Gac Sanit. 2021;35(5):438–444 Original Energy poverty, its intensity and health in vulnerable populations in a Southern European city Juli Carrerea,b,c,d,∗ , Andrés Peraltac,e,f , Laura Oliverasa,b,c , María José Lópezc,g , Marc Marí-Dell’Olmoa,c,g , Joan Benache,f,h , Ana M. Novoaa,c a Institut d’Investigació Biomèdica (IIB Sant Pau), Barcelona, Spain b Department of Experimental and Health Sciences, Universitat Pompeu Fabra, Barcelona, Spain c Agència de Salut Pública de Barcelona, Barcelona, Spain d Asociación Bienestar y Desarrollo, Barcelona, Spain e Health Inequalities Research Group, Employment Conditions Knowledge Network (GREDS-EMCONET), Department of Political and Social Sciences, Universitat Pompeu Fabra, Barcelona, Spain f Johns Hopkins University, Pompeu Fabra University Public Policy Center, Barcelona, Spain g CIBER de Epidemiología y Salud Pública (CIBERESP), Spain h Transdisciplinary Research Group on Socioecological Transitions (GinTRANS2), Universidad Autónoma de Madrid, Madrid, Spain a r t i c l e i n f o a b s t r a c t Article history: Objective: To describe and compare the health status between vulnerable population participating in a Received 28 May 2020 program to tackle energy poverty (Energía, la justa) and the non-energy poor Barcelona (Spain) population Accepted 23 July 2020 and to analyse among participants the effect of energy poverty intensity on health. Available online 29 October 2020 Method: Cross-sectional study based on data from program participants obtained before the intervention. Participants (1799 women and 671 men) were compared to non-energy poor population of Barcelona Keywords: (1393 women and 1215 men) sampled from the Barcelona Public Health Survey (2016). Standardized Energy poverty prevalence and prevalence ratios of self-perceived poor health, asthma, chronic bronchitis, and depres- Fuel poverty Social determinants of health sion and/or anxiety were estimated, and among participants the association between health status and Social health inequalities energy poverty intensity was estimated with multivariate models. Health Results: The probability of poor self-perceived physical and mental health ranged from 2.2 to 5.3 times Vulnerable populations greater in the program participants than non-energy poor Barcelona population. Among program partic- ipants, those most affected were found to be the most likely to report poor health outcomes, regardless of other sociodemographic factors. Conclusions: The prevalence of poor physical and mental health was considerably higher among women and men participating in the programme compared to the non-energy poor population and was even worse among those who were more severely affected. Public policies that tackle energy poverty and its consequences are urgently needed in Spain. © 2020 SESPAS. 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/). Pobreza energética, su intensidad y salud en población vulnerable en una ciudad del sur de Europa r e s u m e n Palabras clave: Objetivo: Describir y comparar el estado de salud entre la población vulnerable que participa en un Pobreza energética programa de lucha contra la pobreza energética (Energía, la justa) y la población sin pobreza energética Determinantes sociales de la salud de Barcelona (España), y analizar entre las personas participantes el efecto de la intensidad de la pobreza Desigualdades sociales en salud energética en la salud. Salud Método: Estudio transversal basado en datos de los/las participantes en el programa obtenidos antes de Población vulnerable la intervención. Se compararon los/las participantes (1799 mujeres y 671 hombres) con población sin pobreza energética de Barcelona (1393 mujeres y 1215 hombres) seleccionada de la Encuesta de Salud Pública de Barcelona (2016). Se estimaron la prevalencia estandarizada y las razones de prevalencia de mala salud, asma, bronquitis crónica y depresión o ansiedad autopercibida. Entre los/las participantes se estimó la asociación entre el estado de salud y la intensidad de la pobreza energética con modelos multivariantes. Resultados: Los/las participantes presentaron entre 2,2 y 5,3 veces peores resultados en salud que la población sin pobreza energética. Entre los/las participantes se encontró que las personas más afectadas eran las que tenían más probabilidades de reportar peores resultados de salud, independientemente de otros factores sociodemográficos. ∗ Corresponding author. E-mail address: julicarrere@gmail.com (J. Carrere). https://doi.org/10.1016/j.gaceta.2020.07.007 0213-9111/© 2020 SESPAS. 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/). J. Carrere et al. / Gac Sanit. 2021;35(5):438–444 439 Conclusiones: La prevalencia de mala salud física y mental era considerablemente mayor en las mujeres y los hombres que participaron en el programa en comparación con la población sin pobreza energética, y era aún peor entre los/las más severamente afectados. En España se necesitan con urgencia políticas públicas que aborden la pobreza energética y sus consecuencias. © 2020 SESPAS. 