A mixed studies systematic review on the health and wellbeing effects, and underlying mechanisms, of online support groups for chronic conditions.

This systematic review was conducted in concordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Supplementary Table  1 ) 29 . The systematic review was pre-registered prior to the search with Prospero, registration number: CRD42023399258. The final review deviated from the pre-registration protocol reported, as the authors did not repeat the search for conditions that were not in the initial search strategy (e.g., endometriosis). The number of research questions differ from the pre-registration. We are no longer comparing outcomes between different types of chronic conditions due to the small number of papers for most of the included chronic conditions. In line with recommendations 30 , the following databases were searched for peer-reviewed publications on September 11th, 2024: Embase 1974 to September 11, 2024 Ovid MEDLINE® ALL 1946 to September 11, 2024 APA PsycInfo 1806 to September Week 1 2024 Web of Science Core Collection Embase 1974 to September 11, 2024 Ovid MEDLINE® ALL 1946 to September 11, 2024 APA PsycInfo 1806 to September Week 1 2024 Web of Science Core Collection Grey literature searches were also conducted to identify any eligible reports not published via academic publishers, to ensure comprehensiveness, on November 29th 2024, using Google Advanced Search (first 200 items), and Google Scholar (first 200 items). The British Library directory of online doctoral theses (EThOS) was searched on February 14th 2023, without any date restrictions. MedRxiv and PsyArXiv were searched to identify any pre-publication articles uploaded between January 1st, 2023 and September 11th 2024, where we searched the 200 most relevant articles. Search terms were based on the target population (i.e., those with a chronic condition) and intervention (i.e., online support groups). To avoid unintentionally excluding articles, the study outcomes were not included in the search terms as they relate more broadly to health and wellbeing as opposed to specific outcomes (e.g., depression). Search terms were developed by the research team based on previous reviews on similar topics 16 , 31 , the types of chronic conditions listed by the National Health Service (NHS) 1 and preliminary literature searches. See Supplementary Tables  2 – 6 for the full search strategy. The full inclusion and exclusion criteria are detailed in Table  1 . The review included quantitative, qualitative, and mixed method studies (excluding reviews, conference abstracts and protocols). Studies from any country were included, if they were published in English, due to the languages spoken by the research team. The review used the following definition of chronic conditions when deciding eligibility of studies: a health problem that requires ongoing management over a period of years or decades and is one that cannot currently be cured, but can be controlled with the use of medication and/or other therapies 1 . Table 1 Inclusion and exclusion criteria Area of study Inclusion criteria Exclusion criteria Population Individuals experiencing symptoms of, or diagnosed with, a chronic condition Individuals not experiencing symptoms of, or not diagnosed with, a chronic condition. Intervention All types of online peer support groups, involving more than two participants, will be included. The support groups can be asynchronous (e.g., social media group, forum, email) or synchronous (e.g., video call) and can be text- and/or video-based. The review will also include support groups with or without a moderator (led by either a peer (i.e., someone with lived experience) or professional). In-person support groups, unless they are a comparator for online support groups, will be excluded. The study will also exclude complex interventions (i.e., when a support group is one component of a broader intervention), unless the study specifically reports the independent effects of online support groups. Comparator The review will include studies with or without a control group. Control groups may include, but are not limited to, in-person support groups, waitlist or standard care. N/A Outcome Changes in physical health may include but are not limited to: symptom presence; symptom duration; symptom severity; and changes in limitations in activities due to such physical symptoms 144 , 145 . Changes in wellbeing may include, but are not limited to: psychological functioning, social relationships, meaning making and changes in limitations in activities due to such symptoms 144 , 145 . These can be clinical or self-report measures. Outcomes that are not related to the health and wellbeing of support group members. For example, studies assessing the acceptability of support groups. Inclusion and exclusion criteria Changes in physical health may include but are not limited to: symptom presence; symptom duration; symptom severity; and changes in limitations in activities due to such physical symptoms 144 , 145 . Changes in wellbeing may include, but are not limited to: psychological functioning, social relationships, meaning making and changes in limitations in activities due to such symptoms 144 , 145 . These can be clinical or self-report measures. Results of the literature searches were exported into the review screening website Rayyan 32 . The first author conducted initial title and abstract screening, where each title was categorised to be either ‘included’ or ‘excluded’ for full-text screening. To improve the robustness of the review process 20% of articles underwent title and abstract screening by the third author 33 . The authors agreed on 96% of the articles. The ‘include’ articles then underwent full text screening by the first author whereby all articles were categorised into either ‘include’ or ‘exclude’. During full-text screening, the inclusion criteria were tightened to ensure the studies specifically refer to online support group use; to focus on individuals currently experiencing the chronic condition; to exclude studies that focused exclusively on chronic mental, as opposed to chronic physical, health conditions; and to exclude mechanisms relating to offline influences. The third author screened all of the excluded studies 33 , where there was 100% agreement. The screening process can be seen in the PRISMA flowchart in Fig.  1 . Fig. 1 PRISMA flow diagram of the identification of studies. PRISMA flow diagram of the identification of studies. Data were extracted in a tabular form on Microsoft Excel by the first author. Results were synthesised using a data-based convergent approach (also called an integrated approach), whereby quantitative and qualitative studies are analysed using the same synthesis method and results are presented together 34 , 35 . Quantitative data underwent data transformation, which involved creating textual descriptions of quantitative findings. Findings of health and wellbeing were coded in themes using thematic synthesis. Each finding was first coded as ‘outcome’ or ‘mechanism’. To organise the data, each health outcome was coded into themes (e.g., physical health). Findings within each theme were then coded into sub-themes based on the specific finding (e.g., pain), with similar codes being grouped together. This was an iterative process with the grouping of codes and themes changing following discussions amongst the research team. For mechanisms, each finding was reported in relation to their respective health outcome and were thematically grouped for the discussion (e.g., support). A thematic synthesis was deemed appropriate for the research questions, rather than a meta-analysis, due to the heterogeneity of quantitative studies across health outcomes, in terms of research design and findings. Indeed, whilst all included studies explore the role of online support groups on group members’ health and wellbeing, some focus on specific aspects of the groups (e.g., receiving information or level of engagement), whilst others compare the groups to control groups, such as receiving education or treatment as usual. The Mixed Methods Appraisal tool (version 2018) was used to evaluate the quality of included studies 36 . This was a suitable appraisal tool as it was designed for systematic reviews that include qualitative, quantitative, and mixed methods studies. The tool comprises of two questions that apply to all studies, followed by five questions relevant to each methodology. The first author carried out the quality appraisal. Further information on research design is available in the  Nature Portfolio Reporting Summary linked to this article.

In total, 21,599 results were extracted from electronic databases and grey literature searches. Duplication screening was conducted on Rayyan, resulting in 13,527 articles for title and abstract screening. Full text screening was conducted on 329 papers, with 100 papers being included in the final review. A summary of the characteristics of each study can be found in Table  2 . Numbers of articles excluded for each reason during the full text screening can be found in Supplementary Table  7 . Across the 100 included papers, 24 chronic conditions were included, with breast cancer ( n  = 21), other types of cancer ( n  = 18), HIV/AIDS ( n  = 9) and diabetes ( n  = 7) being the most frequently studied. A full list of included chronic conditions can be found in Table  3 . Table 2 Included studies First author First author location Study design Chronic Condition Online support group description Comparator description Health and Wellbeing Outcome Algtwei et al. (2017) 146 UK Cross-sectional survey ( n  = 199) Cancer Online supports with at least 50 members and at least 25 threads within the past 30 days N/A Depression Anxiety Quality of life Self-efficacy Ashtari and Taylor (2022) 58 USA Cross-sectional interview ( n  = 30) Ehlers-Danlos Syndrome Online peer support groups hosted by Facebook, Instagram, Twitter, Reddit and Inspire Not participating in an online support group to find information (20%) Pain Broad mental health and wellbeing Behaviour change Treatment decision-making Feeling less alone Identity Ashtari and Taylor (2023) 66 USA Mixed methods: Cross-sectional survey ( n  = 413) Cross-sectional interview ( n  = 30) Ehlers-Danlos Syndrome Not specified N/A Broad mental health and wellbeing Feelings of belonging Babyar (2016) 115 USA Cross sectional survey ( n  = 57) Cystic Fibrosis Not specified N/A Adherence Bartlett and Coulson (2011) 102 UK Cross-sectional survey ( n  = 246) Various conditions Active online support groups with more than 5 members with the primary aim of providing support for patients with a chronic physical illness N/A Broad social wellbeing Treatment decision-making Illness acceptance Self-esteem Optimism and control Batenburg and Das (2014) 61 The Netherlands Longitudinal survey ( n  = 133) Breast cancer Not specified N/A Broad mental health and wellbeing Depression Batenburg and Das (2014) 62 The Netherlands Cross-sectional survey ( n  = 175) Breast cancer Dutch peer-led message boards available 24/7 with messages posted in the last month N/A Broad mental health and wellbeing Depression Batenburg and Das (2015) 86 The Netherlands Cross-sectional survey ( n  = 114) Breast cancer Online 24 h-available peer led message boards on breast cancer with messages posted in the last month N/A Broad mental health and wellbeing Depression Baydoun et al. (2021) 59 Canada RCT ( n  = 125) Breast cancer 90 minute once weekly professionally moderated and guided discussions in a synchronous online text based chat session (plus NuCare) NuCare: self-guided psychoeducational manual focused on adopting new ways of coping Broad mental health and wellbeing Depression Anxiety Loneliness and isolation Coping Bazrafshani et al. (2022) 45 Iran Mixed methods: Cross-sectional interview ( n  = 29) Delphi study ( n  = 26) HIV/AIDS Online social networks e.g., Telegram, WhatsApp and Facebook N/A Symptoms and functioning Broad mental health and wellbeing Feelings of belonging Loneliness and isolation Behaviour change Motivation Adherence Treatment decision making Self-efficacy Empowerment Optimism and hope Self-esteem Coping Beaudoin and Tao (2007) 95 USA Cross-sectional survey ( n  = 372) Cancer Email discussion groups and instant messages N/A Depression Coping Brady et al. (2017) 112 UK Cross-sectional interview ( n  = 21) Diabetes Not specified N/A Behaviour change Motivation Treatment decision-making Self-efficacy Empowerment Identity Changrani et al. (2008) 56 USA RCT ( n  = 68) Cancer Virtual Community for Immigrants with Cancer. 8 people in each online group, meeting weekly for 90 minutes and facilitated by trained bilingual facilitators to discuss issues of interest Usual care Pain Depression Quality of life Post-traumatic growth Chen et al. (2020) 101 China Cross-sectional survey ( n  = 1241) Diabetes 73% were members of groups with more than 100 members N/A Connections and friendships Behaviour change Quality of life Self-efficacy Chung et al. (2021) 99 USA Cross-sectional survey ( n  = 124) Parkinson’s Disease (PD) Run by Parkinson’s patients; not commercial operations; active; numbers from 1600 - 2900 members N/A Quality of life Cooper et al. (2021) 42 USA Quasi-experiment ( n  = 9) HIV Secret Facebook group N/A Symptoms and functioning Quality of life Costello et al. (2019) 39 UK Longitudinal survey ( n  = 329) Variety of conditions HealthUnlocked: a host to 700 online health communities built in collaboration with patient organisations, who moderate the communities. Registered members can follow communities, post questions, updates or reply to other posts. They can post text and images N/A Symptoms and functioning Self-efficacy Coulson (2013) 72 UK Cross-sectional open-ended survey ( n  = 249) Inflammatory bowel disease (e.g., Crohn’s disease or ulcerative colitis) Asynchronous online support communities (e.g., discussion forums) N/A Broad mental health and wellbeing Anxiety Loneliness and isolation Behaviour change Illness acceptance Feeling less alone Feeling understood and reassured Optimism and hope Self-esteem Cummings et al. (2002) 67 USA Cross-sectional survey ( n  = 64) Hearing impairment Beyond hearing: email distribution list ( n  = 240) for people who have hearing loss. All emails sent to the list are automatically emailed to other group members. N/A Broad mental health and wellbeing Day (2022) 73 UK Cross-sectional interview (n = 11) Long Covid Facebook and Whatsapp groups N/A Broad mental health and wellbeing Feeling understood and reassured Optimism and hope Egerton et al. (2022) 87 Australia Post-intervention interview ( n  = 10) Knee osteoarthritis My Knee Community: expert-moderated online discussion forum organised into categories and threads. Moderators monitor the posts in the group, post information and answer questions if they requested a response from a healthcare professional. Members could add posts to threads or create new threads Web-based patient education resource including factsheets, videos and other tools (e.g., risk assessments) Broad mental health and wellbeing Behaviour change Motivation Feeling understood and reassured Fullwood et al. (2019) 124 UK Cross-sectional survey ( n  = 271) Various conditions Not specified N/A Illness acceptance Garrett et al. (2024) 76 USA Cross-sectional interview ( n  = 15) Long Covid Social media, specifically Facebook groups N/A Broad mental health and wellbeing Anxiety Connections and friendship Loneliness and isolation Feeling less alone Feeling understood and validated Han et al. (2014) 40 USA Quasi-experiment ( n  = 325) Breast cancer Discussion forum component of CHESS: a text-based, asynchronous bulletin board monitored by a trained facilitator N/A Symptoms and function Depression Connections and friendship Self-efficacy Han et al. (2019) 94 USA Quasi-experiment ( n  = 236) Breast cancer Discussion forum component of CHESS: a text-based, asynchronous bulletin board monitored by a trained facilitator N/A Depression Coping Hansen et al. (2021) 111 Norway Cross-sectional survey ( n  = 540) Type 2 Diabetes Not specified Not participating in an online support group Behaviour change Healy et al. (2022) 125 USA Cross-sectional interview ( n  = 35) HIV Existing, organically developed youth WhatsApp groups at HIV care clinics Not part of a WhatsApp group Feeling less alone Herrero et al. (2019) 63 Spain Cross-sectional survey ( n  = 307) Diabetes Facebook group, diabetes-related forum, health related forum and other types of online forums Not belonging to an online support group (39.1% of the sample) Broad mental health and wellbeing Motivation Self-esteem Herrero et al. (2021) 110 Spain Cross-sectional survey ( n  = 307) Diabetes Facebook groups Diabetes related forums Health related forums Not belonging to an online support group (32% of the sample) Behaviour change Hodson and O’Meara (2023) 93 Canada Cross-sectional interview ( n  = 12) Cancer Social media, specifically Facebook groups N/A Depression Anxiety Holbrey and Coulson (2013) 77 UK Cross-sectional open-ended survey ( n  = 55) Polycystic Ovary Syndrome (PCOS) A volunteer run, peer-led discussion forum with 4,600 members. It has been identified as