Atencion Primaria 53 (2021) 102020 Atención Primaria www.elsevier.es/ap ORIGINAL ARTICLE Improving mental health and wellbeing in elderly people isolated at home due to architectural barriers: A community health intervention Ferran Dabana,d,e,∗ , Irene Garcia-Subiratsa,d , Victoria Porthéa,c , Ma José Lópeza,c,d , Begoña de-Eytob , M Isabel Pasarína,c,d,e , Carme Borrella,c,d,e , Lucía Artazcoza,c,d,e , Anna Péreza,d , Elia Díeza,c,d,e a Agència de Salut Pública de Barcelona, Barcelona, Spain b Creu Roja Barcelona, Barcelona, Spain c Centro de Investigación Biomédica en Red Epidemiología y Salud Pública (CIBERESP), Madrid, Spain d Institut d’Investigació Biomèdica de Sant Pau, Barcelona, Spain e Department of Experimental and Health Sciences, Universitat Pompeu Fabra, Barcelona, Spain Received 22 April 2020; accepted 13 November 2020 Available online 26 March 2021 KEYWORDS Abstract Elderly; Objectives: To explore the health effects of a community health intervention on older peo- Loneliness; ple who are isolated at home due to mobility problems or architectural barriers, to identify Social isolation; associated characteristics and to assess participants’ satisfaction. Health outcomes; Design: Quasi-experimental before---after study. Health inequalities; Setting: Five low-income neighbourhoods of Barcelona during 2010---15. Community health Participants: 147 participants, aged ≥59, living in isolation due to mobility problems or archi- intervention tectural barriers were interviewed before the intervention and after 6 months. Intervention: Primary Health Care teams, public health and social workers, and other com- munity agents carried out a community health intervention, consisting of weekly outings, facilitated by volunteers. Measurements: We assessed self-rated health, mental health using the General Health Ques- tionnaire (GHQ-12), and quality of life through the EuroQol scale. Satisfaction with the programme was evaluated using a set of questions. We analysed pre and post data with McNemar tests and fitted lineal and Poisson regression models. ∗ Corresponding author. E-mail address: fdaban@aspb.cat (F. Daban). https://doi.org/10.1016/j.aprim.2021.102020 0212-6567/© 2021 The Authors. Published by Elsevier España, S.L.U. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/). F. Daban, I. Garcia-Subirats, V. Porthé et al. Results: At 6 months, participants showed improvements in self-rated health and mental health and a reduction of anxiety. Improvements were greater among women, those who had not left home for ≥4 months, those with lower educational level, and those who had made ≥9 outings. Self-rated health [aRR: 1.29(1.04---1.62)] and mental health improvements [ˇ: 2.92(1.64---4.2)] remained significant in the multivariate models. Mean satisfaction was 9.3 out of 10. Conclusion: This community health intervention appears to improve several health outcomes in isolated elderly people, especially among the most vulnerable groups. Replications of this type of intervention could work in similar contexts. © 2021 The Authors. Published by Elsevier España, S.L.U. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). PALABRAS CLAVE Mejorando la salud mental y el bienestar de las personas mayores aisladas en casa Personas mayores; debido a las barreras arquitectónicas: una intervención de salud comunitaria Soledad; Resumen Aislamiento social; Objetivos: Explorar los efectos sobre la salud de una intervención de salud comunitaria en per- Resultados en salud; sonas mayores aisladas en casa debido a problemas de movilidad o a barreras arquitectónicas, Desigualdades en identificar las características asociadas y evaluar la satisfacción de las personas participantes. salud; Diseño: Estudio cuasi-experimental antes-después. Intervención en salud Emplazamiento: Cinco barrios de baja renta de Barcelona durante 2010-15. comunitaria Participantes: Se entrevistó a 147 participantes, ≥59 años, antes y 6 meses después de la intervención. Intervención: Equipos de atención primaria, trabajadores sociales, de salud pública y otros agentes comunitarios desarrollaron una intervención que consistía en salidas semanales, facil- itadas por voluntarios. Mediciones: Se evalúo la salud autopercibida, la salud mental utilizando la escala GHQ-12 y la calidad de vida mediante la escala EuroQol. La satisfacción se evaluó mediante un conjunto de preguntas. Analizamos los datos previos y posteriores con pruebas de McNemar y modelos de regresión lineal y de Poisson ajustados. Resultados: A los 6 meses, los