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 cardiovascular diseases;18 poor mental health and well-being;19 and the exacerbation of existing health conditions.13,14 These Energy poverty is a significant social problem in the European adverse health effects are unequally distributed among differ- Union (EU) being the Eastern and Southern EU countries the most ent population groups. Children, the elderly, women, immigrants, affected.1 This problem has become more evident in the Southern and the lower-class population are more vulnerable to energy EU countries after the bursting of the housing market bubble in poverty, widening health inequalities.14,20 In contrast, the evi- 2008, the subsequent Great Recession and the concomitant auster- dence for the southern EU regions is scarce and mostly based on ity policies,2 as exemplified in Greece3 and Spain.4 Energy poverty European comparative studies.2,21–23 Hence, there is a lack of evi- is a dynamic, multidimensional phenomenon that can be defined dence on health inequalities between social groups. For example, as a household’s inability to ensure socially and materially required low-income women, such as single-parent families, are at a dis- levels of domestic energy services.5 There is widespread recogni- proportionate risk of poverty and are more susceptible to energy tion that the interaction of three drivers constitutes the critical poverty.24 Also, the responsibility for the care and household tasks determinant of energy poverty: the socioeconomic situation of the still falls mainly on women, often making them responsible for household, the energy performance of the dwelling, and the energy managing energy debts or requesting subsidies. Women’s health prices.6 These drivers are, in turn, influenced by more structural could be more adversely affected because gender patterns expose determinants such as energy, housing and labour policies and mar- them to spending more time at home with harmful indoor envi- kets, economic policies, and welfare states.7 ronments and the stress or feelings of helplessness associated with Southern European regions despite having a warmer climate the energy poverty managing.25 are among the most vulnerable to energy poverty. Two distinc- Energía, la justa (EJUSTA) was a pilot programme to tackle energy tive structural determinants account for this high vulnerability.2,7,8 poverty conducted in Barcelona in 2016. It was implemented by First, the social protection system within the welfare state regime, the city council and local non-governmental organisations. EJUSTA which is described as fragmented and poor developed, has revealed aimed to reduce inadequate and unaffordable household energy less ability to counteract unfair living conditions when primary and to improve health targeting vulnerable populations attended social networks, mainly the family, are unable to assist individuals.9 by the city social services. An opportunity was provided to evaluate Second, the lack of adequate heating systems and the poor quality this program, both its effectiveness and the experience perceived by of dwellings contribute to the higher prevalence in the southern the participants, and this paper presents the baseline study with the European region.1 Such a situation is attributable to the interac- following aims: to describe and compare the health status between tion between deficient housing policies and the market-oriented vulnerable population participating in EJUSTA program and the housing production that prioritises its exchange-value over its use- non-energy poor Barcelona population and to analyse among par- value. The latter negatively affects people’s right to affordable, ticipants the effect of energy poverty intensity on health. adequate, and quality housing. There are two widely recognised approaches to measuring Method energy poverty: household income/expenditure-based indicators and survey-based consensual indicators with a binary logic accord- Design, study population, and information sources ing to household self-assessment about the presence or absence of energy poverty drivers.10 The consensual approach can cap- A cross-sectional study of the vulnerable population affected by ture more comprehensive elements of energy poverty, such as energy poverty in Barcelona city participant of EJUSTA program or social exclusion and material deprivation. In contrast, it is less the non-energy poor population of Barcelona (from now on BCN) accurate in capturing severity and persistence of energy poverty, to compare the health status. since it may differ from household to household depending on Two sources of