one of the most active for women suffering with PCOS N/A Broad mental health and wellbeing Anxiety Feelings of belonging Treatment decision-making Self-efficacy Optimism and hope Holdren et al. (2023) 105 USA Mixed methods: Cross-sectional interview ( n  = 20) Cross-sectional survey ( n  = 291) Cancer Closed Facebook cancer peer support groups N/A Connections and friendship Feeling less alone Huber et al. (2017) 43 Germany Cross-sectional survey ( n  = 1641) Cancer Largest German online support group with 90,000 postings from 4,400 registered users Face to face support group with a group leader that organises meetings and invites guest speakers Symptoms and functioning Broad mental health and wellbeing Depression Anxiety Distress Quality of life Treatment decision-making Illness acceptance Feeling less alone Self-esteem Huber et al. (2017) 119 Germany Cross-sectional survey ( n  = 686) Cancer German prostate cancer online group ( n  = 3357) N/A Treatment decision making Hurtado Illanes (2024) 68 Spain Cross-sectional qualitative survey ( n  = 26) Chronic patient Virtual community support Blogs Broad mental health and wellbeing Feelings of belonging Behaviour change Treatment decision-making Self-efficacy Iliffe and Thomspon (2019) 74 UK Cross-sectional interview ( n  = 12) Alopecia Alopecia UK Facebook peer support group N/A Broad mental health and wellbeing Feelings of belonging Loneliness and isolation Treatment decision-making Illness acceptance Feeling less alone Identity Coping Kaal et al. (2018) 106 The Netherlands Cross-sectional survey ( n  = 30) Cancer An online community with a community manager (non-health care professional) who is online for 4 hours per day. N/A Connections and friendship Loneliness and isolation Feeling understood and reassured Self-esteem Kever et al. (2022) 41 USA Non-randomised control trial: Efficacy outcomes ( n  = 31) Focus group evaluation ( n  = 17) Multiple Sclerosis (MS) Participants met in an online group of 10 individuals led by a psychologist where group members were encouraged to share challenges, fears, thoughts and coping strategies. Each meeting starts with a ‘win’. Discussions evolved based on participants’ current needs and the input they provided. No treatment control (10 participants) Symptoms and functioning Broad mental health and wellbeing Depression Anxiety Distress Quality of life Loneliness Self-efficacy Feelings of belonging Kim et al. (2012) 85 USA Quasi-experiment ( n  = 177) Breast cancer Discussion forum component of CHESS: a text-based, asynchronous bulletin board monitored by a trained facilitator N/A Broad mental health and wellbeing Illness acceptance Kim et al. (2017) 100 USA Quasi-experiment ( n  = 221) Breast cancer Discussion forum component of CHESS: a text-based, asynchronous bulletin board monitored by a trained facilitator N/A Quality of life Coping Klemm (2012) 88 USA RCT ( n  = 50) Breast cancer Asynchronous online support group moderated by social workers. Moderators introduced weekly topics and facilitated discussions among group members. Asynchronous peer-led online support groups, with open discussion without the use of preselected topics or input from a moderator Depression Kosugi et al. (2021) 147 Japan Cross-sectional survey ( n  = 334) Cancer Cancer Parents: one of the largest online peer support groups in Japan for adult cancer patients ( n  = 3000) N/A Loneliness and isolation Koufopoulos et al. (2016) 114 UK RCT ( n  = 216) Asthma Asthma Village: an online community for patients with asthma where participants could leave comments or see who else was online There was no experimenter intervention Asthma Diary: an online diary for recording ICS preventer use. Participants could not read the posts of others or interact with others Treatment adherence Lange et al. (2017) 60 Germany Quasi-experiment ( n  = 44) Cancer Internet chat programme where prostate cancer patients met for 60-90 minutes. Each session was guided by psychotherapists and had a topic focus Treatment as usual Broad mental health and wellbeing Depression Anxiety Quality of life Coping Leavitt et al. (2020) 90 USA Non-randomised controlled trial survey ( n  = 28) Multiple Sclerosis (MS) eSupport: private facilitated video calls with 6 members per group eJournal: weekly online semi-structured journalling activities Depression Loneliness Letourneau et al. (2012) 104 Canada Pre/Post quantitative survey ( n  = 10) Post-intervention qualitative survey and interview ( n  = 14) Asthma and Allergies Weekly synchronous chat sessions centred around specific topics; and an internet community bulletin board; and emails N/A Connections and friendships Loneliness and isolation Self-efficacy Coping Lieberman et al. (2003) 57 USA Pre/Post survey ( n  = 32 adult) Breast carcinoma Electronic support group: 8 people per group meeting for 1.5 hours once a week with no set agenda. Also had access to a private newsgroup where they could chat, post pictures and share their stories N/A Pain Depression Anxiety Post-traumatic growth Coping Lieberman (2007) 50 USA Longitudinal survey ( n  = 77) Breast cancer Breast cancer bulletin boards that had more than 20 postings a day N/A Symptoms and functioning Depression Quality of life Lieberman and Winzelberg (2009) 53 USA Pre/Post survey ( n  = 91) Breast cancer 5 online peer led asynchronous groups with large active membership with women usually responding within a day to new members. There’s a core of active members that provide continuity and share views on how to cope with the disease and how to fight and actively combat the disease. N/A Symptoms and functioning Broad mental health and wellbeing Depression Broad social wellbeing Litchman (2015) 49 USA Cross-sectional survey ( n  = 183 adults) Cross-sectional interview ( n  = 20) Cross-sectional survey ( n  = 178) Diabetes 4 online communities: 1) TuDiabetes:sharing of blog posts, discussions, events, photos and videos on over 22,000 forums and 400 groups 2) Diabetic Connect: discussion boards and options to share and rate receipes and medication and updated new stories and Ask an Expert 3) Facebook 4) Twitter N/A Symptoms and functioning Broad mental health and wellbeing Quality of life Connections and friendships Feelings of belonging Behaviour change Treatment decision-making Self-efficacy Empowerment Feeling less alone Feeling understood and reassured Lopez-Olivo et al. (2022) 38 USA RCT ( n  = 204) Rheumatoid Arthritis Secret Facebook community and access to an educational website. Online community was peer led by 2 advocates / spokespeople for rheumatoid arthritis. Different topics introduced each week for discussion Educational website Symptoms and functioning Pain Broad mental health and wellbeing Depression Anxiety Quality of life Behaviour change Self-efficacy Empowerment Mackie et al. (2024) 107 Australia Cross-sectional interview ( n  = 28) Breast cancer Teleconference N/A Connections and friendship Empowerment: Optimism and hope Self-esteem Meade et al. (2013) 108 UK Cross-sectional interview ( n  = 6) Neuromuscular disorder Emails to previous participants and general advertisement on the message board N/A Connections and friendship Loneliness and isolation Feeling less alone Feeling understood and reassured Coping Meng et al. (2021) 84 USA Cross-sectional survey ( n  = 386) Cancer Two online communities providing basic social networking features (e.g., personal profile / adding friends) and discussion forums where cancer patients can exchange social support N/A Broad mental health and wellbeing Mills et al. (2024) 46 UK Cross-sectional interview ( n  = 21) Long Covid Facebook, WhatsApp and Zoom groups. N/A Symptoms and functioning Broad mental health and wellbeing Feelings of belonging Connections and friendship Loneliness and isolation Self-efficacy Feeling less alone Mo and Coulson (2010) 28 UK Cross-sectional survey ( n  = 640) HIV/AIDS Using an online support group Not using an online support group Symptoms and functioning Pain Broad mental health and wellbeing Broad social wellbeing Coping Mo and Coulson (2010) 148 UK Cross-sectional survey ( n  = 340) HIV/AIDS Online support groups with at least 25 message threads within the last 30 days and at least 50 members N/A Symptoms and functioning Pain Broad mental health and wellbeing Depression Distress Quality of life Broad social wellbeing Loneliness and isolation Self-efficacy Optimism and hope Coping Mo and Coulson (2012) 122 Hong Kong Cross-sectional survey ( n  = 340) HIV/AIDS Not specified N/A Self-efficacy Coping Mo and Coulson (2013) 123 Hong Kong Cross-sectional survey ( n  = 340) HIV/AIDS Public groups with at least 10 new messages posted to the group each day N/A Depression Loneliness and isolation Illness acceptance Optimism and hope Mo and Coulson (2014) 78 Hong Kong Cross sectional open-ended survey ( n  = 115) HIV/AIDS Not specified N/A Broad mental health and wellbeing Anxiety Connections and friendships Loneliness and isolation Behaviour change Feeling understood and reassured Feeling less alone Optimism and hope Morehouse et al. (2021) 69 USA Cross-sectional quantitative and qualitative survey ( n  = 76) Myalgic encephalomyelitis /chronic fatigue syndrome (ME/CFS) 18 online support groups with a total of 2000 members N/A Broad mental health and wellbeing Feelings of belonging Loneliness and isolation Behaviour change Feeling understood and reassured Self-esteem Optimism and hope Coping Parrish (2011) 70 USA Cross-sectional survey ( n  = 472) Cancer Private Listserv group and its subgroup N/A Broad mental health and wellbeing Distress Quality of life Connections and friendship Feeling less alone Identity Parrocha and Bernadas (2024) 126 Philippines Cross-sectional interview ( n  = 15) Cardiovascular disease Private Facebook group with 3500 members N/A Feeling less alone Pester et al. (2022) 55 USA Randomised controlled trial n  = 119 Chronic pain Professional-led Facebook groups: investigators posted nearly every morning to disseminate training components Standard Facebook groups: encouraged by the investigators to offer mutual support for the duration of the group Pain Depression Anxiety Petrovski and Zivkovic (2019) 44 Macedonia Cross-sectional analysis of electronic health records ( n  = 728) Diabetes Internet group: standard medical protocol plus active members of national closed Facebook group on diabetes Non-internet group: standard medical protocol with regular clinic visits Symptoms and functioning Ronen et al. (2023) 89 USA Pre/Post quantitative survey ( n  = 55) HIV Interactive WhatsApp groups (n = 25) facilitated by study staff for 6 months who sent weekly structured messages. N/A Depression Behaviour Motivation Adherence Self-efficacy Rose et al. (2024) 120 USA Cross sectional survey (n = 219) Cancer 5 Facebook groups for scalp cooling (n = 247 - 9047 members) N/A Feelings of belonging Treatment decision-making Rowlands et al. (2023) 79 UK Cross-sectional focus groups ( n  = 10) Various conditions Not specified N/A Broad mental health and wellbeing Feeling understood and reassured Optimism and hope Russell et al. (2022) 98 USA Cross-sectional interview ( n  = 20) Long Covid Long Covid online community including Reddit forums, Facebook groups and Slack channels N/A Anxiety Empowerment Illness acceptance Salzer et al. (2010) 37 USA RCT ( n  = 78) Breast cancer Unmoderated closed Listserv (email list) Review of information on a cancer-related website Symptoms and functioning Broad mental health and Wellbeing Distress Self-efficacy Optimism and hope Seçkin (2007) 71 USA Cross-sectional survey ( n  = 350) Cancer Active online support groups (i.e., new messages posted on a daily basis) N/A Broad mental health and wellbeing Self-efficacy Illness acceptance Optimism and hope Coping Seçkin (2011) 127 USA Cross-sectional survey (n = 255) Cancer Not specified N/A Optimism and hope Setoyama et al. (2011) 96 Japan Cross-sectional survey ( n  = 220) Breast cancer Online communities that include only those with breast cancer and are not moderated by healthcare providers. N/A Depression Anxiety Shaw et al. (2006) 54 USA Quasi-experiment ( n  = 144) Breast cancer Discussion forum component of CHESS: a text-based, asynchronous bulletin board monitored by a trained facilitator N/A Symptoms and functioning Broad mental health and wellbeing Shaw et al. (2007) 51 USA Quasi-experiment ( n  = 97) Breast cancer Discussion forum component of CHESS: a text-based, asynchronous bulletin board monitored by a trained facilitator N/A Symptoms and functioning Broad mental health and wellbeing Self-efficacy Illness acceptance Shaw et al. (2008) 83 USA Quasi-experiment ( n  = 231) Breast cancer Discussion forum component of CHESS: a text-based, asynchronous bulletin board monitored by a trained facilitator N/A Broad mental health and wellbeing Shim et al. (2011) 52 USA Quasi-experiment ( n  = 106) Breast cancer Discussion forum component of CHESS: a text-based, asynchronous bulletin board monitored by a trained facilitator N/A Symptoms and functioning Broad mental health and wellbeing Self-efficacy Shoebotham and Coulson (2016) 26 UK Cross sectional qualitative survey ( n  = 69) Endometriosis Not specified N/A Pain Broad mental health and wellbeing Distress Loneliness and isolation Treatment decision-making Self-efficacy Empowerment Illness acceptance Feeling understood and reassured Coping Sparling et al. (2017) 64 USA Cross-sectional survey ( n  = 508) Multiple Sclerosis Not specified N/A Broad mental health and wellbeing Steadman and Pretorius (2014) 80 South Africa Cross-sectional interview ( n  = 10) Multiple Sclerosis (MS) Online Facebook support group for people with MS N/A Broad mental health and wellbeing Feelings of belonging Connections and friendship Loneliness and isolation Stewart et al. (2013) 103 Canada Post intervention interview ( n  = 27) Pre/Post survey ( n  = 27) Asthma and Allergies Online weekly video ‘real-time’ support (45–120 minutes) led by peer mentors and health professionals. Groups were matched by condition and gender. N/A Feelings of belonging Loneliness and isolation Behaviour change Self-efficacy Feeling less alone Coping Tam et al. (2023) 117 USA Cross-sectional survey ( n  = 97) Cancer Facebook groups for head and neck cancer ( n  = 967–6,096) N/A Treatment decision-making: Tankha (2023) 149 USA RCT ( n  = 119) Any type of chronic pain Professional-led Facebook groups: investigators posted nearly every morning to disseminate training components Standard Facebook groups: encouraged by the investigators to offer mutual support for the duration of the group Self-efficacy Terborg (2023) 150 UK Cross-sectional survey ( n  = 90) Tinnitus Posted on Tinnitus Talk Not using Tinnitus Talk Behaviour change Thewlis (2021) 65 Ireland Cross-sectional survey ( n  = 76) Chronic illness Not specified Not in an online support group Broad mental health and wellbeing van Uden-Kraan et al. (2008) 22 The Netherlands Cross-sectional interview ( n  = 32) Various conditions Active online support groups (i.e., 50 posts a month) N/A Broad mental health and wellbeing Feelings of belonging Connections and friendships Loneliness and isolation Behaviour change Treatment decision-making Self-efficacy Illness acceptance Feeling less alone Feeling understood and reassured Optimism and hope Self-esteem Coping van Uden-Kraan et al. (2008) 151 The Netherlands Cross-sectional survey ( n  = 528) Various conditions Posters on Dutch online groups Lurkers on the same group (defined as never having posted) Social wellbeing Illness acceptance Self-esteem Optimism and hope van Uden-Kraan et al. (2009) 23 The Netherlands Cross-sectional survey ( n  = 528) Various conditions Active online support groups (i.e., 30 posts a month) N/A Broad social wellbeing Connections and friendships Loneliness and isolation Treatment decision-making Illness acceptance Optimism and hope Self-esteem Vanstrum et al. (2022) 91 USA Cross-sectional survey ( n  = 549) Vestibular disorders Facebook groups N/A Depression Treatment decision-making Optimism and hope Coping Vanstrum et al. (2024) 92 USA Cross-sectional survey ( n  = 97) Vestibular disorders Online support groups with over 20,000 members for general vestibular symptoms Face to face support group Depression Anxiety Treatment decision-making Coping Vilhauer (2009) 97 USA Post-intervention interview ( n  = 20) Pre/post survey n  = 18 Breast cancer Unmoderated email-based support groups. Maximum group membership was limited to 10 or 11 women Wait list for support group Anxiety Feelings of belonging Connections and friendships Isolation and loneliness Treatment decision making Empowerment Illness acceptance Optimism and hope Walsh et al. (2024) 81 USA Cross-sectional interview ( n  = 25) Cancer 3 specified online support groups N/A Broad mental health and wellbeing Connection and friendship Willis et al. (2018) 113 USA Cross-sectional interview ( n  = 20) Arthritis Not specified N/A Behaviour change Illness acceptance Willis et al. (2018) 116 USA Cross-sectional interview ( n  = 20) Arthritis Online health communities dedicated to arthritis and were absent of health care professionals and medical interventions. Each group included discussion boards and archived threads on topics such as medications, exercise and special diet plans