Identifying risk and preventing progression to Type 2 diabetes in vulnerable and disadvantaged adults: a pragmatic review

Authors


Correspondence to: Jayne Taylor. E-mail: jayne.taylor8@nhs.net

Abstract

Aim

To identify effective approaches to recognize diabetes risk and prevent progression to Type 2 diabetes in vulnerable groups, whose diabetes risk may be difficult to identify or manage.

Methods

UK-based interventions that assess diabetes risk and/or target known risk factors were identified through four main sources: submissions to two calls for evidence by the National Institute for Health and Clinical Excellence; local practice examples collected via a targeted email questionnaire; selected electronic databases; and a focused search of relevant websites. No restriction was placed on the study type or evaluation methods used. Key themes and sub-themes on outcomes, as well as facilitators and barriers to successful delivery, are reported.

Results

Twenty-four interventions met all inclusion criteria: 15 included a risk identification element and 14 included preventative activities. A range of risk identification tools were used to improve diagnosis of unmet diabetes-related health needs and raise awareness of diabetes risk factors. All preventative interventions focused on lifestyle change. No interventions monitored blood glucose as an outcome and only one reported improvements in baseline risk scores. Facilitators included tailored and flexible programme design, outreach delivery in familiar locations and effective inter-agency working. Barriers included literacy and language difficulties, transient participant populations, low prioritization of diabetes prevention and cost.

Conclusions

It is possible to engage successfully with high-risk adults in vulnerable groups to achieve positive health outcomes relevant to the prevention of diabetes. However, more robust evidence on longer-term outcomes is required to ensure that programmes are targeted and delivered appropriately.

Introduction

There is an established body of evidence describing the successful components of interventions to recognize diabetes risk and prevent progression to Type 2 diabetes in adults identified to be at high risk of developing the disease. Clinical trials have shown that the risk of progression to diabetes can be reduced by almost 60% through lifestyle interventions that target weight loss, increasing levels of physical activity and dietary improvements [1, 2]. These findings have subsequently been confirmed in pragmatic implementation trials in real-life settings [3-5]. However, little is known about the effectiveness of such interventions for adults whose increased risk of developing Type 2 diabetes is more challenging to identify or manage. This group of people come from a wide range of disadvantaged and vulnerable backgrounds; for example, those living in poverty and socially excluded communities (e.g. prisoners, travellers, homeless people), as well as those with learning disabilities or severe mental illness. Members of certain ethnic or faith groups also face a raised diabetes risk that may be difficult to recognize or problematic to treat in practice.

The elevated risk of developing Type 2 diabetes in disadvantaged and vulnerable adults is well documented. For example, people in the lowest socio-economic groups are 2.5 times as likely, and black and minority ethnic groups up to six times as likely, to develop diabetes compared with the general population. Similarly, the prevalence of diabetes is up to five times higher among prisoners and 2–3 times higher in people with a severe mental illness, compared with the general population [6]. Much of this increased risk is attributed to lifestyle factors, including obesity, physical inactivity and an unhealthy diet, all of which are more common in deprived communities and those living in vulnerable circumstances. Many within these groups also face a range of practical and cultural barriers to accessing health care, preventative services and health checks or screening designed to identify relevant risk factors [6].

This article describes the results of a descriptive review of interventions to identify and reduce the risk of progression to diabetes that are targeted specifically at adults from vulnerable and disadvantaged groups. The review takes a pragmatic approach in an attempt to synthesize the current evidence base. The review was commissioned by the National Institute for Health and Clinical Excellence (NICE) to support the development of public health guidance on preventing progression to Type 2 diabetes in individuals at high risk [7].

Methods

Inclusion criteria

The review sought to identify UK-based evidence on effective and cost-effective approaches to:

  1. identify and monitor adults with impaired glucose regulation or who are at high risk of developing Type 2 diabetes (impaired glucose regulation signifies impaired glucose tolerance or impaired fasting glycaemia, or both);
  2. prevent or delay the progression to Type 2 diabetes in adults at high risk, through lifestyle, pharmacological or surgical interventions.