information were used: 1) the EJUSTA Survey the affected dimensions such as access, affordability and needs, database, contain the baseline information of the EJUSTA partici- among other factors.5 Furthermore, the energy poverty does not pants, those were recruited consecutively at public social service occur as an isolated problem. It is frequently experienced in com- offices during the 5-months (March to July 2016) of the execution bination with other housing problems or insecurities in other of the program, and 2) the database of the Barcelona Public Health vital areas as work/employment and food.11 For example, people Survey (BPHS) 2016 edition. BPHS provides information from a rep- at risk of poverty experience day-to-day difficulties in fulfill- resentative sample of non-institutionalised Barcelona population ing their essential needs like food, housing, utilities, and health obtained with a multi-stage random sampling. The questionnaire costs.12 As a result of their limited opportunities and challenges used was adapted from the BPHS using the same validated items.26 in balancing reduced household budgets, they are likely to suf- So, identical health measures were included in both surveys along fer adverse health consequences, an issue that needs to further with demographic, socioeconomic, housing and energy poverty exploration. questions. Trained independent interviewers delivered the survey Energy poverty has severe health and health inequality in both population face to face (only one member per household). implications.13,14 The critical pathways of energy poverty to ill- For the present study, EJUSTA participants that wished to health are the poor thermal and humidity conditions, inadequate receive free home energy counselling visits were also invited to dwelling quality, and the unaffordability of domestic energy participate in the research and their informed consent obtained. services.13 These conditions have been associated with higher Individuals who declined to participate in the study also received excess winter mortality and morbidity due to respiratory15–17 and 440 J. Carrere et al. / Gac Sanit. 2021;35(5):438–444 the intervention. Inclusion criteria were ≥18 years, signed the Social class was measured using the Spanish Society of Epidemi- informed consent and affected by energy poverty based on the con- ology’s classification,31 which is based on the current or last sensual approach:21 inability to keep the home warm or inability occupation. Then, the responses were classified as a non-manual to pay utility bills on time or presence of dampness and rot in the class and manual class. We sampled the manual class population dwelling. Of 3094 participants, 2470 met the selection criteria (204 without any energy poverty measures. Results can be found in did not sign the informed consent; 248 were non-energy poor; 172 Table I in online Appendix. unknown energy-poor condition). The BCN individuals were sam- Among the EJUSTA sample, a descriptive analysis of the energy pled from the BPHS following the criteria of ≥18 years and people poverty measures was performed. Then, the prevalence of age- not affected by energy poverty using the same energy poverty standardised health outcomes by energy poverty intensity level measures but in the reverse sense. Therefore, individuals with one we calculated. The association between poor health outcomes and or more energy poverty indicator was excluded. Of the 4000, 2608 energy poverty intensity was analysed by estimating the PR and met the criteria (560 were < 17 years, 799 reported at least one their 95%CI for all health outcomes by using Poisson regression energy poverty measure, 33 unknown energy-poor condition). models with robust variance. The included explanatory variables The Clinical Research Ethics Committee has approved the study were all covariables, except the inability to face an unexpected protocol of Parc de Salut Mar (Number 2016/6681/I). expense of D 750 or more, because of collinearity with energy poverty intensity. Orthogonal polynomial contrasts were used to Study variables test the gradient association between poor health outcomes and energy poverty intensity. Four physical and mental health outcomes were analysed in this All the analysis was carried out stratifying by gender, using the study. Self-reported health was measured using the question “How statistical software STATA13. is your health in general?”, and the answer was categorised as good (excellent, very good, and good) and poor (fair, and poor). This has proven to be a good indicator of health status, health service util- isation and, mortality.27 Presence or absence of asthma, chronic Results bronchitis and depression and/or anxiety in the previous 12 months were measured. Among the self-reported morbidity indicators, res- Table 1 shows substantial differences