participantes mostraron mejoras en la salud percibida, en la salud mental y en la reducción de la ansiedad. Las mejoras fueron mayores entre las mujeres, las personas que no habían salido de casa durante ≥4 meses, las de bajo nivel educativo y las que habían realizado ≥9 salidas. La salud percibida (aRR: 1,29 [1,04-1,62]) y las mejoras en salud mental [(␤: 2,92 [1,64-4,2]) permanecieron significativas en los modelos multivariados. La satisfacción media fue de 9,3 sobre 10. Conclusión: Esta intervención de salud comunitaria parece mejorar varios resultados de salud en las personas mayores aisladas, especialmente en los grupos más vulnerables. Replicar este tipo de intervención podría funcionar en contextos similares. © 2021 Los Autores. Publicado por Elsevier España, S.L.U. Este es un artı́culo Open Access bajo la licencia CC BY-NC-ND (http://creativecommons.org/licenses/by-nc-nd/4.0/). Introduction roles (children growing up and/or retirement) and increased prevalence of chronic or disabling illnesses.6 Social isolation among the elderly is associated with In elderly, difficulty in engaging in community activities increased mortality, poorer self-rated health, lower qual- has been associated with self-perceived depression and loss ity of life and greater risk of dementia.1,2 Social isolation of interest in pleasurable activities.7 Conversely, elderly refers to the lack of structural and functional social sup- people who have more social relationships and social partici- port that includes (a) the objective assessment of its size pation have better mental, social and physical health.3 Some and frequency, and (b) the subjective assessment of the studies have shown that interventions to promote social par- perceived value of emotional, instrumental and informa- ticipation for older people can prevent loneliness and social tional support provided by others.3 Loneliness is closely isolation associated among the least wealthy groups.5,8---10 related to social isolation and has been defined as the sub- Besides that, architectural barriers that hinder or prevent jective experience of the absence or loss of companionship people from going out on their own, have been described as which is always involuntary.3---5 Loneliness and social isola- factors associated with the perception of loneliness.4 tion are more prevalent among older people due to factors In Barcelona in 2016, 31.6% of buildings did not have such as the death of relatives and friends, loss of social an elevator, and this percentage was higher in the most 2 Atencion Primaria 53 (2021) 102020 Table 1 Description of the activities carried out by the working group to develop a community health intervention from 2010 to 2015. Phases/activities Working group Community Primary Social Public Red Crossb Neighbourhood City Develop- Health Care Services Healtha organisations Councilc ment Plan Design Health needs × × × × × × × assessment Intervention × × × × × × × design Participants × × × × detection Inclu- × × × × sion/exclusion of participants Implementation Outings × × × coordination Recruitment and × × maintenance of volunteers Outings × × × × equipment needed Evaluation Monitoring and × × × evaluation Results × × x dissemination a Barcelona Public Health Agency. b Red Cross in Barcelona. c Health Department of Barcelona City Council. deprived neighbourhoods.11 In addition, in 2018, 21.6% of Methods the citizens of Barcelona were over 65 years old, and 25.6% of them were living alone.12 Since 2007, the municipal Design strategy Barcelona Health in the Neighbourhoods has been working to reduce health inequalities through community We used a quasi-experimental before---after design. The health interventions.13,14 As part of this community health study analyses the intervention carried out from May 2010 strategy, a working group composed by Community Develop- to December 2015 in five disadvantaged neighbourhoods in ment Plan, Primary Health Care, Social Services, Barcelona Barcelona. Neighbourhoods were prioritised according to Public Health Agency, Health Department of Barcelona City the Family Disposable Income Index in 201015 and prevalence Council, Red Cross in Barcelona and neighbourhoods’ organ- of older people living alone.11 isations (Table 1), designed and conducted an intervention consisted of weekly outings in five deprived neighbour- hoods in Barcelona, to alleviate loneliness among older Participants people who were living in isolation in their homes for long periods, mainly due to mobility limitations and/or The inclusion criteria were people aged ≥59 years old and the lack of an elevator in their buildings. The objec- who had been living in isolation in their homes for two tive of this paper is to: (1) explore the effects of the or more months due to mobility limitations and/or lack of intervention on health outcomes, (2) identify what char- an elevator in their buildings. Bedridden persons, persons acteristics of the participants and the intervention are with severe dementia and those without medical authori- associated with health effects and (3) assess participants’ sation to leave their home were excluded. Primary health satisfaction. 