N/A Treatment adherence Identity Wilson and Stock (2021) 82 UK Cross-sectional interview ( n  = 15) Long-term conditions (e.g., asthma, chronic fatigue, lupus, fibromyalgia) Not specified N/A Broad mental health and wellbeing Anxiety Feelings of belonging Loneliness and isolation Feeling less alone Identity Wu et al. (2021) 152 Taiwan Cross-sectional survey ( n  = 452) Chronic patients Online communities established by hospitals or healthcare professionals N/A Adherence Yao et al. (2015) 47 China Mixed methods: Cross-sectional interview = ( n = not reported) Cross-sectional survey n  = 349 Hepatitis B The Anti-Heptatitis B Forum - one of the most popular and active N/A Symptoms and functioning Broad mental health and wellbeing Anxiety Quality of life Feelings of belonging Connections and friendship Optimism and hope Identity Zheng et al. (2016) 48 China Cross-sectional survey ( n  = 326) Hepatitis B The Anti-Heptatitis B Forum - one of the most popular and active N/A Symptoms and functioning Broad mental health and wellbeing Quality of life Zhu et al. (2018) 128 China Cross-sectional interview (n = 18) Breast cancer Discussion forum N/A Self-esteem Zigron and Bronstein (2019) 75 Israel Cross-sectional interview ( n  = 23) Ulcerative colitis and Chron’s Disease Closed Facebook health community with 1875 registered members. Community members can share and consult in these areas, recommend treatments, share experiences N/A Broad mental health and wellbeing Feelings of belonging Motivation Empowerment Optimism and hope Table 3 Chronic Conditions identified in the review Chronic Condition Number of papers References Breast cancer 21 37 , 40 , 50 – 54 , 57 , 59 , 61 , 62 , 83 , 85 , 86 , 88 , 94 , 96 , 97 , 100 , 107 , 128 Other/multiple types of cancer 18 43 , 56 , 60 , 70 , 71 , 81 , 84 , 93 , 95 , 105 , 106 , 117 – 120 , 127 , 146 , 147 Various conditions included 11 22 , 23 , 39 , 65 , 68 , 79 , 82 , 102 , 124 , 151 , 152 HIV/AIDS 9 28 , 42 , 45 , 78 , 89 , 122 , 123 , 125 , 148 Diabetes 7 44 , 49 , 63 , 101 , 110 – 112 Arthritis 4 38 , 87 , 113 , 116 Multiple Sclerosis 4 41 , 64 , 80 , 90 Long Covid 4 46 , 73 , 76 , 98 Asthma 3 103 , 104 , 114 Chronic pain 2 55 , 121 Hepatitis B 2 47 , 48 Inflammatory bowel conditions (Crohn’s disease and ulcerative colitis) 2 72 , 75 Ehlers-Danlos Syndrome 2 58 , 66 Vestibular disorders 2 91 , 92 Cystic fibrosis 1 115 Endometriosis 1 26 Hearing impairment 1 67 Myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) 1 69 Parkinson’s Disease 1 99 Polycystic Ovary Syndrome 1 77 Alopecia 1 74 Cardiovascular disease 1 126 Tinnitus 1 150 Neuromuscular disorders 1 108 Included studies Depression Anxiety Quality of life Self-efficacy Pain Broad mental health and wellbeing Behaviour change Treatment decision-making Feeling less alone Identity Mixed methods: Cross-sectional survey ( n  = 413) Cross-sectional interview ( n  = 30) Broad mental health and wellbeing Feelings of belonging Broad social wellbeing Treatment decision-making Illness acceptance Self-esteem Optimism and control Broad mental health and wellbeing Depression Broad mental health and wellbeing Depression Broad mental health and wellbeing Depression Broad mental health and wellbeing Depression Anxiety Loneliness and isolation Coping Mixed methods: Cross-sectional interview ( n  = 29) Delphi study ( n  = 26) Symptoms and functioning Broad mental health and wellbeing Feelings of belonging Loneliness and isolation Behaviour change Motivation Adherence Treatment decision making Self-efficacy Empowerment Optimism and hope Self-esteem Coping Depression Coping Behaviour change Motivation Treatment decision-making Self-efficacy Empowerment Identity Pain Depression Quality of life Post-traumatic growth Connections and friendships Behaviour change Quality of life Self-efficacy Symptoms and functioning Quality of life Symptoms and functioning Self-efficacy Broad mental health and wellbeing Anxiety Loneliness and isolation Behaviour change Illness acceptance Feeling less alone Feeling understood and reassured Optimism and hope Self-esteem Broad mental health and wellbeing Feeling understood and reassured Optimism and hope Broad mental health and wellbeing Behaviour change Motivation Feeling understood and reassured Broad mental health and wellbeing Anxiety Connections and friendship Loneliness and isolation Feeling less alone Feeling understood and validated Symptoms and function Depression Connections and friendship Self-efficacy Depression Coping Broad mental health and wellbeing Motivation Self-esteem Facebook groups Diabetes related forums Health related forums Depression Anxiety Broad mental health and wellbeing Anxiety Feelings of belonging Treatment decision-making Self-efficacy Optimism and hope Mixed methods: Cross-sectional interview ( n  = 20) Cross-sectional survey ( n  = 291) Connections and friendship Feeling less alone Symptoms and functioning Broad mental health and wellbeing Depression Anxiety Distress Quality of life Treatment decision-making Illness acceptance Feeling less alone Self-esteem Broad mental health and wellbeing Feelings of belonging Behaviour change Treatment decision-making Self-efficacy Broad mental health and wellbeing Feelings of belonging Loneliness and isolation Treatment decision-making Illness acceptance Feeling less alone Identity Coping Connections and friendship Loneliness and isolation Feeling understood and reassured Self-esteem Non-randomised control trial: Efficacy outcomes ( n  = 31) Focus group evaluation ( n  = 17) Symptoms and functioning Broad mental health and wellbeing Depression Anxiety Distress Quality of life Loneliness Self-efficacy Feelings of belonging Broad mental health and wellbeing Illness acceptance Quality of life Coping Asthma Village: an online community for patients with asthma where participants could leave comments or see who else was online There was no experimenter intervention Broad mental health and wellbeing Depression Anxiety Quality of life Coping Depression Loneliness Pre/Post quantitative survey ( n  = 10) Post-intervention qualitative survey and interview ( n  = 14) Connections and friendships Loneliness and isolation Self-efficacy Coping Pain Depression Anxiety Post-traumatic growth Coping Symptoms and functioning Depression Quality of life Symptoms and functioning Broad mental health and wellbeing Depression Broad social wellbeing Cross-sectional survey ( n  = 183 adults) Cross-sectional interview ( n  = 20) Cross-sectional survey ( n  = 178) 4 online communities: 1) TuDiabetes:sharing of blog posts, discussions, events, photos and videos on over 22,000 forums and 400 groups 2) Diabetic Connect: discussion boards and options to share and rate receipes and medication and updated new stories and Ask an Expert 3) Facebook 4) Twitter Symptoms and functioning Broad mental health and wellbeing Quality of life Connections and friendships Feelings of belonging Behaviour change Treatment decision-making Self-efficacy Empowerment Feeling less alone Feeling understood and reassured Symptoms and functioning Pain Broad mental health and wellbeing Depression Anxiety Quality of life Behaviour change Self-efficacy Empowerment Connections and friendship Empowerment: Optimism and hope Self-esteem Connections and friendship Loneliness and isolation Feeling less alone Feeling understood and reassured Coping Symptoms and functioning Broad mental health and wellbeing Feelings of belonging Connections and friendship Loneliness and isolation Self-efficacy Feeling less alone Symptoms and functioning Pain Broad mental health and wellbeing Broad social wellbeing Coping Symptoms and functioning Pain Broad mental health and wellbeing Depression Distress Quality of life Broad social wellbeing Loneliness and isolation Self-efficacy Optimism and hope Coping Self-efficacy Coping Depression Loneliness and isolation Illness acceptance Optimism and hope Broad mental health and wellbeing Anxiety Connections and friendships Loneliness and isolation Behaviour change Feeling understood and reassured Feeling less alone Optimism and hope Broad mental health and wellbeing Feelings of belonging Loneliness and isolation Behaviour change Feeling understood and reassured Self-esteem Optimism and hope Coping Broad mental health and wellbeing Distress Quality of life Connections and friendship Feeling less alone Identity Pain Depression Anxiety Depression Behaviour Motivation Adherence Self-efficacy Feelings of belonging Treatment decision-making Broad mental health and wellbeing Feeling understood and reassured Optimism and hope Anxiety Empowerment Illness acceptance Symptoms and functioning Broad mental health and Wellbeing Distress Self-efficacy Optimism and hope Broad mental health and wellbeing Self-efficacy Illness acceptance Optimism and hope Coping Depression Anxiety Symptoms and functioning Broad mental health and wellbeing Symptoms and functioning Broad mental health and wellbeing Self-efficacy Illness acceptance Symptoms and functioning Broad mental health and wellbeing Self-efficacy Pain Broad mental health and wellbeing Distress Loneliness and isolation Treatment decision-making Self-efficacy Empowerment Illness acceptance Feeling understood and reassured Coping Broad mental health and wellbeing Feelings of belonging Connections and friendship Loneliness and isolation Post intervention interview ( n  = 27) Pre/Post survey ( n  = 27) Feelings of belonging Loneliness and isolation Behaviour change Self-efficacy Feeling less alone Coping Broad mental health and wellbeing Feelings of belonging Connections and friendships Loneliness and isolation Behaviour change Treatment decision-making Self-efficacy Illness acceptance Feeling less alone Feeling understood and reassured Optimism and hope Self-esteem Coping Social wellbeing Illness acceptance Self-esteem Optimism and hope Broad social wellbeing Connections and friendships Loneliness and isolation Treatment decision-making Illness acceptance Optimism and hope Self-esteem Depression Treatment decision-making Optimism and hope Coping Depression Anxiety Treatment decision-making Coping Post-intervention interview ( n  = 20) Pre/post survey n  = 18 Anxiety Feelings of belonging Connections and friendships Isolation and loneliness Treatment decision making Empowerment Illness acceptance Optimism and hope Broad mental health and wellbeing Connection and friendship Behaviour change Illness acceptance Treatment adherence Identity Broad mental health and wellbeing Anxiety Feelings of belonging Loneliness and isolation Feeling less alone Identity Mixed methods: Cross-sectional interview = ( n = not reported) Cross-sectional survey n  = 349 Symptoms and functioning Broad mental health and wellbeing Anxiety Quality of life Feelings of belonging Connections and friendship Optimism and hope Identity Symptoms and functioning Broad mental health and wellbeing Quality of life Broad mental health and wellbeing Feelings of belonging Motivation Empowerment Optimism and hope Chronic Conditions identified in the review Most studies were conducted by authors based in the USA ( n  = 46), followed by the UK ( n  = 18), and the Netherlands ( n  = 7). 30 studies did not report participant location. Amongst those that did, most participants were based in the USA ( n  = 18), the Netherlands ( n  = 7), the UK ( n  = 9) and China ( n  = 5) or were international, but with a high proportion of participants in the USA ( n  = 7). These numbers were identified based on inclusion criteria or recruitment details (e.g., recruited via a specific hospital, university, or a location-specific support group). Studies were published between 2002 and 2024. The years with the largest number of published studies were 2021 ( n  = 11), 2022 ( n  = 10), 2023 ( n = 8) and 2024 ( n  = 8). Most studies recruited participants from either online support groups or through hospitals. Sample sizes of the included studies ranged from 6 to 1641 participants. The effects of online support groups were tested with a variety of methods with the most frequent being cross-sectional quantitative surveys ( n  = 43), cross-sectional interviews ( n  = 26) and quasi-experimental studies ( n  = 22). Experimental studies introduced participants to a new online support group, often created for the experiment whereas cross-sectional studies and longitudinal surveys were naturalistic as they typically assessed the impact of groups in which participants were already a member. Interventions lasted between 1 and 6 months, whilst the duration of support group membership, in cross-sectional studies, when reported, ranged between less than 1 week to 15 years with reported mean duration being between 12 and 31 months. Groups created for the purpose of the research were mostly moderated by the researchers, psychologists, healthcare professionals or patient organisations, whereas studies exploring naturalistic groups often did not report how the group was moderated. 70 papers looked at asynchronous groups (e.g., discussion forums, Facebook groups, WhatsApp groups, email lists), eight papers (of which seven were experimental) looked at synchronous groups (e.g., real time text-based chat groups, or video or teleconference calls), and two explored a combination of both synchronous and asynchronous groups. The MMAT checklist can be found in Supplementary Tables  8 – 12 . The authors of the MMAT recommend against calculating an overall score for each study, as it is not informative, and instead suggest describing the overall quality of the studies included within the review 36 . Overall, the quality of the studies was satisfactory. Most quantitative descriptive and qualitative papers used opportunistic sampling as participants were recruited via adverts posts in online support groups, so it was often not possible to identify non-response bias. Many authors acknowledged use of a self-selected sample, with participants potentially differing those who did not take part. Similarly, most studies did not discuss characteristics of the target population, so it is not possible to identify whether the samples are representative of other individuals with the chronic condition or representative of online support group members. Furthermore, many studies used standardised scales and statistical analyses, but these often differed for each health outcome thus making it difficult to compare across studies. For example, at least eight scales were used to measure self-efficacy. Furthermore, some papers only reported percentage agreements to health and wellbeing-related statements. Randomised and non-randomised (including longitudinal intervention and naturalistic studies) typically used standardised measures and accounted for confounders in their analysis (e.g., demographics or baseline scores). When participants did leave the study, some studies reported their reasons and statistical differences in baseline scores, but not all. The sections below present findings in relation to six health and wellbeing outcomes: physical health, mental health, quality of life, social wellbeing, behaviour and decision-making, and adjustment. Broad mental health and wellbeing was the most frequently explored outcome ( n  = 45), followed by self-efficacy ( n  = 22), and depression and loneliness and isolation ( n  = 21). Table  4 details the types of research method used for each outcome. Tables  5 – 7 present findings on how usage characteristics and group type may influence each health outcome and provide a summary of the findings. Table 4 Type and number of study designs measuring each main health outcome and mechanism Health outcome Total number of papers Mechanisms only Randomised controlled trial Quasi-experimental Cross-sectional quantitative Cross-sectional qualitative Physical health Symptoms and functioning 17 7 2 8 5 2 Pain 6 0 3 1 0 2 Mental health Broad mental health and wellbeing 45 10 3 9 15 22  Depression 21 6 5 10 4 2 Anxiety 18 0 3 3 3 9 Distress 6 0 1 2 2 1 Quality of life 13 3 2 5 5 1 Social wellbeing Broad social wellbeing 3 2 0 1 2 0 Feelings of belonging 16 0 0 2 3 11 Connections and friendship 18 0 0 3 5 10 Loneliness and isolation 21 0 1 5 3 12 Behaviour and decision-making Behaviour change 16 1 2 1 4 9 Motivation 6 0 0 2 1 3 Treatment adherence 5 1 1 1 3 2 Treatment decision-making 18 1 0 1 11 7 Self-efficacy 22 3 3 7 5 8 Empowerment 10 0 1 1 2 6 Adjustment Illness acceptance 16 5 0 2 7 7 Feeling less alone 17 0 0 1 4 12 Feeling understood and reassured 13 1 0 0 2 11 Optimism and hope 19 2 1 1 6 11 Self-esteem 12 0 0 0 7 5 Post-traumatic growth 2 0 1 1 0 0 Identity 7 0 0 0 1 6 Coping 18 4 1 6 6 5 Table 5 The role of usage characteristics on health outcomes Health outcome Level of engagement (e.g., active vs passive user) Intensity of usage (e.g., daily vs non-daily user) Membership length Physical health Symptoms and functioning Higher engagement and better symptoms and functioning 39 , 49 Higher engagement and reduced symptoms and functioning 28 , 40 No relationship between engagement level and symptoms and functioning 28 , 39 , 40 , 49 , 53 . Less intensity of use and better symptoms and functioning 148 No relationship between intensity of use and symptoms and functioning 49 , 148 Pain No relationship between engagement and pain 28 , 55 . No relationship