in the demographic, piratory morbidity and the mental health disorders stand out for socioeconomic and housing characteristics between the EJUSTA the high concordance between the registers and the self-reported participants and the BCN. For example, the number of people born values.28,29 outside the EU-15 was close to half in the EJUSTA (women, 49.0%; The main independent variable was energy poverty measured men, 42.2%), but was much lower in the BCN (women, 17.5%; men, by three self-reported questions:21 “Could your household afford 15.6%). Single mothers among the EJUSTA doubled that of the BCN to keep the dwelling heated at an appropriate temperature dur- (21.3% versus 10.4%, respectively). Both educational attainment and ing the winter months?”, “Has your household been in arrears on socioeconomic position were lower in the EJUSTA than in the BCN, utility and community bills in the past 12 months?” and “Does and most participants in the EJUSTA were tenants (women, 76.8%; your dwelling have any leaks, dampness on the walls, floors, ceil- men 72.7%), while most of BCN were owners (women, 66.2%; men ings or in the foundations, or rotten floors, windows or door 63.4%). frames?”. The answers were combined to create two variables: 1) The EJUSTA reported worse health conditions than the BCN in all “EP (de)composition”: a variable with seven categories based on all physical and mental health outcomes (Table 2). Women reported the combinations of the EP measures. 2) “EP intensity”: a 3-point a higher prevalence in all health outcomes in both populations, categorical variable: “low EP” (yes in one measure); “moderate EP” although we did not observe gender-differentiated association pat- (yes in two of them); and, “high EP” (yes in all). terns. The probability of experiencing poor health was several Relevant covariables included were age, sex, place of birth times higher in the EJUSTA sample than in the BCN sample. The (Spain and EU-15; outside the EU-15); household composition strongest associations were found in chronic bronchitis and depres- (one-person household; couple without children; couple with chil- sion and/or anxiety. The corresponding PRs were 4.94 (95%CI: dren; single parent; other); educational level (less than primary; 3.68–6.63) for women and 5.43 (95%CI: 3.79–7.78) for men in primary; secondary or more); employment status (employed; the case of chronic bronchitis and PR 3.23 (95%CI: 2.85–3.68) for unemployed; retirees and pensioners; permanent disability); and women and 4.00 (95%CI: 3.29–4.86) for men in the case of depres- ability to face an unexpected financial expense of D 750 or more sion and/or anxiety. The sensitivity analysis performed using a (yes/no) that provide information about of individuals’ preparation subsample of BCN manual social class revealed similar association against financial risks.30 Tenure status (owner; tenant; social hous- patterns (Table I in online Appendix). The most noteworthy associa- ing) was also measured because of its relationship with dwelling tions were again found in chronic bronchitis and depression and/or conditions. anxiety. A high proportion of EJUSTA participants reported that they Statistical analysis could not afford to keep their households adequately warm (women, 78.0%; men, 74.8%) and a high proportion of them had A descriptive analysis of all the variables in EJUSTA and BCN arrears on utility bills (women, 68.3%; men, 68.1%). The leaks, samples was conducted. Pearson’s ␹2 test tested the significance damp or rot indicator was less frequently reported (women, of the differences between the groups. The prevalence of age- 29.5%; men, 27.4%). The Euler diagram in Figure 1 shows the standardised health outcomes was calculated for both groups using impact according to the measured dimensions of energy poverty, the direct method, with the age distribution of the EJUSTA sam- displaying a substantial but not homogeneous overlap. ple as a reference in both populations. Next, the prevalence ratios Table 3 shows the association between energy poverty intensity (PR) and their 95% confidence intervals (95%CI) for all health out- and health status among the EJUSTA sample. For most health vari- comes were estimated using Poisson regression models with robust ables, point estimates of PRs were indicative of a gradient associa- variance adjusted by age. tion. The linear trend test confirmed them, except for chronic bron- As a sensitivity analysis, all the analyses were repeated, select- chitis in men. For all health indicators, there were significant differ- ing non-energy poor manual class adult population from BPHS. ences in poor health between low energy poverty and high energy J. Carrere et al. / Gac Sanit. 