We applied the Population, Concept, Context (PCC) structure to set our inclusion and exclusion criteria (Table 1) [20]. The population group included people of any age with no additional limitations set. The concept, and focus of this review, was the role of fuel poverty on mental health. Fuel poverty was defined as having difficulties or being unable to afford to pay for gas or electricity to do basic household tasks like cook, turn on the lights, run a fridge, wash, and maintain a healthy indoor temperature. All measures of fuel poverty including self-reported difficulty with fuel and energy bills, quantitative financial measures of expenditure on fuel, and proxy measures such as reporting living in a “cold home” were included. Mental health was defined broadly, encompassing mental wellbeing and common mental disorders. In terms of context, all countries were considered. The context of fuel poverty was the home. Other residence types (such as institutionalised populations) were not included. Studies assessing only fuel availability, fuel type or physical barriers to fuel access were excluded due to incompatibility with our conceptualisation of fuel poverty and incomparability with the majority of the studies identified by the review. We limited inclusion of studies to empirical research and grey literature. Both quantitative and qualitative studies were included as no restrictions on methodological approach were imposed.

We conducted our search on: Medline (Ovid), PubMed Central, APA PsychInfo (Ovid), Web of Science, Scopus, Embase (Ovid), Social Policy Practice (Ovid), Econlit (Ovid) and The Cochrane Library. Grey literature was searched in databases including: OpenGrey, Grey Literature Report, the WHO database, Department for International Development research output database. The search was conducted on the 29th March 2023 with no date restrictions imposed. A second search was conducted on the 12th of February 2024 to capture any further articles. Search strings are available in the appendix.

All identified studies were screened against our pre-defined eligibility criteria (Table 1). Title and abstract screening were completed by LM and CL. Full text screening was performed by NB and SK. Where there was uncertainty at any stage of the screening process, this was resolved with discussion between authors in line with scoping review methodology [21, 22]. Included studies were extracted by two authors (NB & SK) into a table created by the authors prior to the extraction phase. Extracted data were aggregated and checked for consistency by two authors (NB & SK) and any discrepancies resolved. Themes, definitions, quantitative measures, and results from included studies were extracted. We employed a narrative approach to the synthesis of the data extracted from the included studies in relation to fuel poverty and mental health associations and the pathways involved in the relationship.

Public contributors were involved throughout the project to help guide the research and to ensure the relevance of the project and its findings to those with lived experience of fuel poverty. We regularly engaged with a steering group of public contributors from the NIHR Applied Research Collaboration for Greater Manchester and held public involvement sessions with members of two community centres (Larkspur and Edberts House) in Newcastle upon Tyne.

Our database search returned a total of 7740 studies, after removal of duplicates 4761 studies went through title and abstract screening of which 114 full-text articles were screened for inclusion (Figure 1). A total of 47 studies were included in the final analysis, 42 from peer-reviewed journals and 5 from grey literature.

Included studies were mostly quantitative (n = 26) compared to qualitative (n = 19), with just two studies using a mixed methods approach.

In total, 60% of studies were published within the last 3 years 2023 (n = 3), 2022 (n = 19), 2021 (n = 7); highlighting the topical nature of fuel poverty research. The remaining studies (n = 18) dated over the last two decades back to 2005.

Most included studies were based in the United Kingdom (UK) (n = 20), eight were based in Spain, six in China, four in Australia, three in the United States of America, two in Ireland and one study from each of the following countries: Belgium, Canada, Hungry, New Zealand, North Macedonia, Norway and Poland. Table 2 provides a full breakdown of these data.

Many different definitions of fuel poverty were used among the included studies. Often, the quantitative measure used to capture those in fuel poverty was also operationalised as the definition of the concept. Broadly, the definitions used can be categorised into four types, based on: unmet energy needs (22 studies); a ratio based on energy expenditure and income (five studies); indicative qualities of the physical household environment (three studies); and finally, any combination of the former definitions (five studies). These definitions were variously operationalised as measures taking the form of objective calculations or as self-reported measures. A total of 12 studies provided no clear definition for fuel poverty, with five of these being quantitative studies and seven being qualitative.