The review focused on interventions that target adults in disadvantaged communities and vulnerable groups, including those listed in Table 1. Because of the paucity of published evidence on Type 2 diabetes identification and prevention in these groups, the scope of this review was widened beyond interventions that specifically measure and monitor impaired glucose regulation, to include interventions which target known risk factors for Type 2 diabetes, including obesity and sedentary lifestyle, as well as clinical risk markers, such as hypertension and history of cardiovascular disease. Specific inclusion criteria are described in Table 1. Only those interventions which have been evaluated, or for which outcome measures were available (e.g. screening uptake, attendance, health needs identified, behavioural and health outcomes), were considered for inclusion, but no restriction was placed on the type of study or evaluation methods used.

Table 1. Inclusion criteria
  1. a

    Impaired glucose regulation signifies impaired glucose tolerance or impaired fasting glycaemia, or both, indicating that a person has blood glucose levels above normal but not high enough for a diagnosis of Type 2 diabetes.

UK-based projects or interventions that meet ALL of the following five criteria:
1. Interventions designed to identify and monitor adults with impaired glucose regulationa
and/or interventions which contribute to the prevention or delay of progression to Type 2 diabetes, including pharmacological and surgical interventions, plus lifestyle interventions which address one or more of the following:
1.1 Physical activity—reduce sedentary behaviour and/or increase physical activity levels
1.2 Weight loss—achieve/maintain a healthy body weight, BMI or waist circumference
1.3 Diet—improve dietary intake; for example, through diets that lower or control glycaemic index, (saturated) fat intake, carbohydrate intake, total calorie intake
2. Targeted at individuals or groups of individuals
3. For adults aged 18 years and above with:
3.1 impaired glucose regulation
or
3.2 characteristics that put them at high risk of developing diabetes
4. Focus on vulnerable groups whose diabetes risk may be missed or difficult to manage. These groups may include, but are not restricted to, the following:
—frail older people—travellers
—adults with a physical disability—refugees, asylum seekers and recent migrants
—people with severe mental illness (e.g. schizophrenia, bipolar disorder, on antipsychotic medication) or learning disabilities—homeless people
—some minority ethnic or cultural groups
—those not registered with a general practitioner—some faith communities
—prisoners—those living in poverty
5. Formal evaluation completed or evidence of outcomes recorded. Evidence may be quantitative or qualitative, published (since 2000) or unpublished

Search strategy

Evidence on relevant interventions was identified via four main sources:

  1. Analysis of submissions received in response to two calls for evidence made by NICE to inform the development of two separate pieces of public health guidance on Type 2 diabetes prevention [7, 8].
  2. Local practice examples collected via a targeted email questionnaire sent to 193 key informants, including registered stakeholders of the aforementioned NICE guidance, as well as relevant health and public health networks [including JISCMail Minority Ethnic Health, Health Equity and Public Health networks, plus the UK Health and Learning Disability Network, National Health Service (NHS) Health Check learning network and local diabetes research networks]. A copy of the questionnaire sent to key informants is available from the authors on request; evaluation reports submitted in response to this targeted email were also accepted.
  3. A search of selected electronic databases, using search terms describing one or more vulnerable group, plus ‘diabetes’ and Type 2 diabetes risk factor terms. The list of electronic databases consulted for this review is provided in the Supporting Information (Appendix S1), along with broad search terms used.
  4. A focused search of websites relevant to UK-based diabetes prevention programmes (see also Supporting Information, Appendix S1, which provides a list of the websites searched).

Individual projects and authors of relevant studies identified via electronic databases and the focused web search were contacted directly for further information, where appropriate and possible.

Recent literature reviews relevant to one or more vulnerable group, including UK- and non-UK-based studies, were also sought via the various search routes described.

Analysis

All available information on the identified studies, projects and programmes (collectively referred to as ‘interventions’) was analysed descriptively to uncover key themes and sub-themes in relation to outcomes, as well as facilitators and barriers to successful delivery. This information was supplemented with relevant contextual and supporting evidence from the literature reviews identified by the search to produce a descriptive account of the available evidence. The heterogeneity of the reports precluded a meta-analysis.

Searching and screening activity for the review was carried out between March and May 2011. Analysis and reporting was completed between June and September 2011. All tasks were undertaken by one of the authors (JT), with input and advice provided by NICE's public health technical team (including HC and CW) and an expert subgroup of NICE Programme Development Group committee members (including CC, JH, RH and RIGH). A full review report was presented to the Programme Development Group and used to inform the development of NICE public health guidance. However, the views expressed in this article represent those of the authors only and should not be taken as a reflection of NICE guidance or of NICE.

