse (46.7%). Furthermore, participants with physical limitations affecting their ability to carry out daily or instrumental activities had higher loneliness rates (61% and 39%, respectively).[11]

Another study conducted in Bangladesh found a statistically significant association between loneliness and being widowed, as well as being female.[12]

A study from Indonesia reported that 64% of elderly individuals experienced feelings of loneliness. Multivariate logistic regression analysis identified several factors associated with a higher likelihood of loneliness, including being female, having fewer children, poor self-rated health, the presence of multiple chronic conditions, cognitive impairment and symptoms of depression. These findings underscore the growing concern of loneliness as a public health issue among the aging population in Indonesia, emphasizing the need for timely intervention by government bodies, healthcare professionals, and social service organizations.[15]

A study conducted in an urban area of Portugal reported that 36.7% (95% CI: 35.1–38.3%) of individuals attending the health center experienced loneliness. Factors such as being aged 80 years or older, living alone, having less than nine years of formal education, dissatisfaction with income, and family dysfunction were linked to higher levels of perceived loneliness. Conversely, being married or in a stable partnership, as well as continuing professional employment, were associated with lower feelings of loneliness.[16]

A study from Uganda revealed that older adults residing in urban areas were more likely to experience loneliness than those living in rural communities. Loneliness was also more prevalent among widowed elders compared to those who were married. Contributing factors included the absence of a television, lack of pension support, and problems related to joint health. The study highlighted the need for targeted interventions such as developing age-friendly urban environments, encouraging participation in social groups for older persons, expanding decentralized healthcare services for the elderly, and establishing a dedicated financial support system for senior citizens.[17]

Research carried out in Yazd found that 71% of respondents reported no loneliness, while 24% experienced it at a moderate level and 5% at a severe level. Statistical analysis showed significant associations (P < 0.05) between loneliness and various factors such as education level, marital status, number of children, past occupation, present employment, living in a nursing home, insufficient income, sleep quantity and quality, and self-rated health status.[18]

Primary healthcare centers play an important role in enhancing the well-being and confidence of older adults by addressing their fears and worries. Hence, it is crucial to recognize and evaluate loneliness and social isolation whenever mental health concerns are identified or reported among the elderly.

The National Mental Health Survey of India (2015–2016) highlighted the burden of various mental health conditions but placed limited focus on loneliness.[19] It is essential to examine loneliness in relation to other mental disorders, particularly among high-risk groups with multiple coexisting conditions. Further investigations are needed to determine its prevalence across diverse populations and subgroups, understand the lived experiences of loneliness, explore associated psychiatric comorbidities, and identify both risk and protective factors influencing it.

In this study, although age, literacy status and socioeconomic status were significantly associated with loneliness in the bivariate analysis, these variables lost statistical significance after adjustment in the multivariable logistic regression model. This attenuation suggests that their influence was largely mediated or confounded by more proximal determinants such as marital status, family structure, presence of chronic illness, and functional impairment, which remained significant independent predictors. Similar findings have been reported in studies from Chandigarh[10] and Indonesia,[15] where age and socioeconomic factors did not retain statistical significance after controlling for psychosocial and health-related variables. These results highlight that social connectedness, physical health, and functional ability have a more direct impact on loneliness among the elderly than distal sociodemographic characteristics. Therefore, interventions should focus on enhancing family and community support systems and promoting functional independence rather than targeting demographic or economic attributes alone.

Future research and interventions should draw on experiences from diverse settings to tackle the growing issue of loneliness in India. Since loneliness often coexists with noncommunicable diseases (NCDs), particularly in older adults, addressing it becomes especially important in a country where NCD prevalence is already high. Recognizing loneliness as a contributing factor to many conditions commonly associated with aging can help reduce both the public health burden and existing health inequalities. Building awareness among healthcare providers, researchers, and policymakers through training and capacity-building initiatives is vital for early identification and management. Furthermore, strong collaboration between researchers, community health workers, institutions, and other stakeholders is necessary to create effective strategies for both research and practice aimed at mitigating loneliness in India.

Strengths and limitations

This study has several notable strengths. As a community-based investigation conducted in an urban field practice area, it offers practical insights into the burden of loneliness among older adults. The application of standardized and validated instruments, such as the 3-item UCLA Loneliness Scale and the Katz Index for functional assessment, adds to the robustness of the findings. Furthermore, the use of multivariable analysis enabled the identification of independent predictors while accounting for potential confounding variables.

However, certain limitations should be considered while interpreting the results Cross-sectional design limiting causal inference; Possibility of recall and reporting bias; The present study was conducted as part of a larger study assessing depression among the elderly. However, the current manuscript specifically focuses on loneliness as an independent entity of public health importance

The relationship between depression and loneliness was not explored in this manuscript and warrants further analysis.

Although cluster sampling was not formally employed, more than one elderly individual could be included from the same household, which may have introduced intra-household correlation. A design effect of 1.2 was applied as a precautionary measure during sample size estimation; however, in the absence of prior data, the intracluster correlation coefficient was not estimated.

The absence of a standardized cognitive screening tool may have led to potential selection bias.

The present study used the 3-item UCLA Loneliness Scale, a brief and validated screening tool suitable for community-based surveys. However, comparisons with studies employing the full 20-item UCLA scale should be interpreted with caution, as differences in scale length, scoring, and sensitivity may limit direct comparability of prevalence estimates.

Data collection for this study commenced in January 2023, a period following the COVID-19 pandemic, which may have had an impact on the levels of loneliness observed. Factors such as extended periods of social distancing, decreased interpersonal interactions, bereavement, and interruptions in routine healthcare and community activities during the pandemic could have heightened feelings of isolation among older adults. Although the present study did not specifically evaluate the effects of the pandemic, its lingering influence should be taken into account when interpreting the findings.

CONCLUSION

This study reveals a substantial burden of loneliness among older adults in an urban setting. Multivariable analysis showed that female gender, unmarried/widowed/divorced status, nuclear family structure, presence of chronic illness, and impaired activities of daily living were independently associated with loneliness. These findings underscore the importance of both social circumstances and health status in influencing loneliness among the elderly. Interventions should prioritize vulnerable groups, enhance family and community support, and promote better management of chronic conditions and functional limitations. Further research with larger, more representative samples across diverse settings, preferably using longitudinal designs, is needed to strengthen the evidence base. A comprehensive understanding of the multiple determinants of loneliness will be essential for addressing its increasing public health impact.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.