between intensity of use and pain 148 . Mental Health Broad mental health and wellbeing No relationship between engagement level and broad mental health outcomes 28 , 49 , 61 , 62 , 86 . No relationship between intensity of use and broad mental health outcomes: 49 , 61 , 62 , 148 . Depression Higher engagement and increased depression scores 53 No relationship between engagement and depression 28 , 40 , 55 , 61 , 62 , 86 Higher intensity and better depression scores 123 No relationship between intensity of use and depression 61 , 62 , 91 Longer duration and better depression scores 123 , 146 No relationship between duration and depression 91 Anxiety No relationship between engagement and anxiety 55 No relationship between intensity of use and anxiety 91 Longer duration and better anxiety scores 146 . No relationship between duration of use and anxiety 91 Distress No relationship between engagement and distress 28 Quality of life Quality of life No relationship between engagement and quality of life 28 Longer duration of use and better quality of life 146 Social Wellbeing Social wellbeing Higher engagement and better social wellbeing 28 , 151 No relationship between engagement and social wellbeing 53 Higher intensity of use and poorer social wellbeing 148 Longer duration of use and better social wellbeing 102 Loneliness and isolation No relationship between engagement and loneliness 28 Higher intensity of use and reduced loneliness 123 , 147 Connections and friendships Higher engagement and better connection outcomes 49 No relationship between engagement and connection 40 Higher intensity of use and better outcome 49 Behaviour Behaviour change Higher engagement and better behavioural outcomes 49 No relationship between engagement and self-management 150 Higher intensity of use and better behavioural outcomes 49 Treatment decision-making Higher engagement and better treatment-decision making outcomes 49 , 119 , Higher intensity of use and better treatment decision-making outcomes 49 , 91 , 119 . No relationship between intensity and treatment decision-making 91 . Longer membership duration and better treatment confidence 102 No relationship between membership duration and treatment decision-making 91 . Self-efficacy Higher engagement and higher self-efficacy 39 , 101 No relationship between engagement and self-efficacy 28 Higher intensity of use and higher self-efficacy 101 Longer duration of use and better self-efficacy outcomes 146 Empowerment Higher engagement and better empowerment outcomes 49 Higher intensity of use and better empowerment outcomes 49 . Adjustment Illness acceptance No relationship between engagement and illness acceptance 28 , 151 No relationship between membership duration and illness acceptance 102 Feeling less alone Higher engagement and feeling less alone 49 Higher intensity of use and feeling less alone 49 Feeling understood Higher engagement and feeling more understood 49 . Higher intensity of use and feeling more understood 49 . Self-esteem No relationship between engagement and self-esteem 151 No relationship between duration of use and self-esteem 102 Optimism and hope No relationship between engagement and optimism and hope 28 , 151 No relationship between intensity of use and optimism and hope 91 , 127 Longer duration of use and better optimism scores 123 No relationship between duration of use and optimism and hope 91 , 102 , 127 Coping No relationship between engagement and coping 28 Higher engagement and better coping 28 . Higher intensity of use and better coping outcomes 148 . No relationship between intensity of use and coping 91 , 148 Longer membership duration and better coping outcomes 91 Table 6 The relationship between group type and health outcome Group type Finding Professionally moderated vs unmoderated groups In a longitudinal RCT, pain, depression, anxiety, and self-efficacy scores increased in both a researcher-moderated and unmoderated Facebook group 55 , 121 . Similarly, a longitudinal RCT found no effect of group type (professional vs peer-led) on depression scores 88 . In a cross-sectional survey, more than 85% of participants reported being happy that the online support group was peer-to-peer (i.e., run by regular people ‘like me’ who have also experienced the condition) rather than trained professionals (e.g., therapists, doctors or nurses) 70 . Similarly, another cross-sectional survey found that self-efficacy was associated with medical adherence in both healthcare professional groups (i.e., established by hospitals or healthcare professionals) and non-healthcare professional groups, but this association was stronger in non-healthcare professional groups 152 . Online vs face-to-face groups In one cross-sectional survey, there was no difference in depression or anxiety outcomes between those attending an online or face-to-face group, but those attending a face-to-face group reported more positive wellbeing and less distress 43 . Two cross-sectional surveys reported no differences in treatment decision-making between participants in an online or face-to-face group 43 , 92 . Furthermore, participants in another cross-sectional survey reported that they felt more comfortable in an online group, compared to face-to-face, as they knew no-one was looking at them when they shared their stories / feeling / problems or asking questions 70 . An interview study found that whilst some participants wanted to attend in person groups, they are often unable to attend due to the severity of their symptoms therefore find online groups can be advantageous 46 Local groups (groups with a focus on a local area, rather than groups with an international or non-specific geographical focus) Local groups provided opportunities for patients to connect and maintain a sense of belonging offline, as well as in the online groups 47 and some joined such groups with the intention of meeting in person one day 46 Table 7 Summary of findings Category Health and Wellbeing Outcome Summary of findings Physical health Symptoms and functioning Summary : there is mixed evidence for the impact of online support groups on symptoms and functioning. While some studies report improvements, others found no change. Mechanisms : informational support and sharing experiences may positively influence symptoms and functioning. Pain Summary : the evidence from cross-sectional and experimental studies suggests that online support groups may be able to help with group members’ pain. Mental health Broad mental health and wellbeing Summary: the effects of online support groups on mental health is mixed. While many studies report improvements in mental health, more report either no or negative effects. Mechanisms: influencing factors for positive wellbeing include receiving emotional support, having an outlet for feelings, sharing experiences, helping others and through improved social wellbeing. Influencing factors for negative wellbeing include negative social comparison, information overload, reading emotional information, unsuitable groups and lack of replies. Depression Summary: no study reported a negative effect of online support groups on depression, but it is unclear whether online support groups have a positive or neutral effect on depression. Mechanisms: receiving emotional support may positively influence depression scores, whilst negative social comparison can negatively influence it. Some studies also highlight the complexity of factors such as emotional expression. Anxiety Summary: there was mixed evidence regarding the impact of online support groups on anxiety with studies reporting, positive, negative and no effects. Mechanisms: anxiety can be mitigated through emotional and informational support which helps manage unfamiliar symptoms, but it can be heightened through reading horror stories. Distress Summary: findings are mixed with studies reporting positive, negative and no effects on distress. Mechanisms: distress may increase when posts are skewed to be sad or negative. Quality of life Quality of life Summary: no studies reported a negative effect of online support groups but the evidence is mixed relating to a positive or neutral effect on quality of life. Mechanisms: emotional support may enhance quality of life. Social wellbeing Broad social wellbeing Summary: one study looked at social wellbeing more broadly and found that 52% of participants reported improvements. Mechanisms: factors associated with enhanced social wellbeing include emotional support. Feelings of belonging Summary: evidence, from mostly qualitative data, suggests that group members feel a sense of belonging to the online support group. However, when it is hard to join in conversations or there is no response to message one study found that it can lead to people feeling like an outsider and sometimes leaving the group. Mechanisms: being part of a group with people living with the same condition, which helps people feel part of a group, have discussions and develop a shared identity. Connections and friendship Summary: many members of online support groups form new friendships and social connections, but this is not the case for everyone and can be difficult to do. Mechanisms: positive association between perceived credibility and competence of discussion on online communities and social capital within online groups. Loneliness and isolation Summary: the evidence on isolation and loneliness outcomes is mostly positive but there is also the potential for feelings of isolation to resurface when logging off from the groups. Mechanisms: being an active member and making new friends positively influences loneliness and isolation. Behaviour and decision-making Behaviour change Summary: the evidence suggests that online support groups can encourage positive behaviour change such as engaging in preventative behaviours, changing risky behaviours, purchasing assistive devices and trying other people’s dietary habits. Mechanisms: behaviour change may be facilitated by reading others’ experiences and through the shared advice and the perceived credibility of such discussions. Motivation Summary: studies suggest that participation in online support groups may increase motivation to make a positive lifestyle change or keep up with self-management. Mechanisms: motivation may be influenced by sharing experiences, seeing success stories, and receiving non-judgmental advice. Treatment adherence Summary: most studies suggest no changes in adherence after participation in an online support group Mechanisms: treatment adherence may be influenced by sharing experiences and discussing treatments and medication with others. Treatment decision-making Summary: findings suggests that online support groups may influence treatment decision making (e.g., changing initial treatment and assessing benefits and side-effects), although there is large variability in the proportion of participants reporting this.  Mechanisms: sharing experiences and information and receiving emotional support may influence treatment-decision making. Intensity of usage may also influence treatment decision making (i.e., being a daily user) but this depends on the measurement. Self-efficacy Summary : most studies reported positive effects of online support groups on self-efficacy, although there is large variation in the proportion of participants reporting improvements and some studies reported no changes over time, nor any added benefits compared to educational controls. Mechanisms : receiving informational and emotional support, helping others, religious expression and using positive emotion words has been found to be associated with self-efficacy. Empowerment Summary: most studies report improvements in empowerment. Mechanisms: feeling empowered may be related to the information shared and being part of a collective voice as it enables group members to feel in control. Adjustment Illness acceptance Summary: findings suggest that online support groups may help group members accept their illness and have a positive appraisal of their condition. Mechanisms: illness acceptance may be influenced by social support, comparison with others and findings others in a similar situation. Feeling less alone Summary: the evidence suggests that group members may feel less alone by being part of an online support group. Mechanisms: participants reported feeling less alone after seeing others having similar feelings, emotions and experiences and receiving emotional support. Feeling understood and reassured Summary: the evidence suggests that online support groups may help group members feel understood and reassured Mechanisms: participants felt understood because of the shared experience and such feelings were associated with perceived credibility of discussions. Optimism and hope Summary: the evidence, from mostly qualitative studies, suggests that members of an online support group feel optimistic and hopeful towards the future after using a group. Mechanisms: optimism and hope are influenced by reading success stories, positive comparison, receiving emotional support and finding positive meaning. Self-esteem Summary: most studies suggest that online support groups may enhance self-esteem Mechanisms: studies suggest that self-esteem is associated with emotional and social support. Post-traumatic growth Summary: the evidence suggests no change in post-traumatic growth. Identity Summary: the evidence suggests that participating in online support groups may help group members' personal and group identities as they rediscover their sense of self, return to a lost version of themselves, feel normal again and connect with others. Mechanisms: sharing humour, being part of a majority and illness acceptance may facilitate changes in identity. Coping Summary: the findings are mixed regarding the effect of online support groups on coping. While many studies report a positive influence, some report either a negative or no effect on coping. Mechanisms: Connecting with others who understanding, being accepted and receiving social support may positively influence coping. Type and number of study designs measuring each main health outcome and mechanism The role of usage characteristics on health outcomes Higher engagement and better symptoms and functioning 39 , 49 Higher engagement and reduced symptoms and functioning 28 , 40 No relationship between engagement level and symptoms and functioning 28 , 39 , 40 , 49 , 53 . Less intensity of use and better symptoms and functioning 148 No relationship between intensity of use and symptoms and functioning 49 , 148 Higher engagement and increased depression scores 53 No relationship between engagement and depression 28 , 40 , 55 , 61 , 62 , 86 Higher intensity and better depression scores 123 No relationship between intensity of use and depression 61 , 62 , 91 Longer duration and better depression scores 123 , 146 No relationship between duration and depression 91 Longer duration and better anxiety scores 146 . No relationship between duration of use and anxiety 91 Higher engagement and better social wellbeing 28 , 151 No relationship between engagement and social wellbeing 53 Higher engagement and better connection outcomes 49 No relationship between engagement and connection 40 Higher engagement and better behavioural outcomes 49 No relationship between engagement and self-management 150 Higher intensity of use and better treatment decision-making outcomes 49 , 91 , 119 . No relationship between intensity and treatment decision-making 91 . Longer membership duration and better treatment confidence 102 No relationship between membership duration and treatment decision-making 91 . Higher engagement and higher self-efficacy 39 , 101 No relationship between engagement and self-efficacy 28 Longer duration of use and better optimism scores 123 No relationship between duration of use and optimism and hope 91 , 102 , 127 No relationship between engagement and coping 28 Higher engagement and better coping 28 . Higher intensity of use and better coping outcomes 148 . No relationship between intensity of use and coping 91 , 148 The relationship between group type and health outcome In a longitudinal RCT, pain, depression, anxiety, and self-efficacy scores increased in both a researcher-moderated and unmoderated Facebook group 55 , 121 . Similarly, a longitudinal RCT found no effect of group type (professional vs peer-led) on depression scores 88 . In a cross-sectional survey, more than 85% of participants reported being happy that the online support group was peer-to-peer (i.e., run by regular people ‘like me’ who have also experienced the condition) rather than trained professionals (e.g., therapists, doctors or nurses) 70 . Similarly, another cross-sectional survey found that self-efficacy was associated with medical adherence in both healthcare professional groups (i.e., established by hospitals or healthcare professionals) and non-healthcare professional groups, but this association was stronger in non-healthcare professional groups 152 . Summary of findings Summary : there is mixed evidence for the impact of online support groups on symptoms and functioning. While some studies report improvements, others found no change. Mechanisms : informational support and sharing experiences may