3 F. Daban, I. Garcia-Subirats, V. Porthé et al. care teams contributed to establish and assessed health- leaving home, and educational level and (b) intervention related exclusion criteria. Services usually in contact with dose estimated through the number of outings. The depen- the target population in the neighbourhoods, such as Pri- dent variables were self-rated health, mental health, and mary Health Care, Social Services, Community Development quality of life. Self-rated health was assessed through the Plan and neighbourhoods’ organisations, were responsible question ‘How do you assess your health in general?’ with for detecting people with inclusion criteria (Table 1). a five categories Likert scale, which was dichotomised into good (fair, good and very good) or poor (poor or very poor).19 Intervention Mental health was assessed using the General Health Ques- tionnaire (GHQ-12), which is made up of 12 items that are The intervention was developed by a working group who answered using a 4-point Likert (0---4) scale. We analysed made the health need assessment and intervention design the GHQ-12 as a dichotomous variable where people with (Table 1). This consisted of weekly outings, facilitated by three or more positive answers were considered to be at volunteers using a portable climbing wheelchair. The coor- risk of poor mental health,20 and also as a continuous vari- dination and recruitment of volunteers was mainly in charge able, that is the sum of all the items (ranging from 0 to 36), of the Community Development Plan and Red Cross. Equip- where high values denote high mental disorder.21 The Lik- ment needed to carry out the outings such as wheelchairs ert scoring of the GHQ yielded a normal distribution and we and a portable climbing wheelchair was mainly provided by have used this variable as a continuous variable in our anal- primary health care, Red Cross and the Barcelona Public ysis. Quality of life related to health was evaluated using Health Agency (Table 1). The intervention promoted social the EuroQol scale (EQ-5D-3L),22 which assesses five dimen- support and participation through activities proposed by the sions including mobility, personal care, daily activities, pain participants, such as visiting friends, short walks, going to or discomfort, and anxiety/depression. EQ-5D-3L has three the market or attending church.16 In addition to the indi- levels of severity in each item, and it was dichotomised vidual weekly outings, group outings were conducted once into: (a) ‘no problems’ or ‘some problems’ and (b) ‘severe a month to foster new social relationships among partici- problems’.23 User’s satisfaction with the programme was pants, such as city sightseeing, cultural activities and group evaluated through a set of questions used in Barcelona dynamics. The evaluation process was mainly in charge of Health in the Neighbourhoods interventions to assess gen- the Barcelona Public Health Agency, Community Develop- eral satisfaction, dimensions of frequency, duration, time, ment Plan and Red Cross (Table 1). Economic and human place and punctuality and whether the participants would resources to maintain the sustainability of the interven- recommend it to other people.24,25 tion came from Barcelona Public Health Agency and Health Department of Barcelona City Council. Data analysis Sample and data collection We performed a univariate analysis to describe individual characteristics and intervention dose and a bivariate anal- Trained interviewers administered a face-to-face question- ysis to assess changes in self-rated health, mental health naire to participants before entering the programme and and the five quality of life dimensions, within the explana- after six months, during which time at least four outings tory variables, before and after the intervention through were performed. From a total of 147 initial participants, an McNemar