Results

Overview of included projects, interventions and studies

The approach taken to identifying and retrieving information on relevant projects and interventions was inclusive, but pragmatic. The reported findings are therefore indicative, rather than an exhaustive account of all relevant UK-based interventions.

Thirty-one interventions or papers were identified from the call for evidence submissions and the targeted email questionnaire. Ten of these were excluded, five on the basis of the target population: three studies did not target a vulnerable or disadvantaged group; one was targeted at adults with Type 2 diabetes; and another failed to target adults at high risk of developing diabetes. Three of the excluded papers were either commentaries or perspectives studies and did not describe interventions to identify diabetes risk or prevent progression to disease. Two further studies were in progress at the time of the review and no outcome data were available.

A further four unique interventions or studies were identified via electronic databases and relevant websites, three of which were included in the review (the excluded study was in progress and no outcome data were available at the time of the review).

Eight literature reviews were also identified and are described in Table 2.

Table 2. Included literature/systematic reviews
StudyReview question, inclusion/exclusion criteriaMethods
Obesity interventions for people with a learning disability [9]

Effectiveness and experiences of non-surgical, non-pharmacological interventions to promote weight loss in obese adults (age ≥ 16 years) with a learning disability

Quantitative and qualitative study designs

English language studies published 1998–2009

Integrative literature review

Databases searched: CINAHL, Proquest, MEDLINE (PubMed), PsycINFO databases, Cochrane Library. Plus hand searching of references and key journals' studies

Twelve papers selected for inclusion following screening

Health checks for people with learning disabilities [10]

Effectiveness of health checks for people with learning disabilities

Included studies: peer reviewed, all study designs, any outcome

English language studies published 1989–2010

Systematic review and narrative synthesis

Databases searched: MEDLINE, CINAHL, Web of Science, PsycINFO, plus in press/grey literature search via a specialist health research group

Thirty-eight studies (≥ 5000 subjects) selected for inclusion following screening (three randomized controlled trials, one non-randomized matched control study, one pooled analysis)

Healthy weight for adults with learning disabilities

(multi-method study, including rapid evidence review plus qualitative research with carers, commissioners and healthcare staff) [11]

Descriptive rapid evidence review of interventions that aim to support overweight/obese adults with learning disabilities to lose weight

Primary outcomes: weight loss, BMI reduction

All study designs included

Databases searched: PsycINFO, Global Health, MEDLINE, EMBASE, Database of Abstracts of Reviews of Effects

Other sources: NHS Evidence (Learning Disability specialist collection), UK Health and Learning Disability Network, national policy documents

Three review papers and three intervention studies included following screening

Effectiveness of medications used to attenuate antipsychotic-related weight gain and metabolic abnormalities [12]

Effectiveness of weight reduction medication in treating antipsychotic-related weight gain

Included studies: placebo-controlled randomized controlled trials (double-blind and open label). Primary outcomes: change in body weight and BMI, secondary outcomes include changes in glucose

Systematic review and meta-analysis

Databases searched: MEDLINE, Web of Science, PsycNET, EMBASE, plus hand searching of references.

Thirty-two studies (1482 patients) selected for inclusion following screening

Changes in weight and metabolic parameters during treatment with antipsychotics and metformin [13]

Effectiveness of metformin for patients without diabetes on antipsychotic treatments in reducing weight gain or improving metabolic parameters

Included studies: double blind and open-label clinical studies. Primary outcome: weight change (loss or attenuation)

English language studies only

Systematic review and narrative synthesis

Databases searched: Biosis Previews, Current Contents, EMBASE, MEDLINE

Eleven studies selected for inclusion following screening

Non-pharmacological management of antipsychotic-induced weight gain [14]

Effectiveness of non-pharmacological adjunctive interventions aimed at preventing or controlling antipsychotic-induced weight gain, compared with standard care or active comparator intervention. At least 75% of patients diagnosed with schizophrenia-spectrum disorders.