positively influence symptoms and functioning. Summary: the effects of online support groups on mental health is mixed. While many studies report improvements in mental health, more report either no or negative effects. Mechanisms: influencing factors for positive wellbeing include receiving emotional support, having an outlet for feelings, sharing experiences, helping others and through improved social wellbeing. Influencing factors for negative wellbeing include negative social comparison, information overload, reading emotional information, unsuitable groups and lack of replies. Summary: no study reported a negative effect of online support groups on depression, but it is unclear whether online support groups have a positive or neutral effect on depression. Mechanisms: receiving emotional support may positively influence depression scores, whilst negative social comparison can negatively influence it. Some studies also highlight the complexity of factors such as emotional expression. Summary: there was mixed evidence regarding the impact of online support groups on anxiety with studies reporting, positive, negative and no effects. Mechanisms: anxiety can be mitigated through emotional and informational support which helps manage unfamiliar symptoms, but it can be heightened through reading horror stories. Summary: findings are mixed with studies reporting positive, negative and no effects on distress. Mechanisms: distress may increase when posts are skewed to be sad or negative. Summary: no studies reported a negative effect of online support groups but the evidence is mixed relating to a positive or neutral effect on quality of life. Mechanisms: emotional support may enhance quality of life. Summary: one study looked at social wellbeing more broadly and found that 52% of participants reported improvements. Mechanisms: factors associated with enhanced social wellbeing include emotional support. Summary: evidence, from mostly qualitative data, suggests that group members feel a sense of belonging to the online support group. However, when it is hard to join in conversations or there is no response to message one study found that it can lead to people feeling like an outsider and sometimes leaving the group. Mechanisms: being part of a group with people living with the same condition, which helps people feel part of a group, have discussions and develop a shared identity. Summary: many members of online support groups form new friendships and social connections, but this is not the case for everyone and can be difficult to do. Mechanisms: positive association between perceived credibility and competence of discussion on online communities and social capital within online groups. Summary: the evidence on isolation and loneliness outcomes is mostly positive but there is also the potential for feelings of isolation to resurface when logging off from the groups. Mechanisms: being an active member and making new friends positively influences loneliness and isolation. Summary: the evidence suggests that online support groups can encourage positive behaviour change such as engaging in preventative behaviours, changing risky behaviours, purchasing assistive devices and trying other people’s dietary habits. Mechanisms: behaviour change may be facilitated by reading others’ experiences and through the shared advice and the perceived credibility of such discussions. Summary: studies suggest that participation in online support groups may increase motivation to make a positive lifestyle change or keep up with self-management. Mechanisms: motivation may be influenced by sharing experiences, seeing success stories, and receiving non-judgmental advice. Summary: most studies suggest no changes in adherence after participation in an online support group Mechanisms: treatment adherence may be influenced by sharing experiences and discussing treatments and medication with others. Summary: findings suggests that online support groups may influence treatment decision making (e.g., changing initial treatment and assessing benefits and side-effects), although there is large variability in the proportion of participants reporting this. Mechanisms: sharing experiences and information and receiving emotional support may influence treatment-decision making. Intensity of usage may also influence treatment decision making (i.e., being a daily user) but this depends on the measurement. Summary : most studies reported positive effects of online support groups on self-efficacy, although there is large variation in the proportion of participants reporting improvements and some studies reported no changes over time, nor any added benefits compared to educational controls. Mechanisms : receiving informational and emotional support, helping others, religious expression and using positive emotion words has been found to be associated with self-efficacy. Summary: most studies report improvements in empowerment. Mechanisms: feeling empowered may be related to the information shared and being part of a collective voice as it enables group members to feel in control. Summary: findings suggest that online support groups may help group members accept their illness and have a positive appraisal of their condition. Mechanisms: illness acceptance may be influenced by social support, comparison with others and findings others in a similar situation. Summary: the evidence suggests that group members may feel less alone by being part of an online support group. Mechanisms: participants reported feeling less alone after seeing others having similar feelings, emotions and experiences and receiving emotional support. Summary: the evidence suggests that online support groups may help group members feel understood and reassured Mechanisms: participants felt understood because of the shared experience and such feelings were associated with perceived credibility of discussions. Summary: the evidence, from mostly qualitative studies, suggests that members of an online support group feel optimistic and hopeful towards the future after using a group. Mechanisms: optimism and hope are influenced by reading success stories, positive comparison, receiving emotional support and finding positive meaning. Summary: most studies suggest that online support groups may enhance self-esteem Mechanisms: studies suggest that self-esteem is associated with emotional and social support. Summary: the evidence suggests that participating in online support groups may help group members' personal and group identities as they rediscover their sense of self, return to a lost version of themselves, feel normal again and connect with others. Mechanisms: sharing humour, being part of a majority and illness acceptance may facilitate changes in identity. Summary: the findings are mixed regarding the effect of online support groups on coping. While many studies report a positive influence, some report either a negative or no effect on coping. Mechanisms: Connecting with others who understanding, being accepted and receiving social support may positively influence coping. Physical health outcomes included symptoms and functioning, and pain. The two RCTs found no effect of online support groups on symptoms and functioning over time 37 or compared to website controls 37 , 38 . Similarly, longitudinal surveys found no effect of joining an asynchronous group on their health status 39 nor any differences between users and non-users on functional wellbeing 40 . However, in post-intervention interviews following a non-randomised control trial, 53% of participants reported that participating in an online support group contributed to a reduction in their symptoms 41 and 86% of (seven) participants agreed that the posts in a secret Facebook page were helpful in improving their recovery 42 . Furthermore, one cross-sectional survey found lower self-reported symptom scores and higher function scores in online support group members compared to members of a face-to-face support group 43 . A cross-sectional analysis of health records found that patients with diabetes from a closed Facebook group had lower blood sugar levels compared to those not in the Facebook group, but there were no differences in other health outcomes 44 . Additionally, two cross-sectional qualitative studies, reported improved symptoms, enhanced functional wellbeing and expedited recovery 45 , 46 , although this was not the case for all group members 46 Two cross-sectional surveys found that online social, emotional, and informational support was positively related to physical quality of life 47 , 48 . Similarly, two interview studies suggested that sharing experiences and information on the group was attributed to improved symptoms and functioning 45 , 46 . Additionally, in a cross-sectional survey, participants who reported that the online community helped them to learn strategies to improve insurance coverage were more likely to have increased blood sugar levels 49 . However, there were conflicting findings regarding the role of religious expression and insightful disclosure. Of five pre-post content analyses, within intervention and naturalistic settings, three reported that greater religious expression and insightful disclosure by participants were associated with improved self-reported functional wellbeing 50 – 52 , but two did not 53 , 54 . One also reported no association between disclosure of negative or positive emotions and functional wellbeing 52 . Another cross-sectional survey found no relationship between perceived competence of online discussions and diabetes related complications or blood sugar levels and that participants who reported that the online community helped them to learn strategies to improve insurance coverage were more likely to have increased blood sugar levels 49 . One RCT reported a reduction in pain severity and interference amongst members of moderated and unmoderated Facebook support groups 55 . However, two RCTs found no significant change in pain scores between the intervention (weekly moderated synchronous groups and a secret Facebook group plus education) compared to usual care or an educational control 38 , 56 . A pre-post intervention study reported positive outcomes on reactions to pain 57 and two cross-sectional qualitative studies reported that online support groups helped with pain reduction 26 , 58 . In particular, a participant reported that suggestions made on online support groups helped them to stay ahead of their pain, when previously they would have gone to hospital 26 . This section includes broad mental health and wellbeing, depression, anxiety, and distress. Here we consider measures of emotional benefits, emotional health, (psychological or emotional) wellbeing, negative feelings, difficult emotions, mental health and mood. Three RCTs reported mixed findings. One RCT reported improvements in mood scores across all participants after the intervention and at follow-up, but there were no differences between the intervention group (moderated weekly calls plus education) and the control (education only) 59 . However, another found that no differences in stress scores between a secret Facebook group plus education and an educational control 38 , and another RCT found that women in an unmoderated email group had poorer wellbeing at both 4 and 12 months than women using an educational website 37 . Similarly, anger increased over time for participants of a moderated weekly online group and was higher amongst the intervention group than the control group at the end of the study 60 . Moreover, a longitudinal survey reported no change in emotional wellbeing over time 61 . Interviews following a non-randomised controlled trial found that some participants were not as sad as before they joined a weekly professionally moderated group 41 . Furthermore, four cross-sectional quantitative studies also found no association between online support group participation and mental health 62 – 65 , whereas a further six reported a positive effect on mental health and wellbeing 66 – 71 . For example, in one cross-sectional survey, 100% of participants agreed that the online support group made a positive difference in their emotional health 70 and in another cross-sectional survey, 75% indicated that being involved with online support groups increased their satisfaction with daily life, 57.9% reported reduced sadness, and 27.6% expressed that involvement in the online support group had decreased thoughts of suicide 69 . However, one cross-sectional quantitative survey compared online to face-to-face support groups and found that more people attending a face-to-face group (two-thirds of participants) reported positive wellbeing than those attending an online support group (one-third of participants) 43 . 10 cross-sectional qualitative studies reported a positive effect on mental health and wellbeing 26 , 45 – 47 , 58 , 66 , 72 – 75 , but 12 reported reduced wellbeing, including feelings of frustration, fear, upset, sadness, overwhelm, guilt and disappointment 22 , 46 , 69 , 72 , 73 , 76 – 82 . Content analyses in a pre-post intervention reported that the use of a higher percentage of religious words predicted lower levels of self-reported negative emotions, but not emotional wellbeing 51 . Similarly, insightful disclosure in two intervention studies was predictive of lower levels of self-reported negative emotions and improved emotional wellbeing 52 , 54 . Furthermore, although a survey found that 85.8% of participants said that writing down thoughts and feelings made them feel better 70 , a content analysis in an intervention study found no association between disclosure of negative or positive emotions and emotional wellbeing 52 . Additionally, communicating about oneself within an online support group in an intervention study (measured through first-person pronoun use, e.g., ‘I’) was associated with higher levels of negative emotions, but communicating about others (measured through use of relational pronouns, e.g., ‘we’ or ‘you’) was not 83 . Two cross-sectional quantitative studies found that receiving online emotional support was positively associated with emotional wellbeing 84 and receiving online support was associated with psychological quality of life 48 . However, a further two found that giving and receiving social support, and receiving informational support, was not associated with mental health 84 , 85 . Online support network size also had an indirect positive effect on emotional wellbeing in a cross-sectional survey, through online received emotional support 84 . Furthermore, a cross-sectional survey found that pessimistic social comparison (e.g., fearing that future will be similar or feeling frustrated at own situation) negatively affected emotional wellbeing, whereas optimistic strategies (e.g., realising it is possible to improve or realising how well you are doing), did not negatively affect emotional wellbeing 86 . Six cross-sectional qualitative studies suggested that online support groups improved mental health as they provided a space to share experiences, receive and offer social support, have an outlet for feelings, and have opportunities to help others and learn new skills 26 , 45 – 47 , 66 , 73 . This was reported to be particularly important for those whose symptoms left them unable to carry out their usual purposeful activities 46 . Four also suggested that mental health improved due to improvements in social wellbeing and companionship 45 , 47 , 58 , 66 . On the other hand, seven suggested that poorer wellbeing, and increased fear, was influenced by exposure to negative aspects of conditions (including hospitalisation, relapses, suicidal thoughts, and death of other members), as well as complaints by other members and posts that were not solution-focused 22 , 46 , 73 , 79 – 82 . For some, positive stories were also damaging 69 . Qualitative studies also found that online support groups can focus too much on the condition 77 , can be overwhelming in terms of the information 78 , 80 , and can serve as a reminder for one’s own negative health 79 . One also found that some participants experienced personal attacks or ridicule for their views and opinions, which led to feelings of mistrust and fear 78 . Feelings of frustration and disappointment were also reported in three (cross-sectional and intervention) qualitative studies, if participants were unable find online groups suited to their unique needs 82 and when having to wait for reply 72 , 87 . An interview study with Long Covid patients also found that some participants reported feeling frustrated or resentful of group members’ who developed the condition after taking high-risk activities, such as travelling when it was advised not to 76 . Of five RCTs, one reported a reduction in depression over time in both an unmoderated and moderated Facebook group, with effects being sustained after one month 55 . However, three RCTs reported no differences in depression scores between online support groups (weekly moderated groups or a peer-led Facebook group – sometimes plus education), and control groups (education or usual care) 38 , 56 , 59 . Another longitudinal RCT reported no effect of time on depression in both a professionally moderated group and