tests for paired data. Then, we fitted generalised 8% were lost to follow up; thus 135 participants answered estimating equation regression models for the outcome both questionnaires. Reasons for lost to follow-up were variables adjusting for explanatory variables. Poisson regres- change of address, hospitalisation or death. sion models were fitted for self-rated health and quality of Since it was a community programme which the main aim life, and mental health was analysed through its continu- was improving health in disadvantaged populations, all peo- ous variable with a linear model. We analysed each quality ple who met the criteria and who wished to participate were of life dimension separately, as we did not expect that the included, and no sample size was calculated in the design programme would affect all of them. of the intervention. However, to ensure the validity of the In our study there were two types of missing data: losses results, we estimated the sample size necessary to detect a to follow up and items missing within the survey. To deal with change in health perception of 20% (based on the interven- the first type we compared the socio-demographic charac- tion pilot study)17 : accepting an alpha risk of 0.05 and a beta teristics of lost cases with the total number of participants risk of less than 0.2 in a two-sided contrast, we obtained that (t-test for continuous variables and chi2 square test for the minimum sample size required would be 100 people.18 categorical variables). For the missing data in the specific items in the GHQ-12 scale, we applied the simple imputation method based on the person-mean score when less than two Study variables items were missing per case. This method is considered to be adequate due to the high proportion of complete cases.26 The explanatory variables were divided into (a) individ- All analyses were conducted using the statistical package ual characteristics: sex, age, time (in months) without STATA. 4 Atencion Primaria 53 (2021) 102020 Initial participants (with pre-post data) (n=147) Reasons for the lost to follow-up: Participants’ lost to • Change address: n=3 (27%) follow-up 8% (n=12) • Hospitalization: n=5 (45%) • Death: n=1 (9%) • Others: n=3 (18%) Final sample (n=135) Results The study sample was consisted mainly of women (58.5%), Table 2 Socio-demographic characteristics of the partici- people aged ≥85 (44.8%) and had not left their homes for pants from 2010 to 2015. ≥4 months (55.9%) (Table 2). No differences were observed Variables % (n) between follow-up losses and study participants by sex, age and time without leaving home; however, the level of study Sex in the former group was lower than the overall sample. Par- Women 58.5 (86) ticipants made a median of 8 and a mean of 18 outings Men 41.5 (61) during the six months period of study. After the interven- Age tion, participants’ self-rated health improved significantly in 59---74 14.0 (20) all categories of the explanatory variables, except for men 75---84 41.2 (59) or for people aged ≥85 years (Table 3). Among participants ≥85 44.8 (64) who made 9---24 outings, good self-rated health increased from 51.6% before the intervention to 77.4% after; among Time without leaving home participants who made ≥25 outings, it increased from 35.7% <4 months 44.1 (60) to 67.9%. Self-perceived mental health improved among all ≥4 months 55.9 (76) participants, especially among men (45.3% to 66.0%), those Level of studies achieved younger than 85 years old (50.8% to 71.6%), those who had Less than primary 21.6 (27) not left their homes for ≥4 months (44.8% to 65.7%), those Primary 55.2 (69) who had made 9---24 outings (46.7% to 73.3%), and those with Secondary 19.2 (24) a lower level of education (55.6% to 74.1%). The intervention University 4.0 (5) had a positive effect on anxiety, with significant increases of people reporting no anxiety among women (69.7% to 86.8%), Number of outings during the intervention* people aged ≥85 (73.2% to 89.3%), those who had not left 4---8 53.4 (71) their homes for ≥4 months (69.1% to 89.7%), those who had 9---24 24.1 (32) made 9---24 outings (70.0% to 90.0%), and those with a lower ≥25 22.0 (29) level of education (71.4% to 85.7%). We did not observe sig- Mean of outings (sd) 18.0 (21.3) nificant changes in other aspects of quality of life (mobility, Median (interquartile range) 8 (7---18) personal care, daily activities and pain or discomfort). * There were a total of 135 participants in the intervention. After adjusting for explanatory variables, the