The 26 quantitative studies identified households and individuals in a state of fuel poverty using an applied quantitative measure which could be categorised by the aforementioned conceptual definitions. See Supplementary Table SA1 for a detailed summary of results by study.

Studies applying a quantitative measure most commonly used a single method for measuring fuel poverty (n = 16). A Ratio based definition was the most commonly used of these (n = 7). Of these, six used Boardman’s’ ratio by applying a threshold at which if a households energy expenditure is greater than 10% of their income they can be classed as being in fuel poverty [24], one used Hills’ Low Income High Cost (LIHC) indicator to apply two thresholds (one on each income and cost), at below 60% of median income and at costs higher than median modelled bill [25]. Only one of these studies measured income after tax and housing costs [26]. An unmet energy need definition was second most commonly used (n = 5). Of the studies applying an unmet energy need definition to measure fuel poverty, four used an indicator based on financial difficulties regarding energy bills and one used an indicator based on behaviour changes in usage of energy.

Four studies measured fuel poverty using a physical household environment definition, of which, three used a subjective question asking participants questions regarding the ability to keep their homes warm and one study used an objective measure by taking recordings of internal household temperature.

Seven studies used a combination of physical household environment with an unmet need, two used physical household environments with a ratio and one used unmet need with a ratio measure.

The majority of quantitative studies used an instrument measure (n = 20) to quantitatively measure mental health. Of these studies, seven used the Centre for Epidemiologic Studies Depression Scale (CES-D), four used the General Health Questionnaire 12 (GHQ-12), three used the Mental Component of the Short Form 12 Health survey (SF12-MCS), two used the Short Form 36 health survey. Additionally, the following measures were each used once by different studies: the Short Warwick-Edinburgh Mental Wellbeing Scale (SWEMWBS), the Strengths and Difficulties Questionnaire (SDQ) (a screening tool for mental health specifically for children and young people), the Perceived Stress Scale (PSS), and the Composite International Diagnostic Interview Short Form (CIDI-SF).

The five remaining quantitative studies did not use validated instruments to capture mental health, and instead used various self-reported indications of having experienced depressive disorders or anxiety.

Forty-six of the 47 included studies found a detrimental association between fuel poverty and mental health, with one quantitative study reporting no effect [27].

Of the 26 quantitative studies six studies showed heterogeneity in results by different population groups, including age, sex, ethnicity, students in further education, deprivation and parenthood. Three of these studies showed a large negative association between fuel poverty and mental health in elderly populations [11, 28, 29]; two studies reported worse outcomes in parenthood with mothers experiencing higher likelihood of postpartum depression [30, 31]; one study reported negative associations in vulnerable groups experiencing housing insecurity [32]; and one study reported differences between men and women, demonstrating worse outcomes among men [12].

Eleven of the 19 qualitative studies focused on specific population subgroups. Four of these were based on financially vulnerable groups, where households were chosen to partake based on living in social housing, low income and difficulties with fuel bills. All groups reporting worsening mental health due to fuel poverty [33–36]; two studies focused on single parents, both reporting detrimental impacts [37, 38]; one study focused on women and stated detrimental outcomes [39]; another study assessed students with disability and ethnicity of students, showing detrimental outcomes with respect to both groups [40]; and finally one study carried out research on adults with learning difficulties, reporting that their mental health was detrimentally impacted by fuel poverty [41].

Of the studies included, 37 discussed the pathways thought to operate between fuel poverty and mental health. Where possible pathways were discussed, studies frequently mentioned more than one. Broadly, these pathways can be summarised under the following headings: environmental (n = 20); economic (n = 20); behavioural (n = 10) and social (n = 9) (Figure 2).

We include pathways under the environmental heading where studies described impacts to mental health as a result of the physical environment of the home. Most commonly studies described thermal discomfort (n = 15) as pathway to poor mental health via direct impacts to psychological functioning. For example, Sawyer and colleagues report that the physical health problems cold homes worsen, may also further exacerbate poor mental health [42]. The stress caused by the physical deterioration of the home was also discussed in two studies. The authors also further linked this to the physical health problems that come with mould and damp within households [11]. The concept of “spatial shrink,” whereby the extent of the home lived in is significantly reduced as households decide to heat fewer rooms, was mentioned frequently [30, 36, 42–44]. Authors suggest this not only impacts household relationships via reduced privacy [44], but may also impact the way people interact with their homes, eroding a sense of home and security [30].

