Exploring care from extended family through rural women ’ s accounts of perinatal mental illness – a qualitative study with implications for policy and practice

Extended family often have an important role in caring for women experiencing perinatal mental illness; but rural women ’ s perspectives are under-researched. We explored women ’ s experiences of living in rural northern England and receiving care from extended family during periods of perinatal mental illness through 21 quali- tative interviews. Key findings were that companionship, practical support - informal childcare and transport, and emotional support were important forms of care - filling gaps in formal service provision. Findings highlight women ’ s needs for support from extended families in rural areas. The rural infrastructure and inequity in formal services can create vulnerability for women.


Introduction
In high income countries it is estimated between 10 and 15% of women will experience perinatal mental illness (mental illness during pregnancy and the first year after childbirth) (World Health Organization, 2020). Perinatal mental illness (from here referred to as PMI) comprises a spectrum of conditions that may be new or exist prior to pregnancy, including; depression, anxiety, post-traumatic stress disorder, obsessive compulsive disorder and psychosis (Maternal Mental Health Alliance, 2019). Amongst other adverse outcomes, PMI cause suffering for women, may affect the bond between women and their baby and can potentially affect the long-term health and well-being of children (Blackmore et al., 2013;Dubber et al., 2014;Fearon et al., 2010;Glasheen et al., 2013). In the UK, the formal prevention and treatment of PMI are managed in a variety of ways. This includes through universal services (e.g. health visitors, maternity services and primary care), the improving access to psychological therapies programme (from here referred to as IAPT), secondary mental health services (e.g. community mental health teams) and in some areas specialist perinatal mental health services. Regional and geographical differences in the configuration of these formal prevention and treatment services are currently being addressed by policies seeking to equalise access and pathways of care for women (Maternal Mental Health Alliance, 2019;NHS England, 2016;NHS England and NHS Improvement, 2018). To make services equitable, the needs of different populations should be considered and understood. Rurality is one determinant of health which may affect formal and informal care (care provided in personal networks outside of formal services) differently (Sutherns, 2005).
Evidence from high, middle-and low-income countries shows a higher prevalence of the most common PMI -anxiety and depression -in women in rural compared to urban areas (Ginja et al., 2020;Mollard et al., 2016;Villegas et al., 2011). In a study in the US it is noted that features of the rural infrastructure could help to explain this, including the reduced number of formal health and social care prevention and treatment services, distance to travel to services, and poor public transport infrastructures which make accessing dispersed services difficult (Mollard et al., 2016). Cultural explanations from Australian and US studies have suggested that in rural areas there may be a greater stigma associated with mental health and self-reliant attitudes could discourage women from disclosing early symptoms of PMI to health professionals for fear of judgement (Fuller et al., 2000;Jesse et al., 2008).
Alongside formal care, informal care from both partners and extended family (here defined as own parents or step-parents, parentsin-law, and siblings) is vital during periods of PMI regardless of the condition (Lever Taylo et.al., 2019). Lack of perceived and actual social support are consistently found to be risk factors for perinatal anxiety and depression (e.g. Beck, 2002;Biaggi et al., 2016;Lee et al., 2007). Research on partner involvement has found that women often report them as a main source of support (Holopainen, 2002), and perceived support from partners can reduce women's risk of depression (Cox et al., 2008). Having a partner with PMI can lead to mental illness for male partners themselves (Letourneau et al., 2012). There appear to be few studies about care from extended families during PMI. One study in urban and rural populations found that for first-time mothers with depression, their mothers and sisters were their main sources of care, alongside partners (Leahy-Warren et al., 2012). In a recent qualitative study in a mainly urban English population Lever Taylor et al. (2019) highlighted the importance, as well as the complexity, of partner and extended family interaction with formal PMI services. A review of research from the US suggests informal care from family is especially important for women in rural areas due to the limited formal infrastructures of care (Mollard et al., 2016) and that rural women's own mothers are particularly vital for providing practical support (Gjesfjeld et al., 2012;Mollard et al., 2017;Mollard et al., 2016). To our knowledge no studies have explored rural women's experiences of informal care from extended family during PMI in the UK.