unmoderated group 88 . Furthermore, a pre-post intervention reported a reduction in depression following a 16-week intervention of weekly meetings combined with a private asynchronous newsgroup 57 but another pre-post survey found no difference in depression following a 6-month WhatsApp group intervention 89 . Furthermore, a non-randomised controlled trial found improvements in depression scores following weekly moderated sessions, but no differences post-intervention between the intervention and treatment as usual 60 . Similarly, two non-randomised control trials compared depression between participants in moderated weekly video groups and control groups (journalling or no treatment) and found no differences between conditions 41 , 90 , or over time 90 . Additionally, a longitudinal intervention reported no differences in depression scores between users and non-users at six weeks or 3 months 40 and a longitudinal survey also found no change in depression over time 61 . In one cross-sectional survey, 55% of participants reported improvements in depressed feelings 91 , but two cross-sectional surveys found no difference in depression scores between an online and face-to-face group 43 , 92 . However, a cross-sectional interview found that some participants feel more depressed after reading negative posts on a Facebook group 93 . In terms of content expressed, three analyses of posts made on online support groups within experimental and naturalistic settings reported no association between each of empathy expression 94 , religious expression 53 , or insightful disclosure 50 with depression. With regards to support and comparison, one cross-sectional survey reported that depression was negatively predicted by social support 95 , but another found that receiving and offering emotional support and receiving advice was not associated with depression 96 . With regards to social comparison, one cross-sectional survey found that upward contrast negatively predicted depression (but not downward identification, upward identification or downward contrast) 86 . In a cross-sectional survey, depression scores amongst passive users was not associated with conflict (e.g., feeling burdened or misunderstood) or universality (e.g., findings others with similar experiences), but conflict was positively associated with depression for active users 96 . Of three RCTs, one reported reductions in anxiety scores in both a moderated and unmoderated Facebook group, but these effects were only sustained 1-month post-intervention in the unmoderated group, not the moderated group 55 . However, two RCTs found no difference in anxiety scores over time between the online support groups (moderated synchronous text-based sessions or moderated Facebook group) plus education and an educational control 38 , 59 . Furthermore, qualitative and quantitative findings of quasi-experiments reported a reduction in anxiety in a moderated synchronous online support group compared to a no-treatment control group 41 and following an unmoderated email-based support group 97 . However, another quasi-experiment reported whilst anxiety scores improved over time, there were no differences between a moderated synchronous weekly chat group and treatment as usual 60 . Two cross-sectional surveys found no difference in anxiety scores between participants in an online or face-to-face support group 43 , 92 , with nearly 60% of participants in one study reporting improved anxious feelings [126]. Four cross-sectional qualitative studies reported a reduction in anxiety 47 , 78 , 91 , 98 . However, five qualitative studies reported the potential for online support groups to increase anxiety 72 , 76 , 77 , 82 , 93 , with this causing some individuals to limit their usage of the group 77 . A quantitative cross-sectional survey found that giving and receiving emotional support and receiving advice were negatively correlated with anxiety for active users, whereas receiving advice, and universality (e.g., finding others similar to you) was negatively correlated with anxiety for passive users 96 . Qualitative findings suggested that online support groups quelled anxiety as they helped to manage unfamiliar symptoms and provided emotional and informational support 47 , 78 , 97 , 98 . However, they also suggested that online support groups may increase anxiety after reading ‘horror stories’ and messages that can bring attention to specific issues that could be faced in the future 72 , 77 , 82 , 93 . Distress refers to distress from traumatic events and emotional distress more generally. An RCT compared an unmoderated email group with an educational website and found no difference in distress scores over time or between groups 37 . However, whilst one quasi-experiment found that distress significantly decreased over time amongst participants in a moderated synchronous online support group, there were no differences between the intervention and treatment as usual control 60 . Similarly, a non-randomised controlled trial found no differences in distress between a moderated synchronous group and a no-treatment control 41 . Furthermore, a cross-sectional survey found that 100% of participants reported that private email groups helped them deal with their emotional distress 70 . Another cross-sectional survey reported that distress was less frequent in a face-to-face group than an online support group 43 . Qualitative findings also suggested that seeing others’ stories can lead to increased distress 26 . Participants in a cross-sectional qualitative survey reported that distress increased when the posts are skewed to sad or negative 26 . This study also found that positive stories can be distressing, for example reading members’ pregnancy stories can be difficult for those with fertility issues 26 . The following studies refer to a broadly measured quality of life; where sub-scales of quality of life are reported (e.g., role functioning) these are reported in their respective section. One RCT and two quasi-experiments found no differences between an online support group (private Facebook group or moderated synchronous groups) and control (education, usual care, or no treatment) 38 , 41 , 56 . Another RCT found that whilst quality of life scores improved over time following a moderated synchronous group, there were no differences between the intervention and treatment as usual 60 . However, 100% of (seven) participants agreed that the posts in a secret Facebook page were helpful in improving their quality of life 42 . Similarly, a cross-sectional quantitative survey reported that 94.7% in a private email group said that the group made a positive difference to their quality of life 70  and another cross-sectional survey reported lower quality of life scores in a face-to-face group than an online support group 43 . Giving and receiving informational support was not associated with quality of life in a cross-sectional survey and a content analysis within an intervention study 99 , 100 , whereas perceived emotional support was, with this outcome being mediated by contentment 99 . However, in cross-sectional surveys, existential quality of life was associated with receiving online social and emotional support 48 , companionship 47 , and relatedness 47 , but not online informational support 47 . A content analysis within a longitudinal survey of new members of an existing asynchronous bulletin board found that insightful disclosure was not associated with quality of life scores 50 . Furthermore, a cross-sectional qualitative survey reported that their quality of life had improved through the support from group members 101 . A cross-sectional quantitative study found no association between perceived competence of discussions within an online support group and quality of life 49 . Social wellbeing outcomes include broad social wellbeing, feelings of belonging, connections and friendship, and loneliness and isolation. One cross-sectional study found that 52% of participants reported enhanced social wellbeing from being part of an online support group 102 . Two cross-sectional surveys suggested found that exchanging social support 23 , 102 and encountering emotional support 23 were positively associated with social wellbeing. However, there were conflicting findings for sharing experiences as whilst one cross-sectional study found that it was positively associated with social wellbeing 23 , another reported that it was not 102 . A cross-sectional survey and a quasi-experiment reported that enhanced social wellbeing was not predicted by use of religious expression 53 , information exchange 102 , helping others 102 or comparison with others 102 . A pre-post intervention survey had mixed results as although women in an unmoderated email group agreed that they felt a sense of belonging, some also reported leaving groups as they felt different from other members and did not feel close to the group 97 . Furthermore, in interviews following a non-randomised controlled trial, participants reported finding community in the weekly professionally moderated support group 41 . Similarly, 90% of participants in a cross-sectional survey reported a sense of belonging as a result of comments or posts from other members 69 and similar findings were reported in two more cross-sectional surveys 66 , 68 and all 11 cross-sectional qualitative studies 22 , 45 – 47 , 49 , 74 , 75 , 77 , 80 , 82 , 103 . However, two cross-sectional qualitative studies also reported that some group members do not feel a sense of belonging within the group 46 , 77 . Qualitative findings suggested that feelings of belonging arose from interactions with others and were attributed to the common ground amongst group members and to being part of a group of people living with the same condition, which helped group members to fit in, have discussions and develop a shared identity 45 , 46 , 75 , 80 , 82 . However, another interview study found that some participants felt like outsiders due to difficulties in joining conversations, receiving no, or unhelpful, responses, finding posts too negative or too positive, or feeling like their needs are not represented within the group 46 , 77 . The quantitative findings of an intervention study did not find an increase in the number of friendships following a combined synchronous and asynchronous online support 104 . However, in post-intervention interviews, participants reported experiencing improved relationships and being more confident in their ability to make and socialise with friends following the combined synchronous and asynchronous group 104 and amongst participants of an unmoderated email group 97 . Furthermore, a longitudinal intervention reported no differences in bonding scores between users and non-users at six weeks or 3 months 40 . Four cross-sectional surveys reported that between 44 and 66% of participants formed new friendships in asynchronous groups 23 , 49 , 105 , 106 and another found that 94.7% bonded with the other women in an email group 70 . This is echoed in all 10 cross-sectional qualitative studies, as participants reported developing true friendships and bonds and felt connected to others 22 , 46 , 47 , 76 , 78 , 80 , 81 , 101 , 107 , 108 . With regards to offline relationships, sometimes new social contacts replaced friendships lost because of their condition 22 , sometimes they supplemented existing offline friendships 22 , and other times they led to a decline in real-life relationships due to being over-reliant on online relationships and decreased attention to offline relationships 78 . Furthermore, participants in two interview studies reported difficulties forming new relationships 78 , 108 . Participants connected with others through similar diagnoses, symptoms, illness management issues, as well as personal characteristics such as sense of humour 46 , 108 . Participants felt connected to other members through the conveyed emotion, although some participants found this difficult due to the lack of body language and not being an active member 80 , 108 . Furthermore, some participants found it difficult to connect to those with different experiences, such as those who are newly diagnosed, have more severe disabilities, have less family support, or do not share the same political interests 108 . Furthermore, one cross-sectional qualitative study found that although group members felt sad when a fellow member passes away they also felt more connected to each other 81 . Another found a positive correlation between the perceived credibility and competence of discussion on online communities and social capital within online groups 49 . Loneliness and isolation refers to the feelings following the formation of friendship and connections and is distinguished from feeling less alone (included in the adjustment sections) following seeing others with similar experiences. Loneliness and isolation outcomes have been grouped together, despite the differences in definitions 109 , as the terms are used interchangeably within the included studies to refer to an absence or presence of social connections. An RCT compared a moderated synchronous group plus educational website to the website alone and found better loneliness scores in the online support group condition 59 . The quasi-experimental studies reported conflicting results; quantitative findings from a post-intervention study found reductions in loneliness scores after a 12-week synchronous chat session intervention 104 , which is echoed in post-intervention interviews of the same study as well as after a four-month unmoderated email group 97 . However, three quasi-experiments reported no effects online support groups (combined synchronous and synchronous groups or moderated synchronous group alone) on loneliness over time 90 , 103 or compared to either a no treatment, or active, control 41 , 90 . On the other hand nine cross-sectional qualitative studies 22 , 26 , 45 , 46 , 74 , 76 , 78 , 82 , 108 and three cross-sectional quantitate surveys reported reductions in isolation, with this being reported in 47-75% of participants 23 , 69 , 106 . However, two cross-sectional qualitative studies suggested that participants sometimes felt isolation within an online support group 80 and after logging off 72 . Qualitative studies suggest that reductions in isolation occurred by connecting with others, making new friends, feeling part of a group and becoming more outgoing 22 , 26 , 45 , 46 , 74 , 76 , 82 , 104 . This was often particularly needed as the physical constraints of chronic conditions make it difficult socialise 108 . However, participants can feel isolated in online support groups as they lack human touch and connection 80 or because they feel different from others, which can result in them leaving groups 97 . This includes behaviour change, motivation, treatment adherence, treatment decision-making, self-efficacy and empowerment. The quantitative findings of an RCT, comparing an online support group plus an educational website to an educational website alone, found no difference in behaviours relating to disease management or health promotion between the groups after the intervention 38 . However, the qualitative findings of another RCT suggested that participants tried new things and were more active after using the online support group 87 . Similarly a pre-post survey following a WhatsApp group intervention found improvements in behaviour 89 . Similarly, post-intervention interviews following a quasi-experiment suggested that participants learned tips to help with their day-to-day life (e.g., where to place an inhaler) 103 . Furthermore, three cross-sectional surveys reported mixed findings. One found higher scores for self-management of diabetes amongst participants not belonging to an online support group compared to online support group members 110 , whereas two reported improvements for those who had participated in groups. Specifically, one reported increased odds for lifestyle changes for those who had participated in online support groups in the previous year 111 , whilst another reported improvements in self-management and adopting a healthy lifestyle for those in a virtual online community 68 . Furthermore, nine cross-sectional qualitative studies suggested that upon joining an online support group, participants gained the skills for self-management of their condition and started taking better care of themselves (e.g., engaged in preventative activities, changed risky behaviours, purchased assistive devices, and tried other people’s dietary habits) 22 , 45 , 58 , 69 , 72 , 78 , 101 , 112 , 113 . Five qualitative studies suggested that behaviour change was possible after reading about the experiences of others and through sharing advice in online support groups 45 , 58 , 72 , 112 , 113 , and a quantitative survey found that credibility of discussion on online communities positively correlated with self-care 49 . Interviews following a quasi-experiment, including a moderated discussion forum plus education (compared to education alone), suggested that participants were motivated to keep up with self-management 87 , but a pre-post survey found no differences in motivation to adhere to HIV treatment following a WhatsApp group intervention 89 . A