multivari- ate analysis showed that the improvement, before and after 5 F. Daban, I. Garcia-Subirats, V. Porthé et al. Table 3 Self-rated health, mental health and anxiety prevalence among isolated elder participants before and after the programme (n = 135). Before After Difference Bivariate analysis Multivariate analysis % (n) % (n) % p value Coefficient (95% CI) Good self-rated health 1.29 (1.04---1.62)a Sex *** Women 48.7 (38) 69.2 (54) 20.5 Men 60.0 (33) 71.0 (39) 11 Age *** <85 45.1 (32) 66.2 (47) 21.1 ≥85 63.8 (37) 74.1 (43) 10.3 Time without leaving home *** <4 months 53.7 (29) 75.9 (41) 22.2 ≥4 months 55.1 (38) 68.1 (47) 13 * Educational level Without studies or primary 60.0 (51) 70.6 (60) 10.6 * Secondary or university 42.9 (12) 71.4 (20) 28.5 * Number of outings 4---8 59.2 (42) 66.2 (47) 7 9---24 51.6 (16) 77.4 (24) 25.8 * ≥25 35.7 (10) 67.9 (19) 31.1 ** *** Total 53.4 (71) 70.0 (93) 16.6 Good mental health 2.92 (1.64---4.20)b Sex Women 58.6 (41) 74.4 (50) Men 45.3 (24) 66.0 (35) * Age ** <85 50.8 (34) 71.6 (48) ≥85 56.6 (30) 67.9 (36) Time without leaving home <4 months 66.0 (33) 78.0 (39) ≥4 months 44.8 (30) 65.7 (44) ** Educational level *** Without studies or primary 55.6 (45) 74.1 (60) Secondary or university 56.5 (13) 73.9 (17) Number of outings 4---8 63.4 (45) 70.4 (50) 9---24 46.7 (14) 73.3 (22) * ≥25 23.8 (5) 57.1 (12) Total 52.85 (65) 69.11 (85) * No anxiety 1.17 (0.91---1.50)a Sex *** Women 69.7 (53) 86.8 (66) 17.1 Men 81.5 (44) 85.2 (46) 3.7 Age <85 77.5 (55) 83.1 (59) 5.6 ≥85 73.2 (41) 89.3 (50) 16.1 * Time without leaving home <4 months 82.7 (43) 84.6 (44) 1.9 6 Atencion Primaria 53 (2021) 102020 Table 3 (Continued) Before After Difference Bivariate analysis Multivariate analysis % (n) % (n) % p value Coefficient (95% CI) ≥4 months 69.1 (47) 89.7 (61) 20.6 *** Educational level *** Without studies or primary 71.4 (60) 85.7 (72) 14.3 Secondary or university 74.1 (20) 81.5 (22) 7.4 Number of outings 4---8 80.0 (56) 85.7 (60) 5.7 9---24 70.0 (21) 90.0 (27) 20 * ≥25 63.0 (17) 81.5 (22) 17.8 *** Total 74.6 (97) 86.2 (112) 11.6 a Adjusted risk ratio of the GEE regression model adjusting for all explanatory variables. b Adjusted coefficient of the GEE normal regression model adjusting for all explanatory variables. * p < 0.05. ** p < 0.01. *** p < 0.001. the intervention, in self-rated health (aRR 1.29; (CI) 95%: loneliness are those that use new technologies to increase 1.04---1.62) and mental health (ˇ: 2.92; (CI) 95%: 1.64---4.20) socialisation with loved ones or new friendships.28---31 Other remained significant, while no differences were found in types of individual intervention, such as home visits, have anxiety (aRR 1.17; (CI) 95%: 0.91---1.50). shown to be ineffective in reducing loneliness among older Participants reported being very satisfied with the pro- people,3,31 probably due to the relatively small increase in gramme, with an average score of 9.3 out of 10 and 99% their social interaction, and that they do not encourage par- declared they would recommend the programme to other ticipants to leave their homes. A number of group-based people. Dimensions of satisfaction, such as frequency, dura- interventions, such as educational or volunteering activities tion, time, place, and punctuality, were rated as adequate over several months, have proven to be effective in redu- by almost 100% of the participants, although the frequency cing loneliness among elderly people, by generating purpose of the group outings was considered insufficient (data not in their lives, promoting social participation and support, shown). and reducing stigma.3,8,9,31---34 There is also evidence describ- ing that community health interventions can contribute to improved health, wellbeing and social support in disadvan- Discussion taged populations.35---38 One systematic review found that 21 out of 24 studies analysed reported improvements in health- This study shows that this community health intervention related behaviours, a range of health outcomes, health improves participants’ self-rated health and mental health literacy and the use of health services, as well as community and reduces anxiety. We observed significant improvements changes such as empowerment or public health planning. in self-rated health in all sociodemographic groups, except In addition, 60% of the studies included reported reduced for men. The number of outings was also associated with health inequalities.35 good self-rated health. There was a general improvement Furthermore, evidence indicates