Sociological approaches to family and care
Care is not easy to define, though there is agreement that it includes elements of practical and emotional support which help to sustain people in their everyday lives (Lynch, 2007;Thomas, 1993). Responsibilities for informal care are gendered, and women's roles in families have been particularly associated with raising children and providing care in family relationships through illness and/or disability (Morgan, 2011). However, though they feel responsible, families may not always have the capacity to offer the care that others expect (Smart, 2007). Many adult women are engaged in paid work, and it may now be more difficult than in previous generations to offer the care family members may want/need (Gerson, 2009). Importantly, sociological work in a range of populations illustrates that people's experiences of health and illness can be affected by socially shaped gendered expectations they have of extended family members to offer care (e.g. Richardson et al., 2007;Woodgate et al., 2008). Historically, the expectation that women provide care in families has been especially strong in rural communities where gender roles tended to be more defined (Little, 2003) and life perceived to be more communal (Bryant and Pini, 2011). However, rural communities are increasingly heterogeneous with more people moving from outside the areas who do not have the family structures they might once have had or have been perceived to have (Local Government Association and Public Health England, 2017).
To contribute to research about informal care during periods of PMI, and social science scholarship about mental health and rurality more broadly, our paper focuses on rural women's experiences in this health context. We aim to explore the care rural women value from these networks. Our focus is how the place effects of living in a rural area, alongside socially constructed notions of care extended families should offer, shape women's experiences. We do this drawing on empirical qualitative data from the accounts of women who live in rural northern England. These insights can help with understanding the needs of women who experience PMI and identifying where policy and practice could best focus resources. We note, for women who have partners they are often an important part of their informal networks of care during PMI (Lever Taylor et al., 2019). However, because accounts of care, or lack of care, from extended family were pervasive in the accounts of the rural women we interviewed, these personal networks are our focus here.

Methods
Data for this study were collected through qualitative interviews between March and August 2017. The interviews were part of a larger regional mixed methods study which aimed to understand the prevalence of PMI (Ginja et al., 2020) women's experiences of PMI, and practitioner experiences of PMI service provision in rural areas. A case study area in rural northern England was selected as it is a very large region and predominantly rural, with around one third of people living in 'sparse', and one quarter living in 'super sparse' areas. The UK Department for Environment, Food and Rural Affairs (Defra, 2015) defines a rural area as settlements below 10,000 people or which are open countryside, and 'sparse setting' as an area where the number of households is low and the wider area is remotely populated.

Sample and recruitment
Women were eligible to take part if they: 1) lived in an area of the region considered rural according to the Defra Rural/Urban categorisation (Defra, 2015;Local Government Association and Public Health England, 2017) and 2) had experienced (by self-report) mental illness during their pregnancy or in the year after giving in birth, or at the time of interview had symptoms of anxiety or depression based on the Edinburgh Postnatal Natal Depression (McBride et al., 2014) or the Whooley screening tools (Bosanquet et al., 2015).
Women were recruited to the study through a previously completed survey as part of the larger study, flyers distributed through health professionals delivering universal services (health visitors and midwives) a one-off press release through the local media or face-to-face in community groups. The participants who had completed the survey were contacted by the study team via telephone and those recruited through the flyers and press release contacted the research team directly by email, text or telephone. Demographic data were collected from women who expressed an interest in taking part including their age, selfreported PMI and postcode data to identify their level of rurality based on the Defra Rural/Urban categorisation (Defra, 2015;Local Government Association and Public Health England, 2017) and assess socio-economic status using the Index of Multiple Deprivation (Department for Communities and Local Government, 2010). The Defra Rural/Urban categorisation consists of six rural (and four urban) context/settlement combinations (shown in Table 1). In this definition towns have the most dwellings followed by villages, hamlets and isolated dwellings (Defra, 2015).The Index of Multiple Deprivation scale combines six domains to measure deprivation in small geographical areas (ward level). Home postcodes are given an index of multiple deprivation score from 1 to 10, with 1 being most deprived and 10 being least deprived (Department for Communities and Local Government, 2010). Purposive sampling techniques (Patton, 2002) were used to identify the final participants. We aimed to achieve variation in the sample in the areas of the region in which the women lived, category of rurality and age of participants. As the study progressed, we recruited women who could speak to emergent topic areas identified e.g. women who did not have access to their own transport.