cross-sectional survey reported mixed findings on motivation outcomes, as it found an increase in motivation scores amongst participants with Type 2 diabetes but a decrease amongst those with Type 1 63 . Moreover, three cross-sectional qualitative studies reported an increase in motivation to change behaviour 45 , 75 , 112 . Post-intervention interviews suggested that participants were motivated to keep up with self-management after reading posts of other people who were still active despite their pain 87 . This was echoed in cross-sectional qualitative studies which reported that motivation was influenced by seeing other people make healthy lifestyle choices, sharing success stories and receiving non-judgmental personalised advice 45 , 75 , 78 , 112 . One RCT reported no effects of support group membership on medication and infection control adherence, within and between conditions 114 . This is supported by post-intervention interviews following a WhatsApp group intervention 89 and a cross-sectional survey which found that social networking support group membership was not related to self-reported infection control adherence 115 . However, improvements in medication adherence were reported by online support group users in an interview and Delphi study, in the same paper 45 . Two qualitative studies suggested that treatment and medication adherence was facilitated by observing similar patients’ health status, sharing (positive and negative) experiences and being able to discuss with others (e.g., tracking and side-effects) 45 , 116 . However, a cross-sectional survey found no relationship between perceived social support from online peers and reported medical adherence 115 . Treatment decision-making refers to group members’ ability to make decisions relating to their treatment, revising their initial treatment plan and feeling confident in their treatment. In interviews following an unmoderated email-based support group intervention, participants reported being more active in terms of their treatment 97 . Four cross-sectional surveys report that group members learn about existing (50-60%) and alternative (60%) treatments; received treatment advice (20%); and can feel more confident in their chosen treatment 68 , 92 , 102 , 117 , 118 . Six cross-sectional quantitative studies (including one Delphi study) reported that 25-80.5% of participants reported learning about new treatments, having their treatment requests influenced by an online support group, or choosing to change their initial treatment after participating in an online group 43 , 45 , 91 , 92 , 119 , 120 . When comparing to face-to-face support groups, two cross-sectional surveys found no differences in treatment decision-making outcomes between the groups 43 , 92 . Furthermore, seven cross-sectional qualitative studies reported feeling empowered in relation to treatment decision-making and feeling more confident in their treatment 22 , 26 , 45 , 58 , 74 , 77 , 112 . Four qualitative studies reported that support with treatment decision-making occurred through connecting with other group members and sharing experiences and information as it allowed members to assess the benefits and side-effects of treatment and identify best practice 22 , 26 , 45 , 97 . With regards to treatment confidence, two cross-sectional quantitative surveys reported that social comparison 102 and finding recognition 23 predicted treatment confidence. However, there was conflicting evidence regarding the role of receiving emotional support, as although one cross-sectional survey found that it predicted treatment confidence 23 , another did not 102 . These two surveys also reported that treatment confidence was not predicted by information exchange, helping others or sharing experiences 23 , 102 . Two RCTs reported improvements over time in moderated and unmoderated Facebook groups (sometimes plus education) 38 , 121 , although one reported no differences between participants in the online support group and educational control 38 . However, another RCT found that emotional self-efficacy declined amongst participants in an unmoderated email group 37 . A non-randomised controlled trial reported no differences in self-efficacy between a weekly professionally moderated support group and a no-treatment control 41 . Furthermore, a post-intervention survey found no difference over time following a 12-week synchronous online support group 104 , but a pre-post survey found improvements in adherence self-efficacy following a WhatsApp intervention 89 . Other post-intervention interviews reported improvements after an 8-week synchronous online support group supplemented with gamification communication 103 . A longitudinal survey, three cross-sectional surveys and a Delphi study reported improvements in self-efficacy amongst participants, but there was variation in the proportions of people reporting such an effect (19.1–88.5%) 39 , 45 , 68 , 71 , 102 . Eight cross-sectional qualitative studies also reported improvements in self-efficacy 22 , 26 , 45 , 46 , 49 , 77 , 101 , 112 . With regards to content expressed on online support groups, two content analyses within intervention studies and a cross-sectional survey found that writing a higher number of religious expressions 51 , using more positive emotion words 52 , receiving social support 122 and helping others 122 was associated with improved self-efficacy, but disclosing negative emotions was not 52 . Similarly, qualitative studies found that the information and support on online support groups enabled people to take an active role in managing their condition and feel like they can regain control over their personal lives 22 , 46 , 77 , 112 . An RCT comparing a peer-led Facebook group plus online education to education alone found no differences in empowerment at 3 or 6 months 38 . However, post-intervention interviews following an unmoderated email-based support group found that participants felt empowered following the intervention 97 . Furthermore, across two quantitative studies between 73-80.7% of participants reported that online support groups improved empowerment 45 , 49 . Six cross-sectional qualitative studies also suggested that participants feel more empowered by being part of an online support group 26 , 49 , 75 , 98 , 107 , 112 . Qualitative studies suggested that feeling empowered was mostly in relation to the information shared, which enabled group members to feel in control 26 , 75 , 97 , 107 . Participants also reported feeling empowered by helping others 49 , 112 and being part of a collective voice 98 . A quantitative study reported that empowerment was positively associated with perceived credibility of discussions on online communities and behaviours such as requesting or sharing informational and emotional support 49 . This section includes illness acceptance, feeling less alone, feeling understood and reassured, self-esteem, optimism and hope, post-traumatic growth, identity, and coping. Two cross-sectional surveys reported that approximately 30% of participants said that the online support group helped them find meaning in their experience 71 and improved acceptance of their condition 102 . However, another cross-sectional survey reported that face-to-face support group members accepted their illness better than those in online support groups 43 . Seven qualitative studies also found that online support groups helped group members to accept their illness 22 , 26 , 72 , 74 , 113 , view it more positively 72 , reappraise it as something that can be successfully managed 72 , overcome its uncertainty 98 , conceptualise the illness as chronic rather than terminal 97 , and allowed members to understand their condition as defined by the community 113 . Four qualitative studies suggested that illness acceptance was facilitated by emotional expression 74 , comparison with other group members (particularly those with more severe symptoms) 72 , 113 and finding others in a similar situation 26 . Two cross-sectional surveys and two content analyses within interventions found that illness acceptance and positive reframing were not associated with empathy reception 85 , receiving emotional/social support 23 , 85 , 102 , information exchange 23 , 102 , helping others 23 , 102 , finding recognition 23 , sharing experiences 23 , 102 or religious expression 51 . On the other hand, across three cross-sectional quantitative studies positive reframing and illness acceptance was positively associated exchanging social support 85 , 102 , empathy expression 85 and comparison with others 102 , 123 . Additionally, a cross-sectional survey found that those who were inhibited from making contributions to online support groups because they either felt a poor sense of community or had concerns about privacy and disclosure were less likely to feel they had found positive meaning from the online support groups 124 . Post-intervention interviews in one quasi-experiment reported that participants felt less alone following the intervention 103 . Four cross-sectional quantitative 43 , 49 , 70 , 105 and 12 cross-sectional qualitative studies 22 , 46 , 58 , 72 , 74 , 76 , 78 , 82 , 108 , 112 , 125 , 126 also reported that participants felt less alone. In the surveys, this occurred in 76-100% of participants. Participants in cross-sectional qualitative studies reported feeling less alone as they can connect with others 74 , receive support 105 , compare to other group members knowing that others have similar feelings, emotions and experiences 46 , 76 , 78 , 112 , 126 , and have shared understanding and empathy amongst group members 82 . Online support groups are particularly beneficial for connecting those with rare conditions, and helping them to feel less alone 108 . A cross-sectional survey also found that feeling less alone was also positively associated with perceived credibility of discussions on online communities and behaviours such as requesting or sharing informational and emotional support 49 . One cross-sectional survey reported that 15% of participants reported that they felt reassured in a moderated asynchronous online support group 106 . All (11) cross-sectional qualitative studies reported that online support groups enabled participants to feel understood and reassured 22 , 26 , 49 , 69 , 72 , 73 , 76 , 78 , 79 , 87 , 108 . Three of the qualitative studies reported that online support groups reassured group members that they were not ‘crazy’ and that their symptoms were not ‘psychosomatic’ 22 , 73 , 76 . Qualitative studies suggested that participants felt understood and reassured because of the shared experience 22 , 108 , peer support 79 and reading others’ experiences, particularly those who share similar symptoms 26 , 72 , 73 , 78 , 79 , 87 . A cross-sectional survey found that feeling understood was positively associated with perceived credibility of discussions on online communities and behaviours such as requesting or sharing informational and emotional support 49 . One RCT found a deterioration in hope after 4 months in an unmoderated email group, but no differences were found between the online support group and an educational website 37 . Conversely, post-intervention interviews following a quasi-experiment suggested that an unmoderated email-based support group increased hope 97 . Furthermore, four cross-sectional surveys reported increases in optimism and hope, reporting that between 19% and 75% of participants experienced improvements 69 , 71 , 91 , 102 . All (11) cross-sectional qualitative studies reported increases in optimism and hope 22 , 45 , 47 , 69 , 72 , 73 , 75 , 77 – 79 , 107 but two also suggested decreases in these outcomes 69 , 77 . 10 qualitative studies highlighted the importance of reading success stories and comparing to other group members, particularly those who have had the condition for longer and are improving 45 , 47 , 72 , 73 , 75 , 77 – 79 , 97 , 107 , with one qualitative study suggesting that other members serve as positive role models 22 . Cross-sectional surveys suggest that receiving emotional and social support, finding recognition, comparison with others, and positive meaning may predict optimism and hope 23 , 102 , 123 , 127 . However, one cross-sectional survey reported that exchanging support and sharing experiences did not predict optimism and hope 102 . There was conflicting evidence between three cross-sectional surveys regarding receiving information and helping others, as two found that these factors did predict optimism 102 , 123 , but another found that they did not 23 . Six cross-sectional quantitative studies reported that between 26% and 88.4% of participants experienced improvements in self-esteem and self-confidence 23 , 43 , 45 , 69 , 102 , 106 . However, another cross-sectional survey found no difference between those who use Facebook forums and those who do not 63 . Five qualitative cross-sectional studies also report enhanced self-esteem and self-confidence 22 , 45 , 72 , 107 , 128 . Two cross-sectional qualitative studies suggested that enhanced self-esteem was facilitated by receiving appreciation from other group members, through the gratification they felt from being active online, and from giving back to the group by sharing personal experiences 22 , 107 . Two cross-sectional quantitative surveys suggest that self-esteem was not associated with information exchange 23 , 102 , finding recognition 23 , comparison with other members 102 , helping others 23 , 102 or sharing experiences 23 , 102 , but may be predicted by encountering emotional support 23 and exchanging social support 102 . An RCT and quasi-experiment found no changes over time in post-traumatic growth amongst participants in a weekly synchronous online support group compared to usual care 56 and scores prior to the intervention 57 . One cross-sectional survey reported that 93.1% of participants said that group participation had helped them recover their sense of self 70 . Similarly, two cross-sectional qualitative studies reported that participants formed new identities through accepting the changes that come with their condition and by returning to a lost version of themselves 47 , 74 . Four qualitative studies also reported that participants felt “normal” again after participating in the online support group 58 , 82 , 112 , 116 . Across six qualitative studies, three reported that they felt “normal” again as their experiences were normalised 82 , they were part of a majority (vs being an outlier) 58 , and they shared gallows humour 112 . Two also reported that participants formed new identities through accepting the changes that come with the condition 74 and feeling connected to a group 47 . One RCT found better coping outcomes in an educational control compared to a 12-weekly moderated online support group during the intervention 59 . However, after the intervention, coping outcomes on one sub-scale (self-blame) were more favourable in the online support group condition. A pre-post intervention study found an increase in support-seeking coping following weekly synchronous groups supplemented with a gamification social setting, with these quantitative findings echoed in the qualitative evaluation 103 . However, in a similar study by the same research team, they found no differences in coping scores following a 12-week moderated synchronous online support group, but post-intervention interviews suggested that participants sought more support-seeking coping strategies after the intervention 104 . Furthermore, another pre-post study reported reduced coping following a combined synchronous and asynchronous online support group 57 and a non-randomised controlled trial found improvements in coping following weekly moderated sessions, but not differences post-intervention between the intervention and treatment as usual 60 . Four cross-sectional quantitative surveys reported that between 60% and 88.1% of participants found that the online support group helped them to cope with their condition 69 , 71 , 91 , 92 , although one study found that a higher proportion of participants reporting increased coping in face-to-face groups 92 . Five cross-sectional qualitative studies reported coping outcomes with all suggesting that online support groups help people cope with their condition 22 , 26 , 45 , 74 , 108 . One interview study reported that 82.7% of participants found that online interactions helped them learn how to cope with the social, physical and health consequences of the diseases 45 . However, another interview study reported that there are limitations in the extent to which the groups can help as participants recognise that the groups do not substitute the support from health professionals 108 . Qualitative studies suggested that coping was facilitated by connecting with other people who understand 26 , 108 , having individual differences accepted 74 , and receiving social support 45 . Across two cross-sectional quantitative studies and two content analyses within intervention studies, giving and receiving informational support, empathy reception, social support, and finding positive meaning were positively associated with adaptive coping 94 , 95 , 100 , 122 , whereas helping others and empathy expression were not 94 , 100 .