2021;35(5):438–444 441 Table 1 Sociodemographic, socio-economic, and housing characteristics of the EJUSTA participants and the Barcelona non-energy poor population, stratified by sex. Barcelona, 2016. Women Men EJUSTA (n = 1799) BCN (n = 1399) p EJUSTA (n = 671) BCN (n = 1215) p n % n % n % n % Sociodemographic variables Place of birth Spain and EU-15 914 51.0 1,149 82.5 < 0.001 386 57.8 1,026 84.4 < 0.001 Outside the EU-15 877 49.0 244 17.5 282 42.2 189 15.6 Missing 8 - 3 1 Household composition One-person household 351 20.1 185 13.3 < 0.001 203 31.2 115 9.4 < 0.001 Couple without children 173 9.9 397 28.5 112 17.2 383 31.5 Couple with children 621 35.5 537 38.5 243 37.4 508 41.8 Single parent 372 21.3 144 10.4 27 4.2 90 7.4 Other 233 13.2 129 9.3 65 10.0 120 9.9 Missing 49 - 21 - Educational level Less than primary 325 18.4 104 7.5 < 0.001 109 16.7 44 3.6 < 0.001 Primary 853 48.2 382 27.5 323 49.6 358 29.6 Secondary or more 591 33.4 904 65.1 219 33.6 808 66.8 Missing 30 3 20 6 Socioeconomic variables Employment status Employed 404 23.3 684 59.6 < 0.001 93 14.4 708 59.0 < 0.001 Unemployed 791 45.6 94 6.8 345 53.4 86 7.2 Retirees and pensioners 307 17.7 325 23.5 132 20.4 298 24.8 Permanent disability 179 10.3 47 3.4 72 11.2 46 3.8 Others 54 3.1 230 16.7 4 0.6 62 5.2 Missing 64 14 25 14 Ability to face an unexpected expense of D 750 or more Capable 60 3.4 1,102 81.1 < 0.001 24 3.7 1,024 86.2 < 0.001 Incapable 1,704 96.6 258 19.0 626 96.3 164 13.8 Missing 35 32 21 28 Housing variable Tenure status Owner 384 21.7 907 66.2 < 0.001 160 24.2 761 63.4 < 0.001 Tenant 1,006 56.8 383 27.9 340 51.4 373 31.1 Social housing 354 20.0 20 1.5 141 21.3 16 1.3 Other 28 1.6 62 4.5 21 3.2 51 4.3 Missing 27 22 9 14 EJUSTA participants: participants in Energia, la justa program affected by energy poverty; BCN: Barcelona non-energy-poor population. Table 2 Age-standardised prevalence (%) and adjusted prevalence ratio (PR) of physical and mental health in the EJUSTA participants compared with the Barcelona non-energy poor population, stratified by sex. Barcelona, 2016. Women (n = 2849) Men (n = 1595) Prevalence PR 95%CI Prevalence PR 95%CI Poor health BCN 24.0 1 17.5 1 EJUSTA 58.6 2.60a 2.34-2.88 56.7 3.23a 2.78-3.76 Asthma BCN 4.9 1 4.6 EJUSTA 15.6 3.22a 2.45-4.22 11.8 2.23a 1.61-3.10 Chronic bronchitis BCN 3.9 1 3.1 EJUSTA 18.0 4.94a 3.68-6.63 17.1 5.43a 3.79-7.78 Depression and/or anxiety BCN 17.9 1 11.5 EJUSTA 57.5 3.23a 2.85-3.68 46.7 4.00a 3.29-4.86 EJUSTA: participants in Energía, la justa program affected by energy poverty; BCN: Barcelona non-energy-poor population. 95%CI: 95% confidence interval; PR: prevalence ratio. a p-value <0.001. poverty. Depression and/or anxiety in women and asthma in men women, immigrants, people with low levels of education, unem- were statistically significant at all energy poverty intensity levels. ployed and tenants. These population suffered a strikingly higher prevalence of poor self-perceived health, respiratory and mental Discussion morbidity compared to the non-energy poor population. More- over, among those suffering from energy poverty, the most severely This study shows that the vulnerable population affected by affected were found to be the most likely to report poor health energy poverty in a southern European city was disproportionately outcomes, regardless of other sociodemographic factors. 442 J. Carrere et al. / Gac Sanit. 2021;35(5):438–444 Home inadequately warm Arrears on utility bills Leaks, damp or rot 3.2% 3.1% 6.4% 6.1% 4.8% 5.1% 15.0% 13.1% 22.1% 13.9% 22.7% 17.5% 34.6% 32.4% Women Men Figure 1. Energy poverty prevalence and (de)composition among EJUSTA participants, stratified by sex. Barcelona 2016. Table 3 Prevalence (%) and adjusted prevalence ratio of physical and mental health by energy poverty intensity in the EJUSTA participants, stratified by sex. Barcelona, 2016. Women (n = 1726) Men (n = 642) Prevalence PR 95%CI Prevalence PR 95%CI Poor health Low energy poverty 55.0 1 53.5 1 Moderate energy poverty 56.6 1.05 0.95-1.15 53.3 1.01 0.86-1.18 High energy poverty 66.9 1.32 1.17-1.491.46 73.7 1.34b 1.10-1.62 *** ** p-trend Asthma Low energy poverty 14.0 1 8.2 1 Moderate energy poverty 14.8 1.10 0.84-1.45 13.4 1.70a 1.04-3.07 High energy poverty 22.3 1.79b 1.23-2.36 13.6 2.05a 1.01-4.15 ** p-trend * Chronic bronchitis Low energy poverty 15.1 1 13.9 1 Moderate energy poverty 17.2 1.17 0.92-1.50 16.7 1.09 0.72-1.65 High energy poverty 27.3 1.88c 1.40-2.52 21.8 1.67a 1.00-2.82 *** p-trend Depression and/or anxiety Low energy poverty 51.3 1 38.5 1 Moderate energy poverty 56.3 1.13a 1.02-1.24 44.7 1.20 0.98-1.47 High energy poverty 62.8 1.37c 1.23-1.57 64.1 1.80c 1.43-2.27 *** *** p-trend 95%CI: 95% confidence interval; PR: prevalence ratio. Note: All prevalence ratios were adjusted by age, place of birth, household, employment status, education level and tenure status. a p < 0.05. b p < 0.01. c p < 0.001. p-trend: orthogonal polynomial linear contrast. a p-trend <0.05. b