Data collection
The method of semi-structured qualitative interviews with open questions was used so the women could talk about aspects of their experiences of PMI that were most important to them (Doucet et al., 2010). The interviews focused on women's emotional and mental health during the perinatal period and their engagement with formal prevention and treatment services. Participants were also asked directly about their social support networks but accounts of care from extended family were commonly discussed unprompted by women throughout the interviews. Eighteen interviews were conducted face-to-face in women's homes (n = 16) or at community venues (n = 2) and three interviews were conducted by telephone. Offering telephone interviews enabled women to take part who may not otherwise have been able. Two female researchers collected the data for the study (DS and KJ). Interviews lasted between 30 and 80 min and were recorded using a digital recording device.

Data analysis
The analysis was completed in two stages using techniques of thematic analysis (Braun and Clarke, 2006). The audio files of interviews were transcribed, checked for accuracy and anonymised by the researchers. This process also facilitated familiarisation. Two researchers (KJ and DS) inductively identified codes in the data which were constantly compared across the transcripts and used to identity sub-themes. In data workshops, two other team members (EH and RL) helped to agree the broader overarching descriptive themes.
Our interpretation in this paper is based on a thematic analysis of the data specifically about one major theme from the first stage of the analysis -'Support from extended family'. We systematically inductively coded all the data about support from extended family and developed sub-themes. These data were compared with data from other sub-themes from the first stage of the analysis including those that related to rurality, support from formal services and support from partners. Data were then framed within sociological literature about care and family. Together they were used to generate the interpretation presented in this paper. The coding frame for this study is found in Appendix A.

Ethical issues
The project was given ethical approval by the Newcastle and North Tyneside 1 NHS Research Ethics Committee (212364). The researchers gained verbal and written consent from the women prior to the interviews. When some women became upset during the discussions the researchers offered to stop the recording device, comforted women and checked whether they wished to continue. In all cases the women wanted to continue. The researchers verbally de-briefed women at the end of the interviews and written information was provided containing details of support they could access if they wished. Participants were given a £10 voucher as a thank you for taking part. All data were fully anonymised, and the names attached to the illustrative quotations in the following section are pseudonyms. Table 1 shows the characteristics of the 21 women we interviewed. Six women lived in a rural hamlet and isolated dwelling (three of these in a sparse setting), nine women lived in a village (three of these in a sparse setting) and six women lived in a rural town and fringe (three in a sparse setting). Fifteen participants resided in areas in the two least deprived quartiles on the Index of Multiple Deprivation Scale (6-10), four were in the two most deprived quartiles (1-5) and two did not report postcode. Mainly women had experienced the most common mental health conditions, i.e. anxiety and/or depression, which were managed through universal services and/or IAPT. Three women had required an inpatient stay for either postpartum psychosis or post-natal post-traumatic stress disorder. Twelve women were first time mothers. All the women were white British. In this region of northern England 96% of people across all age groups were white British when census data was collected in 2011 (Office for National Statistics, 2011).

Findings
All but one woman had a partner and most women spoke about their partners being an important form of informal care. Although there was a complexity within this. Some partners worked long hours or sometimes in seasonal jobs, typical in rural areas, which could keep them away from home for extended periods of time. Geographical isolation and limited services for pregnant women, mothers and babies, were perceived to restrict their opportunities to build networks with other women. Many of the women's accounts suggested they felt that having extended family living locally should make the perinatal period -a time of emotional, social and physical change easier. In this section we discuss three types of care that women particularly valued from extended family: 1) Companionship -supporting and filling gaps in care, 2) Practical support -informal childcare and access to transport 3) Emotional support -providing reassurance and advice. A fourth theme captures the 'Unmet expectations of care' in the women's accounts.