Included studies: randomized controlled trials. Primary outcomes: mean change of body weight and BMI at intervention end

No language restriction

Systematic review and meta-analysis

Databases searched: Cochrane Central Register of Controlled Trials, MEDLINE, EMBASE, PsycINFO, CINAHL, UMI Proquest Digital Dissertations, SCI-EXPANDED, SSCI, Information Arts and Humanities Citation Index, registers of ongoing clinical trials, plus hand searching of references and key journals

Ten studies (482 patients) selected for inclusion following screening

Grey literature review of health promotion interventions to reduce risk of diabetes in South Asians [15]

Included studies: lifestyle/behaviour change interventions in South Asian populations to reduce diabetes risk; staff awareness-raising interventions

Children as well as adults

Published reviews from 2000

Rapid review of grey literature sources

Sources: university higher degree theses, bibliographies of published reviews, South Asian search engines and journals, websites and other electronic sources, research in progress, expert testimony

Promoting physical activity in South Asian Muslim women through ‘exercise on prescription’ [16]

Aim: to gather knowledge to develop a theoretical framework for the development of pilot intervention

Selection criterion: paper adds knowledge that may be useful for theoretical framework

Broad, but focused, review of empirical and conceptual literature.

Databases searched: MEDLINE, HMIC, Cochrane Library, ERIC, British Education Index, Sports Discus

A total of 24 interventions met all inclusion criteria, covering a wide range of objectives, populations, settings and approaches. Ten interventions focused primarily on identification and monitoring of Type 2 diabetes risk (‘risk identification’), nine involved activities that could help prevent progression to Type 2 diabetes (‘preventative intervention’) and five contained an element of both. Table 3 lists each of these interventions, describing the broad target group and type of intervention. Further details on the 24 interventions, including the source and quality of the evidence, and the main reported outcomes are provided in the Supporting Information (Appendix S2).

Table 3. Overview of included interventions (listed alphabetically)
Intervention nameBroad target groupType of intervention
10% Club (Plymouth)Low income/deprived areaPreventative
Apnee Sehat (Coventry)Ethnic minority/faith groupRisk identification; preventative
CASAHA 50 Plus (London)Refugees, asylum seekers, new migrantsRisk identification; preventative
Choosing the Chance to Change (London) [17]Learning disabilitiesPreventative
Cromwell House weight management clinic (Manchester) [18]Severe mental illnessPreventative
Diabetes UK Early Identification Project (Surrey)Ethnic minority/faith groupRisk identification
Fit for Life (Devon) [19]Learning disabilitiesPreventative
Happy Hearts (Nottingham) [20]Low income/deprived areaRisk identification
Keep Well (Scotland) [21]Low income/deprived areaRisk identification
Keep Well in Prisons (Scotland) [22]PrisonersRisk identification
Keep Well Gypsy and Travellers (Lothian) [23]TravellersRisk identification
Keep Well Gypsy and Travellers (Lanarkshire) [24]TravellersRisk identification
Khush Dil (Edinburgh) [25]Ethnic minority/faith groupRisk identification; preventative
Learning disability diabetes prevention group (Sandwell)Learning disabilitiesPreventative
Lighten Up (Birmingham)Low income/deprived areaPreventative
NHS Health Check: Colchester Mosque (Essex) [26]Ethnic minority/faith group Risk identification
NHS Health Check: Jobcentre Plus (Essex) [27]Low income/deprived areaRisk identification
NHS Health Check: Temporary accommodation (Essex) [28]HomelessRisk identification
‘New life, new you’ (Middlesbrough) [29]Low income/deprived areaRisk identification; preventative
Seek Diabetes (East Midlands) [30]Ethnic minority/faith groupRisk identification
Slimmers' Kitchen (Dudley)Low income/deprived areaPreventative
Slimmer's Kitchen Learning Disability (Dudley)Learning disabilitiesPreventative
Weight Busters (Nuneaton)Low income/deprived areaPreventative
Well-being Support Programme (national pilot [31], Kent local pilot [32])Severe mental illnessRisk identification; preventative

For some target groups (i.e. prisoners, recent migrants), only one intervention was identified. For others, no evidence was found (e.g. sedentary older people, adults with a physical disability). The focus of the included interventions varied from diabetes-specific interventions (mostly risk identification) to identification of associated risk factors and healthy lifestyle improvements with no defined disease-specific endpoint (e.g. interventions to reduce weight or BMI, or increase levels of physical activity).