that effective interven- in mental health and lower rates of anxiety, mainly among tions are not restricted to group interventions or solitary men, people who had not left their home for a long time, interventions.29,39,40 Our intervention combined individual those with a lower level of education, and those who and group outings, to promote social support and partic- had made more than 9 outings during the intervention. ipation, and this could be another key component of the The improvements in self-rated health and mental health intervention’s success. remained significant with the adjusted analysis. Our intervention was implemented at the time when the Evidence shows that people from vulnerable groups are Catalonian Department of Health launched COMSALUT pro- the least able to positively respond to population-approach gramme, with the objective of redirect the Primary Health interventions.27 In this sense, this community health inter- Care towards health promotion and community health, as vention, which forms part of a strategy to reduce social well as tackling social health inequalities. This context may health inequalities in disadvantaged neighbourhoods, seems have enabled synergies and exchanges between Primary to have contributed to alleviating inequalities by benefiting Health Care, Public Health and other community agents to people with more cumulative risk factors (those with a lower boost local community health.41 level of education and ≥4 months without leaving home). One of the main strengths of this intervention is the com- This intervention is one of the few, at the individual munity response to a global health problem among older level, that seems to be effective in reducing social isolation people, especially among those with fewer resources.4,9 This and loneliness. Previous studies have shown that the most community health intervention is based on the informa- effective individual interventions to reduce elder people tion that Primary Health Care, Social Services, and other 7 F. Daban, I. Garcia-Subirats, V. Porthé et al. community agents collect on a daily basis, to access a group of the population that can hardly benefit from usual community interventions. This intersectoral and commu- • This intervention based in periodic outings to pro- nitarian approach strengthens community health and the mote social relationships and social participation co-production of interventions to reduce social inequalities. could contribute to improve self-rated health and Another strength of this study is that it is one of the first com- alleviate mental health risks among the most vul- munity health interventions that evaluates health outcomes nerable elderly people isolated at home. measured with validated scales. • This community health intervention has enabled syn- The main limitation of this study is the absence of a ergies and exchanges between primary health care, comparison group. An experimental design was not feasi- public health, and other community agents to boost ble due to ethical constraints (leaving one of the study arms local community health. at home for 6 months) and the limited number of eligible participants. Another limitation could be reverse causation between the number of outings and self-reported health, Compliance with ethical standards i.e. participants who had good health may have been more likely to have more outings, while those with poorer health Ethics approval may have been unable to leave their homes for this reason. Even though, some elements may reinforce and give support Obtained (2019/8569/I) (undisclosed details to protect to our results: (a) a pilot test of the intervention yielded sim- anonymity). ilar results in self-rated health, mental health and anxiety reduction17 ; (b) a strong positive relationship between the number of outings and health outcomes; (c) the improve- Informed consent ments in variables expected to suffer a maturation bias, as older people’s self-rated and mental health tend to worsen A written informed consent for participation was signed by over time, and (d) the great satisfaction reported by the each participant before intervention. participants. We conclude that community health strategies based on Funding periodic outings to promote social relationships and social participation could contribute to improve self-rated health Undisclosed details to protect anonymity. and alleviate mental health risks among vulnerable elderly people isolated at home due to mobility limitations and Conflict of interests architectural