Companionshipsupporting and filling gaps in care
The women's accounts illustrated the heterogeneity of their extended family networks. Approximately half had family living nearby -which for most meant within a neighbouring village or town. A similar number of women's families lived out of the immediate area. Many women who had family living nearby said these networks had provided everyday companionship and supported their wellbeing during early motherhood. For example, Sally who lived on a farm in the same hamlet as her parents and parents-in-law and saw them every day remarked: Days could go by potentially where people don't even see another person. From my own point of view, as I said, I am very lucky that I do have a good support network… here there is not much to do. (Sally, postnatal depression, rural hamlet and isolated dwelling) A minority of the sub-sample of women with family living nearby felt their family had not offered them the companionship or care they needed, or wanted, during early motherhood. This is explored further in the theme 'Unmet expectations of care'.
Most women who did not have family living locally, discussed lacking day-to-day companionship. Their accounts suggested they could be geographically isolated and feel lonely. Nicola's case illustrates several sub-themes in this sub-sample. She had moved to the village where she lived about a year before she had her first child and had no friends or family locally. She said that there were no groups for new mothers and babies in her village, and the women she met at centralised antenatal groups in local towns lived too far away to be day-to-day companions. The women's accounts suggested these prevention services were limited in all the rural locations, but were particularly limited in the most rural or sparse areas. Nicola described how she was alone with her baby for long periods, while her husband was at work: I expected that my husband would be able to take an equal share in things, but despite the fact that he's great and did as much as he could, in reality, he was going to work, and during the day, it was me and the baby. (Nicola, postpartum psychosis and postnatal depression, rural village) Nicola's experience of her partner not being present all the time and the change in her role is not unique to rural women during early parenthood (Barnes, 2015). However, lack of companionship from her partner and family, combined with geographical isolation had led to feelings of loneliness which she felt contributed to her becoming unwell in the early stages of motherhood.
When some women first became ill, they had not necessarily recognised their symptoms as PMI or had not wanted to seek help because of a concern about how they were perceived as mothers. It was notable however that for the women who had the companionship of extended family locally, these networks often alongside women's partners, had been instrumental in recognising symptoms of mental illness and supporting and encouraging women to seek help. For example, Lara said her sister, who she spent a lot of time with, had observed a change in her mood over a number of weeks and encouraged her to go to her GP. She explained: My sister came to the doctor with me… [the doctor] asked me how I was feeling and [my sister] just told her, she said, "She's crying all the time, she's not sleeping. (Lara, perinatal anxiety, rural town and fringe) It should be noted that not all women with family companionship had sought help sooner than women without family companionship. However, the accounts of women without family companionship suggested without the encouragement of family it was often only when professionals had approached them, or they reached crisis, that they had sought or received help for their symptoms. Grace who lived in a rural village in a sparse setting related delaying help seeking to her hesitancy about recognising symptoms herself and the rural infrastructure. She explained that her mental health deteriorated over a month when her health visitor had not visited her because she was 'over stretched' owing to the large geographical area she had to cover. Grace explained when her health visitor finally came to visit and saw how unwell she was she had been quickly referred to IAPT for perinatal depression.
When their symptoms were worst, or when they reached crisis, the extent of companionship and emotional support many women spoke about receiving from their extended families was striking. Some said extended family members spent a lot of time with them.
I was very panicky being on my own so I would spend a lot of time at my Mum and Dad's house. (Donna, perinatal anxiety, rural hamlet and isolated dwelling) A few women had moved in with their parents or parents-in-law, or family members had moved in with them for a period. Moreover, geographical proximity to family did not seem to make a difference to the care family offered when women were most ill. Several women whose families did not live nearby said family members had temporarily moved to be with them.
Some women attributed the need for intense support from family to limited formal treatment services and long waiting lists for treatment in their area. Gemma explained her sister-in-law who did not live close by had spent time with her and provided informal childcare when she was most ill, which had filled what she perceived as a significant gap in formal care: My sister-in-law… came in like an angel and did everything for me and if I hadn't had that, you know, my son and my husband, he wouldn't have got … well I'm not saying he wouldn't have got fed, but if I had been a single parent it would have been very different. (Gemma, postnatal depression and post-natal post-traumatic stress disorder, rural village in a sparse setting) However, after this initial crisis Gemma explained, without extended family who provided companionship, she could be on her own solely with baby for long periods. Jess discussed a complex history with her parents who had provided informal childcare to her baby during a period of crisis but did not provide companionship most of the time. She explained how she was often alone with her daughter when she was feeling unwell: I don't really have a support network, it's my husband, and that's it".... But I was really low ... I felt like I couldn't look after my daughter properly, which wasn't the case, because she's perfectly fine. She was with me most of the day, because my husband worked, he'd leave at half one, and get home just after midnight. (Jess, preexisting anxiety and depression and post-natal depression, rural town and fringe) On the other hand, the accounts of women who had family nearby who offered companionship all the time suggested they provided a lot of valuable support and could fill gaps in formal provision. For example, Penny said that while she was on the waiting list to access IAPT for perinatal anxiety she had to have extended family 'babysitters', with her constantly when her partner was at work. … it was a very long process and in the middle of all this I'm having to have babysitters come and look after me every day because I couldn't be on my own. (Penny, perinatal anxiety, rural hamlet and isolated dwelling in a sparse setting) A few women, both with and without family companionship, had paid for private treatment to help fill some of the gaps in formal care (e. g. counselling or hypnotherapy), whereas others said they could not afford private care.