p-trend <0.01. c p-trend <0.001. Note: Energy poverty intensity: “low energy poverty” (one energy poverty indicator); “moderate energy poverty” (two energy poverty indicators); and “high energy poverty” (all energy poverty indicators). Health status among EJUSTA and non-energy poor population a socially disadvantaged subpopulation whose health is likely to worsen as more and more social problems coexist. There is abun- The vulnerable populations studied reported severe adverse dant evidence of the association between health and socioeconomic physical, and mental health outcomes compared to the non-energy status.32 However, this does not fully explain the association found poor population. The results support the findings of previous between energy poverty and poor physical and mental health. The studies showing that low-income people, particularly women,24 sensitivity analysis performed, as well as the energy poverty assess- lower social classes, immigrants, and tenants, are more exposed ment by intensity level, confirmed this association. Both results to the structural determinants of energy poverty and its health support the conception of energy poverty as a condition related, but consequences.14,20 As previously noted, the household’s inability different from other socioeconomic vulnerabilities.33 These results to ensure the required levels of domestic energy services tends to were consistent with the evidence found in previous studies related coexist with other problems,11,12 like financial difficulties, unem- to asthma,15,17 chronic bronchitis,16 and poor self-perceived health ployment, and residential insecurity. The population studied was and mental health problems,14,19 suggesting the problem goes J. Carrere et al. / Gac Sanit. 2021;35(5):438–444 443 beyond the mere existence of economic poverty or material depri- term than income poverty, as factors other than income can be vation. addressed to alleviate energy poverty. Such an approach should, at the same time, also be tackled as a matter of social justice. There- Association between energy poverty intensity and health status fore, targeted policies and programmes to alleviate energy poverty among EJUSTA should coexist with broader structural policies aimed at improving living, working, and housing conditions. Both strategies combined Among the EJUSTA population, the self-reported energy poverty would reduce health effects and health inequalities in populations indicators showed a pervasive but uneven overlap in their dimen- affected by multiple social deprivations. sions. It was noted that the higher the energy poverty intensity, the higher the prevalence in almost all health problems studied. However, the energy poverty affects health through different path- ways. For example, respiratory problems have been related to poor dwelling conditions.17 In contrast, mental health problems have What is known about the topic? been more closely linked to worry about debt and affordability, thermal discomfort and concern about its health consequences.19 Energy poverty is a severe problem in Eastern and South- ern European countries. Living in energy poverty is associated Such a suggestive pattern merits further exploration, not only with poor physical and mental health. However, most of the about energy poverty intensity patterns suggested but also how evidence was gathered in other European contexts, which dis- the different pathways combined affect people’s health. In future play differences both in climatic conditions and economic and research, this approach should also address the inability to keep political structural determinants. the home cool in summer, a problem that may become relevant in coming years given the climate change but remains relatively What does this study add to the literature? under-explored within EU.34 Vulnerable populations affected by energy poverty in a Limitations and strengths Southern European city showed high prevalence of poor self- perceived health, respiratory and mental morbidity. Moreover, the higher the energy poverty experienced, the greater the The main limitation was the existence of multiple confounding probability of physical and mental health problems. factors between the target and comparison populations, threat- ening its external validity. Although extrapolation of the health burden related to energy poverty should be done with caution, the evidence gathered from the analysis of the energy poverty intensity suggests an unequivocal and consistent relationship with several Editor in charge health outcomes. The sample was useful and a good first approach, but the comparison group far from ideal. Another limitation is the Carlos Álvarez-Dardet. reliability and accuracy of the self-perceived energy poverty mea- sures used due to possible reporting bias. For example, households Transparency