barriers. Replications of this community health intervention could work in similar contexts. None declared. Acknowledgements Key points The authors are most grateful to people who participated in the study and generously shared their time. We thank What is already known on this topic? Michael Maher and Gavin Lucas for their help in correcting the English version of this article. This article forms part • Social isolation among the elderly is associated with of the doctoral dissertation of Ferran Daban Aguilar at the increased mortality, poorer self-rated health, lower Universitat Pompeu Fabra of Barcelona. quality of life and greater risk of dementia. • Evidence has shown that interventions to promote References social participation for older people can prevent loneliness and social isolation associated among the 1. Hand C, Retrum J, Ware G, Iwasaki P, Moaalii G, Main least wealthy groups. DS. Understanding social isolation among urban aging adults: • Architectural barriers can hinder or prevent people informing occupation-based approaches. OTJR Occup Particip from going out on their own and have been associ- Heal. 2017;37:188---98. ated with the perception of loneliness. In Barcelona, 2. Singh A, Misra N. Loneliness, depression and sociability in old 31.6% of buildings did not have an elevator, and this age. Ind Psychiatry J. 2009;18:51---5. percentage was higher in the most deprived neigh- 3. Dickens AP, Richards SH, Greaves CJ, Campbell JL. Interventions bourhoods. targeting social isolation in older people: a systematic review. BMC Public Health. 2011;11:647. 4. Gené-Badia J, Comice P, Belchín A, Erdozain MÁ, Cáliz L, Tor- What this study adds? res S, et al. Profiles of loneliness and social isolation in urban population. Aten Primaria. 2019;54:224---32. • This community health intervention is a response to 5. Grenade L, Boldy D. Social isolation and loneliness among older a global health problem among older people, espe- people: issues and future challenges in community and residen- cially among the most vulnerable groups. tial settings. Aust Heal Rev. 2008;32:468---78. 8 Atencion Primaria 53 (2021) 102020 6. Dahlberg L, Mckee KJ. Correlates of social and emotional loneli- Barcelona (POIBA project): study protocol of the intervention. ness in older people: evidence from an English community study. J Public health Res. 2018;7:1129. Aging Ment Heal. 2014;18:504---14. 25. Lapena C, Continente X, Sánchez Mascuñano A, Pons Vigués 7. Hao G, Bishwajit G, Tang S, Nie C, Ji L, Huang R. Social partic- M, Pujol Ribera E, López MJ. Qualitative evaluation of a ipation and perceived depression among elderly population in community-based intervention to reduce social isolation among South Africa. Clin Interv Aging. 2017;12:971---6. older people in disadvantaged urban areas of Barcelona. Heal 8. Pool MS, Agyemang CO, Smalbrugge M. Interventions to improve Soc Care Community. 2020;28:1488---503. social determinants of health among elderly ethnic minority 26. Bono C, Ried LD, Kimberlin C, Vogel B. Missing data on the Cen- groups: a review. Eur J Public Health. 2017;27:1048---54. ter for Epidemiologic Studies Depression Scale: a comparison of 9. Niedzwiedz CL, Richardson EA, Tunstall H, Shortt NK, Mitchell 4 imputation techniques. Res Soc Adm Pharm. 2007;3:1---27. RJ, Pearce JR. The relationship between wealth and loneliness 27. Frohlich KL, Potvin L. Transcending the known in public health among older people across Europe: is social participation pro- practice: the inequality paradox: the population approach and tective? Prev Med (Baltim). 2016;91:24---31. vulnerable populations. Am J Public Health. 2008;98:216---21. 10. Cattan M, White M, Bond J, Learmouth A. Preventing social iso- 28. Sharit J, Czaja SJ. The role of technology in supporting older lation and loneliness among older people: a systematic review workers. Gerontechnology. 2018;17:32. of health promotion interventions. Ageing Soc. 2005;25:41---67. 29. Poscia A, Stojanovic J, La Milia DI, Duplaga M, Grysztar M, 11. Agència Salut Pública de Barcelona. Infobarris Barcelona [Inter- Moscato U, et al. Interventions targeting loneliness and social net]. Available from: http://www.aspb.cat/infobarris [cited isolation among the older people: an update systematic review. 12.04.20]. Exp Gerontol. 2018;102:133---44. 