Practical support -informal childcare and access to transport
The difficulty in accessing formal childcare in rural areas due to its limited availability and high costs were often noted. For some women, local family provided informal childcare so they were able to do things that made them feel better, such as socialising with friends or engaging in hobbies, or so some women could return to work. In this sense family could be seen to support the women's autonomy during this period. Companionship and informal childcare from family also helped several women's partners to return to work, this could help to moderate the financial implication of the illness.
Women also valued extended family for providing informal childcare so they could access treatment. PMI health services often did not accommodate women's children during treatment sessions, and therefore some women relied on informal childcare to enable them to access this formal support.
I've been having counselling every week and I've come on in leaps and bounds… my Mum and Dad kept these two (children) and it's so helpful. (Tina, postnatal depression and post-traumatic stress disorder, rural village) For several women informal childcare from family was perceived to have moderated the potentially harmful effects that the illness might have had or be having for their children. This was illustrated in Gemma's account about care from her sister-in-law in the last section, it is also illustrated by Donna who said: [My daughter] was young enough for it to not affect her at the time. Daddy would just take her to the farm, and she would be with Grandma. (Donna, perinatal anxiety, rural hamlet and isolated dwelling) The difficulty of living in rural areas without access to transport was acknowledged by many women, who both had access to their own transport and those who did not. This was particularly notable for women in the most sparsely populated areas where transport was extremely limited and the distance to travel to services was increased. Some women explained that family members had provided transport by car so they could access treatment which was often centralised in a local town. Several women mentioned the infrequency of public transport in rural areas, and others explained the challenge of accessing public transport when they were feeling unwell. Their symptoms could make them anxious or worried about using public transport.
[The counselling service] didn't really ask if I had childcare… it is quite a way to go when you don't drive and you have mental health (difficulties) and you don't want to get on transport and things so I think that would be something that is difficult for people and why sometimes they don't go to appointments and things. (Andrea, perinatal anxiety and depression and obsessive compulsive disorder, rural hamlet and isolated dwelling) Indeed, a few women mentioned they had missed treatment appointments because they could not access transport and did not have childcare available. A few women who were taking medication for their illness explained this meant they could not drive, which without family support could leave them geographically isolated.

Emotional supportproviding reassurance and advice
The women valued emotional support from extended family. This included the opportunity to talk about their experiences, ask for advice and gain reassurance (for example, about childbirth, baby weight, breastfeeding, etc). Geographical proximity to family did not always matter and some women whose family lived out of the area said they valued emotional support from family over the phone. However, it was notable that a number of women perceived that female extended family members, were more able to offer emotional support than male extended family members, often because they had been mothers themselves. For example, Susie whose mother had died since she had her first child, described that she was finding early motherhood more difficult with her second child without emotional support from her mother: 'My husband is very supportive… My Dad's around. We've got friends. But I think you need your mum, don't you?' (Susie, symptoms of perinatal depression, rural town and fringe in a sparse setting) When women were most unwell, having someone who they could talk to who understood and empathised with how they were feeling was a comfort to them and could make them feel better. Again, commonly female family members were perceived as more able to offer this support than male family members. This is illustrated by Penny who contrasted the type of emotional support she felt she had had from her mother-inlaw with the support she had from her father: my mother-in-law used to come and sit with me … My Dad if I went to his house it was just more that he was there if I needed him rather than him giving me the counselling that comes with it as well, the chatting and stuff. (Penny, perinatal anxiety, rural hamlet and isolated dwelling in a sparse setting) However, not all women felt they could speak to extended familyincluding their own mothers. Some women said they did not want to talk to family because they felt they would not understand, or they thought the way they were feeling might reflect badly on them. For example, Tina valued the companionship and informal childcare her family sometimes offered but she explained why she could not open-up to them fully: I can tell my counsellor that I've been having suicidal thoughts and she won't judge me. I can't say that to my Mum. (Tina, postnatal depression and post-natal post-traumatic stress disorder, rural village) Tina mentioned she had tried to seek emotional support from her sister, but her sister had young children and worked long hours, and thus could not offer the care she needed.
In this sense families could not always offer all the care some of the women sought during this period and other support was necessary. Everyday companionship or practical support without emotional support from family could be seen as not fulfilling all the women's need for care. It was notable that many women had tried to seek peer support from other women who also experienced PMI. Several women mentioned there appeared to be few peer support groups exclusively for PMI in the region. Caroline who lived in a rural village described approaching her local children's centre about setting up a PMI peer support group but was told by staff there was not enough need due to the sparsity of the population.