declaration that would be empirically identified as energy-poor with other metrics, may not identify themselves as energy-poor or may deny The corresponding author on behalf of the other authors guar- their circumstances.33 In any case, the measures were asked iden- antee the accuracy, transparency and honesty of the data and tically in both populations. Lastly, this study may show a selection information contained in the study, that no relevant information bias due to the non-random sample used among EJUSTA partici- has been omitted and that all discrepancies between authors have pants. For example, the sample did not include people in extreme been adequately resolved and described. socioeconomic circumstances affected by energy poverty who were not social service users (e.g. illegal immigrants, squatters, or people Authorship contributions in extreme poverty). If these population groups had been included, the health outcomes would probably have been worse and the Study conception, planning, analysis, interpretation, and writ- health associations even more substantial. ing: J. Carrere. Study conception, insights for the analysis, and Despite these limitations, this is the first study analysing the interpretation and writing: A. Peralta, L. Oliveras, M.J. López, M. association between energy poverty intensity and health in the con- Marí-Dell’Olmo, J. Benach and A.M. Novoa. All authors gave final text of Southern Europe. The advantages of a composite indicator approval and agree to be accountable for all aspects ensuring the that compares energy poverty level across households its abil- integrity and accuracy of the manuscript. The corresponding author ity to capture a multidimensional concept that often is appraised attests that all listed authors meet authorship criteria and that no on separate components.35 and poorly studied in relation with others meeting the criteria have been omitted. This article forms health status. This composite indicator, although “rudimentary”, part of the doctoral dissertations of J. Carrere, at Universitat Pompeu has allowed revealing how the energy poverty level was associated Fabra, Barcelona. in gradient-shaped with different health outcomes. This approach may offer to policymakers the opportunity to design programmes according to the level of energy poverty, considering its possible Funding associated health consequences. The study was partly funded by a research grant from the Carlos Conclusion III Institute of Health, Ministry of Economy and Competitiveness (Spain), awarded on the 2015 call under the Health Strategy Action This study highlights the shocking health inequalities among 2013-2016, within the National Research Programme oriented to the vulnerable population adversely affected by the energy poverty Societal Challenges, within the Technical, Scientific and Innova- in an urban context in the south of the EU. In Southern European tion Research National Plan 2013-2016, with reference PI15/02006, regions tailored programmes to reduce energy poverty are urgently co-funded with European Union ERDF funds (European Regional needed because energy poverty is more likely to change in the short Development Fund). 444 J. Carrere et al. / Gac Sanit. 2021;35(5):438–444 Conflicts of interest 17. Mendell MJ, Mirer AG, Cheung K, et al. Respiratory and allergic health effects of dampness, mold, and dampness-related agents: a review of the epidemiologic evidence. Environ Health Perspect. 2011;119: None. 748–56. 18. Barnett AG, Dobson AJ, McElduff P, et al. Cold periods and coronary events: Acknowledgement an analysis of populations worldwide. J Epidemiol Community Health. 2005;59:551–7. 19. Liddell C, Guiney C. Living in a cold and damp home: frameworks for The authors thank the Associació Benestar i Desenvolupament understanding impacts on mental well-being. Public Health. 2015;129: and the Associació Ecoserveis for their support, particularly to 191–9. 20. Gibson M, Petticrew M, Bambra C, et al. Housing and health inequalities: a syn- Mònica Plana Izquierdo. Joan Benach gratefully acknowledges the thesis of systematic reviews of interventions aimed at different pathways linking financial support by ICREA under the ICREA Acadèmia programme. housing and health. Health Place. 2011;17:175–84. 21. Healy JD. Housing, fuel poverty and health: a pan-European analysis. London, UK: Ashgate Publishers; 2004. p. 250. Appendix A. Supplementary data 22. Fowler T, Southgate RJ, Waite T, et al. Excess winter deaths in Europe: a multi- country descriptive analysis. Eur J Public Health. 2015;25:339–45. Supplementary data associated with this article can be found, in 23. Thomson H, Snell C, Bouzarovski S. Health, well-being and energy poverty in Europe: a comparative study of 32 European countries. Int J Environ Res Public the online version, at doi:10.1016/j.gaceta.2020.07.007. Health. 