12. Ajuntament de Barcelona. Departament d’Estadística i difusió 30. Tsai HH, Tsai YF. Changes in depressive symptoms, social sup- de dades [Internet]. Available from: https://www.bcn.cat/ port, and loneliness over 1 year after a minimum 3-month estadistica/catala/index.htm [cited 15.01.20]. videoconference program for older nursing home residents. J 13. Fuertes C, Pasarín MI, Borrell C, Artazcoz L, Díez È, Calzada Med Internet Res. 2011;13:e93. N, et al. Feasibility of a community action model oriented 31. Hagan R, Manktelow R, Taylor BJ, Mallett J. Reducing loneliness to reduce inequalities in health. Health Policy (New York). amongst older people: a systematic search and narrative review. 2012;107:289---95. Aging Ment Heal. 2014;18:683---93. 14. Díez E, Pasarín MI, Daban FCN, Fuertes C, Artazcoz LBC. 32. Robins LM, Hill KD, Finch CF, Clemson L, Haines T. «Salut als barris» en Barcelona, una intervención comunitaria The association between physical activity and social isola- para reducir las desigualdades sociales en salud. Comunidad. tion in community-dwelling older adults. Aging Ment Heal. 2012;14:121---6. 2018;22:175---82. 15. Ajuntament de Barcelona. Distribució territorial de la 33. Findlay RA. Interventions to reduce social isolation renda familiar a Barcelona 2000-2017. Departament amongst older people: where is the evidence? Ageing d’Estadística i difusió de dades [Internet]. Available from: Soc. 2003;23:647---58. http://www.bcn.cat/estadistica/catala/dades/economia/ 34. Coll-planas L, Valle G, Bonilla P, Masat T, Puig T, Montseserin renda/rdfamiliar/evo/rfbarris.htm [cited 10.04.20]. R. Promoting social capital to alleviate loneliness and improve 16. Agència Salut Pública de Barcelona. Barcelona Salut health among older people in Spain. Heal Soc Care Commun. als Barris [Internet]. Available from: http://www.aspb. 2017;25:145---57. cat/documents/informes//promocio-entorns/barris// [cited 35. Cyril S, Smith BJ, Possamai-Inesedy A, Renzaho AMN. Exploring 16.12.19]. the role of community engagement in improving the health of 17. Diez E, Daban F, Pasarin M, Artazcoz L, Fuertes C, Jose Lopez disadvantaged populations: a systematic review. Glob Health M, et al. Evaluation of a community program to reduce isola- Action. 2015;8:1---13. tion in older people due to architectural barriers. Gac Sanit. 36. O’Mara-Eves A, Brunton G, Oliver S, Kavanagh J, Jamal F, 2014;28:386---8. Thomas J. The effectiveness of community engagement in 18. Marrugat J, Vila J. Calculadora de tamaño mues- public health interventions for disadvantaged groups: a meta- tral. GRANMO [Internet]. 2012. Available from: analysis. BMC Public Health. 2015;15:1---23. https://www.imim.es/ofertadeserveis/software-public/ 37. NICE. Community engagement: improving health and wellbe- granmo/ ing and reducing health inequalities. NICE guideline NG44. 19. Idler EL, Benyamini Y. Self-rated health and mortality: a Your responsibility our responsibility [Internet]; 2016. Available review of twenty-seven community studies. J Health Soc Behav. from: https://www.nice.org.uk/guidance/ng44 1997;38:21---37. 38. Daban F, Pasarín MI, Borrell C, Artazcoz L, Pérez A, Fernández 20. Goldberg D. Manual of the general health questionnaire. Wind- A, et al. Barcelona Salut als Barris: twelve years’ experience sor: National Foundation for Educational Research; 1978. of tackling social health inequalities through community-based 21. Rodrigo MF, Molina JG, Losilla J-M, Vives J, Tomás JM. Method interventions. Gac Sanit. 2020. S0213-9111(20)30070-4. effects associated with negatively and positively worded items 39. Gardiner C, Geldenhuys G, Gott M. Interventions to reduce on the 12-item General Health Questionnaire (GHQ-12): results social isolation and loneliness among older people: an integra- from a cross-sectional survey with a representative sample of tive review. Heal Soc Care Community. 2018;26:147---57. Catalonian workers. BMJ Open. 2019;9:e031859. 40. Cotterell N, Buffel T, Phillipson C. Preventing social isolation in 22. Williams A, The EuroQol Group. EuroQol --- a new facility for the older people. Maturitas. 2018;113:80---4. measurement of health related quality of life. Health Policy 41. Fornells JM, González A, Lacasa C, Martin-Zurro A, Martínez E, (New York). 1990;16:199---208. Mena I, et al. Projecte COMSALUT. Comunitat i Salut: Atenció 23. van Reenen M, Janssen B, Stolk E, Secnik Boye K, Herdman Primària i Comunitària [Internet]; 2015. Available from: M, Kennedy-Martin M, et al. EQ-5D-5L User Guide. EuroQol https://comsalut.files.wordpress.com/2015/12/comsalut- Research Foundation; 2019. document-base.pdf 24. Sánchez-Martínez F, Juárez O, Serral G, Valmayor S, Puigpinós R, Pasarín MI, et al. A childhood obesity prevention programme in 9