Unmet expectations of care from family
Some women felt let down after not being offered the care from their families, particularly their own mothers, with a few suggesting that this perceived lack of care had contributed to their symptoms. To illustrate, Caroline discussed that her mother who lived in the next town had not offered day-to-day companionship or the emotional support she had expected. Subsequently, living in a rural village where she said there were few women with young children and baby clinics run irregularly, meant she spent long periods on her own with her baby. She said she lacked someone to turn to for emotional support, advice and reassurance. She described: I thought that I might have done some more things with my Mum… I felt she just didn't really have a lot of time to do things with me, and didn't really want to… She'd sometimes just bob [pop] in for literally five minutes, then just go again and be like, "Oh, I'm going." I felt a bit, not hurt,… I felt a little bit disappointed really. (Caroline, postpartum psychosis and postnatal depression, rural village) Caroline perceived this lack of care from her own mother to have contributed to her feelings of anxiety during the first few months of her son's life. This theme is also illustrated by Karen who had had little contact with her mother after the birth of her first child and had recently had a second baby. Although Karen had her parents-in-law living nearby who sometimes provided companionship and informal childcare she described that the lack of any support from her own mother, father and sister increased her feelings of distress: I don't have a relationship with my Mum and I don't really see much of my Dad, and my sister moved down south and our relationship has kind of fallen apart… that's what it is (began to cry)… (Karen, perinatal anxiety, rural village in a sparse setting)

Discussion
This study provides much needed insights into rural women's experiences and expectations of care from extended families during periods of PMI. Other studies find that families are often heavily involved in care giving during times of mental illness in rural areas (e.g. Murry et al., 2011;Parr and Philo, 2003). Our findings support these and demonstrate that configurations of services in rural areas can result in responsibility being placed on extended families of women with PMI in rural areas, to bridge the gap in formal service provision (Hemard et al., 1999). We noted some differences in the experiences of women living in more and less sparsely populated areas, yet overall, the issues identified were applicable to women in all rural geographical locations in this region, to varying degrees. Care from extended family seemed to be especially important when women could not afford private treatment or did not have their own transport. This observation supports the studies from the US and Australia which have found informal care often from maternal grandmothers is critical for the most economically deprived women in rural areas, as health and social care resources are often remote or unaffordable (Gjesfjeld et al., 2012;Mollard et al., 2017). This increased importance of extended family care for women living in more deprived areas could put additional pressure on families who may have fewer resources themselves. Women without support from extended family (either because they live away from family or because family cannot or do not offer care) can be geographically isolated, lonely, and can find it more difficult to access formal services. This can be detrimental to women's own wellbeing. Indeed, some participants linked lack of day-to day companionship to their illness.
Alongside these place effects of the rural infrastructure, the women's reflexive constructions about the care family members should offer were important to their experiences. Gendered expectations of responsibilities of care by different family members were important to how the women felt. In adult life, mothers continue to be expected to offer care to their own children during illness (McKie et al., 2005). When they do not, this can be distressing for those who expect it (Smart, 2007). This was evident in the accounts of some participants in our study who had not received the care from their mothers they expected and wanted and felt this lack of care had contributed to their illness. In a US study, rural women similarly felt a perceived lack of family support had contributed to their post-natal depression as much as physiological components such as 'hormones' or childbirth experience (Mollard et al., 2017). Thus, the socially created notion that extended families, and their own mothers in particular, should offer care during the perinatal period and PMI, affects women in very tangible ways.
Female family members were most involved in caring for women and their children, but some male extended family members were also providing a lot of care, particularly companionship, childcare and transport. This finding echoes the broader societal patterns of informal care in the UK that, regardless of rurality, women are still more involved in care giving than men (Carers UK, 2015). However, due to changes from the past in men and women's working roles women may be less available, and men more available for care giving, than they may have been in past generations (Boyer et al., 2017;Fraser and Warr, 2009). Although the women in our study valued practical forms of care from male extended family members, their accounts suggested they felt men could not always provide the emotional support they needed to make them feel better. Other studies illustrate the complexity and importance of considering gender for understanding men's care practices (Milligan and Morbey, 2016). Some research with male carers of family members with mental illness has suggested that emotional support may not always be compatible with men's notion of what it is to be caring in these contexts (Fraser and Warr, 2009). Gendered norms about the types of care different family members can offer could mean that women may not always be able to get the emotional care they seek in extended families.
Social science scholarship about rurality and mental health from rural geography has focused on situated understanding, and 'deconstructed' the idea of the rural idyll (Duff, 2012;Parr and Philo, 2003). Rural sociologists have also focused on understanding social and cultural practices through lived experience and have paid particular attention to social divisions and inequalities in rural locations (Shucksmith, 2012). In this article, adopting a more broadly sociological approach we have used conceptual ideas from sociology of care and family to show how care practices at a relational level are subjectively experienced through wider gendered inequalities and social change, and illustrated how they interact with the material realties of living in rural area. The expectations of care from family we highlight are found in other locations/populations and are not unique to rural areas. When considered alongside place they illustrate the importance of incorporating the subjective and the material experience for understanding the complexity of rural (and indeed urban) mental health. Thus, we join others in advocating a multi-disciplinary approach (Boyd and Parr, 2008;Laoire, 2001) and drawing on social science scholarship beyond the rural to understand mental health in rural locations.