Included studies discussing economic pathways described financial insecurity, and often debt, that occurred with fuel poverty (n = 15). Studies often described the challenges households face surrounding the prioritisation of essentials which we term “difficult decision making” (n = 5). Studies commonly referred to the “heat or eat” dilemma, where healthier or heated food options were sacrificed or meals skipped in order to prioritise spending on energy bills [34, 45]. Finally, studies also referred specifically to a fear of receiving bills themselves and these unknown costs (n = 6). All of these were reported to negatively impact mental health with increased stress, worry and anxiety.

Social pathways described hardships encountered as a result of fuel poverty. These included social isolation (n = 6), loneliness (n = 3) and social stigma (n = 6). In studies describing these pathways, fuel poverty was often said to cause withdrawal from social networks and activities due to cost or reluctance to leave the home, and reduced visits from family and friends at home due to housing temperature and condition [42, 43].

Finally, we use the “behavioural” pathways heading to describe significant behaviour changes made by households and individuals in order to lessen the impacts of fuel poverty. These included consciously restricting and carefully planning the use of essential appliances such as washing machines, ovens and central heating in order to optimise the available budget [41, 46] as well as changing hygiene practices, such as forgoing or reducing showering [33]. Studies describing these pathways commonly reported the toll that constant consideration, calculation and modification of these everyday activities has on individuals, as well as the fear of receiving bills which remains. Moreover, studies also frequently cite impacts to sleep (n = 3) and a loss of a sense of home (n = 6).

In general, the qualitative studies more commonly discussed potential explanations for the link between fuel poverty and mental health and explored these themes with participants. Though the quantitative studies often cited potential explanations, these were less thoroughly described and were very rarely empirically investigated in analyses.

A strength of our review is its focus on the impact of fuel poverty on mental health, which has to our knowledge, not yet been examined systematically across the global literature base. Moreover, we applied a broad range of terminology, and included literature of both quantitative and qualitative methodologies allowing for a better approach in exploring the pathways between fuel poverty and mental health. This approach has allowed us to provide a robust summary of the state of this growing and increasingly important research field. Furthermore, we were able to describe some of the explanatory factors for the association between fuel poverty and mental health.

However, our review was also subject to limitations. Firstly, as described above, the absence of a commonly understood definition of fuel poverty meant comparability between studies, especially across countries, was challenging. In addition, despite allowing for a global focus, this review predominantly contained European countries with the exception of the United States, Canada, New Zealand, Australia and China. As we took a scoping review approach this study, we did not perform critical appraisal of the included studies and therefore cannot comment on the quality of the studies or the robustness of their results.

Our findings have several implications for future research. Firstly, clearer conceptualisation of fuel poverty is necessary in future studies in order to elucidate the mechanisms operating between fuel poverty and mental health. Secondly, the value of qualitative research in this field is clear in the contributions which have been made to understanding the relationship to date. Future research may benefit from the adoption of mixed methods approaches in order to advance our understanding of the mechanisms at the population level. Finally, this review identified sub population groups who are more vulnerable to the deleterious impacts of fuel poverty. Further research on the experience of fuel poverty in these vulnerable groups is essential.

This scoping review mapped the current research on fuel poverty and mental health. We found a lack of conceptual clarity on the term “fuel poverty” and identified a wide range of definitions and measures used to date. Overall, the majority of studies, both qualitative and quantitative identified a detrimental association between fuel poverty and mental health. Explanations for this relationship fell into four broad themes: environmental, economic, social, and behavioural. Though it is widely accepted that certain population groups are more vulnerable to the risks of fuel poverty and mental health impacts, few studies analysed these groups specifically. We argue that future work should focus on better understanding these pathways linking fuel poverty and mental health, particularly how these operate in at-risk groups, in order to identify possible interventions.