This systematic review sought to investigate whether health and wellbeing outcomes are influenced by participating in online support groups for chronic conditions. We also sought to identify the factors influencing such outcomes. Summarising the findings of 100 papers, health outcomes were categorised as physical health, mental health, quality of life, social wellbeing, behaviour and decision-making, and adjustment, which broadly aligns with outcomes from a recent umbrella review exploring other types of peer support for people with chronic conditions 129 . The sections below, organised by research question, summarise, and discuss the findings. The findings suggest that online support groups may positively influence pain, social wellbeing, adjustment and behaviour change and decision-making. By surrounding oneself with people with the same condition within an online support group participants reported feeling less alone and more understood, reassured and optimistic. Similarly, participants reported that online support groups helped them find meaning, feel “normal” and either re-discover their old sense of self or discover a new identity, which is particularly important as those with chronic conditions often experience a loss of personal identity 130 . In addition to changes in identity, people with chronic conditions reported experiencing a loss of social connections upon their diagnosis 131 . Cross-sectional quantitative and qualitative findings suggests that online support groups can bridge this gap as many group members reported feeling a sense of belonging, feeling less isolated and developing new friendships, which is in line with previous reviews 18 , 19 . However, it is important the group members also foster offline relationships as some may feel lonely when coming offline and others may focus on online connections as the expense of in-person connections. Furthermore, after reading others’ experiences or advice, participants reported being motivated to keep up with self-management, change their behaviour and adopt new behaviours (e.g., changing their diet or purchasing assistive devices). This is in line with previous reviews reporting that social networking sites, peer support and online support groups may be effective for changing health behaviours 16 , 19 , 132 . For treatment decision-making (including treatment confidence) and empowerment, participants reported that sharing personal experiences and information helped them assess the benefits and side-effects of treatment, which in turn, helped them make decisions about their treatment and feel empowered. However, the benefits of the decisions may be context-dependent and vary according to the revised treatment option and group preferences 133 . For example, if a particular group encourages behaviours or treatments that could be damaging this could have a negative effect, as has been highlighted in the eating disorder literature comparing pro-eating disorder groups to pro-recovery groups 134 . It is also important to be cautious of misinformation and anecdotal evidence that may occur within online support groups. For example, whilst a particular treatment may be successful for one individual this is not to say it will work in someone else. As a result, many recommend speaking to a healthcare professional or conducting your own research before making changes 46 . However, some participants reported leaving the groups if their needs were not met, or they did not feel close with other group members. This highlights the importance of exploring the available online support groups to identify the ones that align with one’s needs and values 46 . Quantitative, and experimental, findings were less likely to report changes in loneliness, friendships, or behaviour change, but this could be due to the research design whereby participants are aware of the short duration of the study or it could be that quantitative measures do not reflect the experienced behaviours and connections. As the cross-sectional studies for adjustment and social wellbeing are naturalistic and mostly qualitative, they provide an insight into the effects of online support groups used by participants in their day-to-day lives, but they cannot establish cause and effect, nor analyse outcomes over time. For symptoms and functioning, depression, coping, quality of life, treatment adherence and self-efficacy the findings had either a positive or no effect, suggesting that whilst online support groups are unlikely to worsen these health outcomes, they may not always improve them. This supports a previous review which found mixed effects of various types of peer support in care settings on physical health outcomes 135 and partially supports a review that found computer-mediated support being associated with less depression and greater quality of life 20 . As people with chronic conditions are more likely to develop depression 136 and the physical symptoms are a key component affecting patients’ day-to-day life 137 , alternative support from a healthcare professional should be sought. It is also important to note that, compared to adjustment and social wellbeing, studies measuring these four outcomes mostly used quantitative and (quasi-)experimental measures which may also explain the findings. For example, the measures used may not reflect the experiences of those with a chronic condition or the experimental nature of the study may not lead to changes in health outcomes, either due to moderation, style or study duration. There is also the potential for some health outcomes to worsen after engaging with an online support group. Whilst some studies reported either no change or improvements in anxiety and broad mental health and wellbeing, participants in a similar number of studies described increases in anxiety and feelings of frustration, sadness, and guilt. This may be, in part, due to the greater number of qualitative studies used to measure these outcomes, as they allow participants to provide greater insight into their experiences. However, it is also likely that online support groups can simultaneously help and hinder mental health and may be dependent on various factors, such as users’ mood when engaging with the groups, group content and external pressures 46 . As a result, participants should be aware of this potentially harmful effect and should be attentive to how they feel when using online support groups and take a break if they notice a deterioration. Health and wellbeing outcomes were influenced by giving and receiving support, and sharing experiences and social comparison, which supports the extant literature 16 , 138 . These findings also partially map onto the SCENA model with regards to connection, exploration and narration 25 . This review distinguishes between the differential role of informational and emotional support for health. Indeed, whilst informational support may aid physical health, adaptive coping and behaviour change and decision-making, emotional support may improve wellbeing, anxiety, illness acceptance and make individuals feel less alone, particularly in the absence of other care. The findings also suggest that giving emotional support may aid positive re-framing whereas receiving support may help with depression. However, whilst this supports a previous review of mechanisms of different types of peer support, which reported that helping others enabled peers to find meaning in their own chronic condition 129 , multiple studies in this review reported that helping others was often not associated with outcome measures, including self-esteem, coping, optimism and social wellbeing. This may be due to the outcome measure or it could be attributed to the duration of participants’ illness as they may be more likely to benefit from helping others at a later stage of their illness journey 46 . Qualitative studies suggesting that reading others’ experiences can also positively influence physical health, adjustment, and behaviour and decision-making. However, quantitative measures did not find any effect on optimism and hope, potentially due to the way in which these measures were operationalised. Furthermore, if group members engage in positive comparison strategies whilst reading these experiences, it may help them feel less alone, normalise the condition, view their condition more positively, support meaning making, and put experiences put into ‘proportion’ 75 , 102 . This provides evidence for social comparison theory, which suggests that in order to evaluate oneself, people often compare to others 139 . This is typically done under uncertainty 139 , which is often the case for people experiencing a chronic condition, particularly novel or under-researched conditions such as Long Covid. However, reading others’ experiences can also negatively influence broad mental health, anxiety and distress, particularly when posts are negatively oriented or include worse symptoms or experiences, as readers feel upset or guilty. Similarly, if group members engage in negative comparison strategies (e.g., feeling frustrated at others doing better or anxious of people being worse) then they may have a negative effect on mental health, so it is important that individuals draw inspiration from other group members rather than dwelling on negative aspects of comparison. Therefore, it is important for group members to be aware of these potential negative outcomes when choosing which posts to engage with. Health outcomes may also differ depending on usage (e.g., level of engagement, intensity of use and membership duration) and group characteristics (e.g., moderated vs unmoderated and synchronous vs asynchronous). For most health outcomes, the included studies suggest that they are not impacted by the extent, or intensity, to which group members engage, but for feeling less alone, feeling more understood, and enhanced social wellbeing it may be beneficial to engage more actively and frequently. However as most of the studies were cross-sectional, it may be that individuals who already have these positive outcomes engage more with the groups. Also, a limited number of studies explored each outcome and characteristic, often with conflicting findings or different definitions and measurements, which makes it difficult to identify the optimal level of interaction with online support groups. It is argued that different group features may afford different benefits and may depend on individual preferences 27 . Most of the studies in this review explored asynchronous groups, such as discussion forums, Facebook groups, or email lists, although one cross-sectional study found that video-based groups foster social wellbeing, compared to large text-based groups which aid informational support 46 . Synchronous groups were only explored in seven papers, most of which were (quasi)experimental, and the quantitative findings of these experiments mostly reported no effects, compared to control groups, on health outcomes. However, it is not possible to establish whether this was due to the design of the support group or the experimental, or quantitative, nature of the study, particularly as the quantitative findings of (quasi)experimental studies with asynchronous groups mostly reported similar results and the qualitative findings of both synchronous and asynchronous groups were more nuanced. Future research should explore this more rigorously. Furthermore, studies comparing face-to-face and online groups found that both groups have similar influences on the health outcomes of group members. Two papers (with the same participants) compared unmoderated and moderated groups, and another two compared face-to-face and online groups, and found similar improvements in both groups. This is in line with a study which found no differences in depressive symptoms between participants allocated to a moderated or peer-led online support group 88 . However, it is not possible to generalise to other moderated groups, as groups can be moderated by researchers, peers or psychologists and can vary in activity from approving posts to actively guiding the conversation. Living with a chronic condition can have various consequences on health and wellbeing, with many turning to online support groups to support these health outcomes. This review can be used by clinicians, online support group administrators and those with a chronic condition to optimise their experience of using online support groups. The following recommendations can be made based on this review: (i) As many health outcomes were not affected by level and intensity of engagement, group members can engage with the groups at their own pace without harming their health; (ii) Online support groups may be able to bridge the decline in offline relationships that can occur with the diagnosis of a chronic condition, but it is important to not do this at the expense of offline relationships; (iii) If group members are looking to make a behavioural change or find support with treatment decision-making, they may benefit from informational support, but should also conduct their own research or speak to a healthcare professional; (iv) If individuals do not know anyone else with their condition, seeking emotional support from an online support group may help them feel less alone and more understood; (v) Learning of others’ experiences, particularly those who are successfully managing the condition, can support illness acceptance and feeling ‘normal’, particularly for conditions with increased uncertainty; and vi) Individuals should be aware that online support groups have the potential to increase distress, anxiety and negative emotions, so it is important that they avoid negatively oriented posts and negative comparison strategies and take a break from groups if their mental health begins to decline. Whilst considering these recommendations, it is important to be aware of the limitations of this review and the included studies. It is also important to consider individual differences that may also affect experiences with online support groups. Limitations of the studies included and of the review itself should be acknowledged. First, the quality of the studies was satisfactory. Most studies were cross-sectional, and survey findings were particularly descriptive, therefore it is not possible to identify a causal relationship between use of the online support groups and health and wellbeing. Also, some naturalistic studies did not include descriptions of the groups used, or whether participants were members of multiple groups thus making it difficult to extrapolate the findings. Indeed, it is possible that members used multiple online support groups, either for the same condition or to support them with multiple conditions, and that group values or content varied between groups, with each group potentially influencing health and wellbeing differently [115]. Most of the groups were also asynchronous, so it is not necessarily possible to extrapolate to synchronous groups, particularly video-based groups. The majority of included studies are also susceptible to selection bias, therefore it is possible that the samples do not reflect the wider population of either the online support groups or the chronic condition. When researching existing online support groups, researchers should endeavour to report as much detail as possible, such as whether members use multiple groups, their engagement level, and group features. Moreover, many of the studies also included mostly White and married participants, so these findings may not extrapolate to other demographics. This is important as chronic conditions may be more prevalent in deprived groups 140 and there may be different support needs between married and single participants 141 . When considering the review itself, it is possible that some studies were not identified within the search. There are also many offline factors, such as offline support and symptom severity, that may also underly any effects of online support groups on health and wellbeing 67 , which were beyond the scope of this review. Finally, most studies were conducted in populations with cancer, which may skew the findings as there are considerable differences between the available formal support for cancer patients compared to conditions such as Myalgic Encephalomyelitis/Chronic fatigue syndrome or Long Covid 142 . There is scope for further research, particularly regarding the effects of different group features, such as group size, composition, platform, duration and moderators 143 . Future research should compare different levels of these features to identify the optimal set-up of these features (e.g., video- or text-based) and the most suitable type of moderator (e.g., peer or healthcare professional). Similarly, there were limited, and sometimes conflicting, findings for usage characteristics so it is important for studies to formally define active and passive users and further explore how this influences health outcomes. As most studies included in this review were cross-sectional, future research should also consider a longitudinal design to see if such effects were sustained over time and to identify possible spill-over effects changes. This review synthesised findings on 25 health outcomes on the effects of online support groups for people with chronic conditions and suggests that online support groups broadly have a positive effect on social wellbeing (e.g., feeling connected to others and less isolated), behaviour (e.g., adopting positive behaviours), and adjustment (e.g., illness acceptance, identity, and feeling understood). For physical health, the findings suggest a positive influence on pain but a mixed result for symptoms and functioning. In terms of mental health, online support groups may have a positive or negative impact on outcomes, such as anxiety and emotional or psychological wellbeing, and this will depend on group content and comparison strategies.

Chronic conditions refer to health problems that require ongoing management over a period of years that cannot currently be cured, but can be controlled 1 . Almost half of the UK population reported living with at least one long-standing health problem in 2020 2 , and globally 41 million people per year are estimated to die from a chronic condition 3 . Although more recent data on the prevalence of chronic conditions is unavailable, it is likely to have increased since the COVID-19 pandemic, with nearly 2 million people reporting symptoms of Long Covid in England and Scotland in March 2023 4 . Living with a chronic condition is associated with reduced health-related quality of life 5 , 6 and leaves many individuals unable to carry out day-to-day activities, socialise or work, which can result in them being dependent on other people 7 , 8 . Alongside experiencing symptoms of a chronic condition, individuals may face challenges such as prejudice 9 , stigma 10 and feeling alone 8 . One way in which individuals can connect with others, and find support, is through online support groups. Online support groups, also referred to as ‘online communities’, ‘online support forums’, and ‘virtual support groups’, are “online services with features that enable members to communicate with each other” 11 ; they have an underlying premise that peers offer meaningful support due to the shared experience of a particular life event 12 . They may be created, and moderated, by peers (i.e., those with a direct lived experience of the condition), caregivers, charities, or health professionals. The growing need for online support groups is showcased by the large membership of many groups. For example, at the time of writing a diabetes Facebook group has reached 102,000 members in four years, with 202 new members in the last week, and 202 posts per month 13 , and a Long Covid support group has reached 66,000 members, with 96 members in the last week and 2000 posts in the last month 14 . An advantage of these online support groups, as opposed to in-person groups, is that they can transcend geographical boundaries and are less restricted by time or location, which is particularly beneficial to those with limited mobility and those living in rural communities 15 . Such groups can be synchronous via audio or video calls, or they can be asynchronous via social media platforms, such as Facebook groups and discussion boards, or via direct messages, such as in WhatsApp groups 15 . Previous reviews, exploring experiences of online support groups for specific chronic conditions, such as HIV 16 and cancer 17 , report that they are a place where group members can receive social support and experience a sense of community, which can result in increased adaptive coping and reduced loneliness. However, they also report that group content can be negative (e.g., distressing personal information or complaints), and that lack of replies and absence of nonverbal communication can lead to misunderstandings and distress. Previous reviews have also explored online support groups for multiple chronic conditions, including how online groups influence daily life 18 and illness self-management 19 . However, these reviews excluded quantitative studies, such as intervention studies, which could provide strong evidence for the impact of online support groups on group member experiences. A meta-analysis exploring health outcomes in relation to online support groups for health conditions did include intervention studies, but only those with a fixed start and end point and included an educational component, which is not representative of existing online support groups 20 . The outcomes included were also limited to social support, depression, quality of life and self-efficacy. Thus, there is a gap in the literature regarding a systematic review on the effects on health and wellbeing of using an online support group which includes both qualitative and quantitative studies and covers a greater number of health and wellbeing outcomes. In addition to understanding the health impacts of online support groups, it is also important to consider how these effects occur. Previous reviews highlight the importance of finding and exchanging information, receiving emotional support, and sharing experiences 21 – 24 . Furthermore, the SCENA Model of Therapeutic Affordances of Social Media 25 has also been applied to online support groups 26 , and suggests that such groups may afford self-presentation ( managing how one presents themselves online) , connection ( connecting with, and supporting, others ), exploration ( seeking information and improving knowledge) , narration ( exchanging experiences) and adaptation ( adapting self-management needs in relation to health status) . Due to the variety of platforms used for online support groups (e.g., video- or text-based), as well as the different ways of engaging with the groups (i.e., being a passive or active member of the group), it is important to consider how these factors also influence the health benefits afforded by online support groups 27 . For example, exploration may be easier in text-based groups where there is an archive of information. Previous studies have also explored the role of engagement 28 and group features (e.g., group size, duration, nature of communication) 20 . For example, more social support was reported when online support groups were of a longer duration and included both synchronous and asynchronous channels 20 . Fewer studies have compared asynchronous and synchronous platforms. Furthermore, there is not a review looking at the potential mechanisms underlying each type of health outcome and synthesising group and usage characteristics as well as support group content in the context of online support groups for chronic conditions. As the number of individuals experiencing, and having their lives disrupted by, chronic conditions increase, it is important to explore potential ways to improve health outcomes. One such way is online support groups. Therefore, it is important to understand the impact of these groups on the health and wellbeing of group members and to identify any influencing factors. This systematic review aims to explore this with the following research questions: What are the effects of online support groups on the observed and self-reported health and wellbeing of individuals with a chronic condition? What are the mechanisms by which online support groups affect the health and wellbeing of individuals with a chronic condition? What are the effects of online support groups on the observed and self-reported health and wellbeing of individuals with a chronic condition? What are the mechanisms by which online support groups affect the health and wellbeing of individuals with a chronic condition?

Transparent Peer Review file Supplementary Tables Reporting summary Transparent Peer Review file Supplementary Tables Reporting summary