Strengths and limitations
This was a small-scale qualitative study undertaken in one rural region of the UK and therefore we do not claim it is representative of women's experiences in other rural areas. As the majority of our participants were in the two most affluent quartiles on the index of multiple deprivation scale the sample has not enabled us to fully investigate the experiences of women in most deprived circumstances, or women who do not have partners. Future work should focus on these contexts as deprivation is one of the most important factors related to prevalence of perinatal mental illnesses (Ban et al., 2012) and care from extended family may be more important for women with children who do not have partners (Reschke et al., 2006). Furthermore, as this was a white British sample the experiences of women in other ethnic groups in rural areas need to be considered, as inequalities may be heightened for these groups women (Templeton et al., 2003).
The qualitative methods used in this study have enabled us to carefully explore a sensitive issue of rural women's experience of PMI. It was clear from the participants' accounts that care from extended family was important to women from a wide range of personal circumstances. Highlighting the experiences of women who have experienced perinatal anxiety and depression, and are only in touch with universal services, is useful as the types of support that can help their mental health may not be prioritised as a result of NHS and Local Authority financial strains.

Implications for policy and practice
Recent key policy documentation in England and the UK (NHS England, 2019; NHS England and NHS Improvement, 2018) have acknowledged the 'invaluable' role that family carers, partners in particular, have in supporting women's' recovery from PMI. In this study we illustrate the crucial role extended families can play in shaping women's experiences of PMI during this period. This study highlights 7 that policy and practice should be aware of women's interdependence in extended family relationships as well as the heterogeneity of these networks in rural areas. It is essential that informal care from extended family is considered and asked about in professional relationships at all stage of the perinatal period. Professionals should be alert to the vulnerability of women who do not have family living in the immediate area who can offer companionship and practical support. They should also be aware of women who do not perceive themselves to have good emotional care in extended family relationships. This is key to a wholistic psychosocial assessment during the perinatal period.
Another key implication of the findings is the part that formal services could play in enabling the types of care we have identified women value from families. While it may not be possible to establish face-to-face groups for mothers in all localities, there may be scope for improving opportunities for companionship by connecting women in rural areas (ideally antenatally to enable relationships to develop) through virtual methods of communication such as video links or telephone contact (where internet access might be poor in rural areas). Moreover, some women reported having minimal contact with their midwives and health visitors and consideration could be given to finding ways to check in with women more regularly using virtual methods. The lack of emotional support some women report receiving from family should be addressed in other relationships, for example through commissioning of peer support and befriending services which (while recognising the importance and sometimes preference for face-to-face contact) could also be set up virtually where face-to-face contact is more difficult. Additionally, there should be a greater acknowledgement of the importance of transport and childcare for women attending treatment and ensure there is adequate provision so that women without extended family care are not disproportionately affected. Policy makers should also consider the support needs of informal carers for women with perinatal illness in rural areas. This would be aligned with the agenda to increase support to unpaid family carers in the UK (Department of Health and Social Care, 2018).

Funding
The work was funded by the County Council in the region of the UK where the work was carried out.

Declaration of competing interest
There are no competing interests. Practical support