2017;14:584. 24. Clancy J, Daskalova V, Feenstra M, et al. Gender perspective on access to energy References in the EU. Brussels: Directorate General for Internal Policies Policy Department C: Citizens’ Rights and Constitutional Affairs; 2017. Nr. 596. 816. Available at: https://www.researchgate.net/profile/Marielle Feenstra/publication/ 1. Bouzarovski S, Tirado S. The energy divide: integrating energy transitions, 335691585 Gender perspective on access to energy in the EU/ regional inequalities and poverty trends in the European Union. Eur Urban Reg links/5d761acf299bf1cb80932836/Gender-perspective-on-access-to-energy- Stud. 2017;24:69–86. in-the-EU.pdf. 2. Oliveras L, Peralta A, Palència L, et al. Energy poverty and health: trends in the 25. Gonzalez I. Gender inequality and energy poverty. A forgotten risk factor. European Union before and during the economic crisis, 2007-2016. Heal Place. Barcelona: Engineering Without Borders; 2017. 2020:102294. 26. Bartoll X, Baranda L, González JM, et al. Manual metodològic de 3. Dagoumas A, Kitsios F. Assessing the impact of the economic crisis on energy l’Enquesta de Salut de Barcelona 2016/17. Barcelona: Agència de poverty in Greece. Sustain Cities Soc. 2014;13:267–78. Salut Pública de Barcelona; 2018. Available at: https://bcnroc. 4. Tirado S, Jiménez L. Energy poverty, crisis and austerity in Spain. People Place ajuntament.barcelona.cat/jspui/bitstream/11703/114216/1/Enquesta-Salut- and Policy. 2016;10:42–56. Barcelona-2016-17-Manual-metodologic.pdf. 5. Bouzarovski S, Petrova S. A global perspective on domestic energy depriva- 27. Idler EL, Benyamini Y. Self-rated health and mortality: a review of twenty-seven tion: overcoming the energy poverty-fuel poverty binary. Energy Res Soc Sci. community studies. J Health Soc Behav. 1997;38:21–37. 2015;10:31–40. 28. Heliövaara M, Aromaa A, Klaukka T, et al. Reliability and validity of interview 6. Boardman B. Fuel poverty: from cold homes to affordable warmth. London: data on chronic diseases The Mini-Finland Health Survey. J Clin Epidemiol. Belhaven Press; 1991. p. 267. 1993;46:181–91. 7. Recalde M, Peralta A, Oliveras L, et al. Structural energy poverty vulnerability 29. De Sanjosé S, Antó J, Alonso J. Comparación de la información obtenida en una and excess winter mortality in the European Union: exploring the association encuesta de salud por entrevista con los registros de Atención Primaria. Gac between structural determinants and health. Energy Policy. 2019;133:110869. Sanit. 1991;27:260–4. 8. Bouzarovski S. Energy poverty in the European Union: landscapes of vulnerabil- 30. Guio A, Gordon D, Marlier E. Measuring material deprivation in ity. Wiley Interdiscip Rev Energy Environ. 2014;3:276–89. the EU: indicators for the whole population and child-specific 9. Ferrera M. The ‘Southern Model’ of welfare in social Europe. J Eur Soc Policy. indicators. Methodologies & Working Papers. 2012. Available at: 1996;6:17–37. https://ec.europa.eu/eurostat/documents/3888793/5853037/KS-RA-12- 10. Herrero S. Energy poverty indicators: a critical review of methods. Indoor Built 018-EN.PDF. Environ. 2017;26:1018–31. 31. Domingo-Salvany A, Bacigalupe A, Carrasco JM, et al. Propuestas de clase social 11. Hulse K, Saugeres L. Housing insecurity and precarious living: an Australian neoweberiana y neomarxista a partir de la Clasificación Nacional de Ocupaciones exploration. AHURI Final Rep. 2008;124:1–51. 2011. Gac Sanit. 2013;27:263–72. 12. Hernández D. Understanding ‘energy insecurity’ and why it matters to health. 32. World Health Organization. Closing the gap in a generation: health equity Soc Sci Med. 2016;167:1–10. through action on the social determinants of health. Final report of the Com- 13. Liddell C, Morris C. Fuel poverty and human health: a review of recent evidence. mission on Social Determinants of Health. Geneve: World Health Organization; Energy Policy. 2010;38:2987–97. 2008. 14. Marmot Review Team. The health impacts of cold homes and fuel poverty. 33. Thomson H, Bouzarovski S, Snell C. Rethinking the measurement of energy Friends of the Earth & the Marmot Review Team. 2011;342. poverty in Europe: a critical analysis of indicators and data. Indoor Built Environ. 15. Rudge J, Gilchrist R. Excess winter morbidity among older people at risk of 2017;26:879–901. cold homes: a population-based study in a London borough. J Public Health. 34. Thomson H, Simcock N, Bouzarovski S, et al. Energy poverty and indoor cooling: 2005;27:353–8. an overlooked issue in Europe. Energy Build. 2019;196:21–9. 16. Rudge J, Gilchrist R. Measuring the health impact of temperatures in dwellings: 35. Charlier D, Legendre B. A multidimensional approach to measuring fuel poverty. investigating excess winter morbidity and cold homes in the London Borough Energy J. 2019;40:27–53. of Newham. Energy Build. 2007;39:847–58.