Much of the identified evidence is drawn from descriptive accounts of project activity (including quantitative service data and feedback from participants and staff) and from before-and-after evaluation studies, mostly involving small sample sizes. None of the formal evaluations included a control group and, in most cases, participants were self-selecting.

Approaches to identify and monitor Type 2 diabetes risk

Seven of the 15 interventions that included a risk identification element were established as part of two national programmes: the NHS Health Check programme in England and Wales (NHS Health Check Colchester mosque, Jobcentre Plus and Temporary Accommodation in Essex) and the Keep Well programme in Scotland [Keep Well Scotland, Keep Well in Prisons and Keep Well Gypsy and Travellers (Lothian and Lanarkshire)]. Both of these national programmes aim to reduce cardiovascular and related risk and also offer some lessons for preventative interventions. Three risk identification interventions were focused specifically on early identification of Type 2 diabetes risk (Diabetes UK Early Identification Project in Surrey, Seek Diabetes East Midlands and CASAHA 50 Plus). Five included risk identification as part of a wider lifestyle programme (Apnee Sehat, CASAHA 50 Plus, Khush Dil, ‘New life, new you’ and the Wellbeing Support Programme).

Risk identification tools

A variety of risk assessment tools were used in the interventions, incorporating different combinations of anthropometric (e.g. weight, height, waist circumference), clinical (e.g. blood glucose) and self-reported (e.g. levels of physical activity, dietary intake, personal and family history) measures. The specific tools used in the interventions included:

  1. QRISK (http://qrisk.org/) used in the NHS Health Check interventions (NHS Health Check Colchester mosque, Jobcentre Plus and Temporary Accommodation in Essex).
  2. ASSIGN cardiovascular risk tool (http://assign-score.com/), used in the four Keep Well interventions [Keep Well Scotland, Keep Well in Prisons and Keep Well Gypsy and Travellers (Lothian and Lanarkshire)].
  3. Diabetes UK Risk Score Assessment Tool (http://www.diabetes.org.uk.uk/Professionals/Risk-score-assessment-tool/), used in the Diabetes UK Early Identification Project in Surrey.
  4. FINDRISC diabetes risk tool (a validated risk assessment tool, using simple measurements including height, weight and waist circumference [34]), used in the ‘New life, new you’ programme.

In some cases (Khush Dil, NHS Health Check Colchester mosque), these tools have been adapted specifically for the target population (e.g. using different BMI or waist circumference cut-offs for South Asian populations, or adjusting risk assessment scores to take account of differential underlying risk), but mostly generic tools were applied.

Outcomes

Risk identification interventions led to improved diagnosis of unmet health needs and conditions specifically related to Type 2 diabetes risk in adults across a range of vulnerable groups. These unmet health needs included raised blood glucose concentration or glycated haemoglobin, overweight or obesity, central adiposity, nutrition imbalance and physical activity below nationally recommended levels. None of the interventions, however, reported outcomes of follow-up activity to monitor changes in identified risk factors.

The reported outcomes suggest that health checks can also help raise awareness of Type 2 diabetes risk factors and encourage engagement in preventative activities among vulnerable groups with low levels of health literacy. For example, the Keep Well in Prisons programme in Scotland reported that some prisoners started using their own money to purchase fresh fruit and vegetables, while others started using the gym, as a direct result of this health screen plus brief intervention programme. Similarly, the NHS Health Check sessions held at a mosque resulted in 14 out of the 18 men assessed being given lifestyle advice, with six being referred directly to a weight-management service. Screening was found to be an effective motivator for lifestyle change in a recent literature review of diabetes prevention initiatives in South Asian communities [15].

Interventions which help to prevent or delay progression to Type 2 diabetes

Fourteen studies and programmes were identified that offered a range of lifestyle interventions targeted at vulnerable adults who were deemed at high risk of developing Type 2 diabetes, including obese and sedentary adults. The aim of the interventions was to support participants in achieving outcomes that may reasonably be assumed to reduce their risk of progression to Type 2 diabetes. Two interventions specifically targeted people who had been identified as having a raised Type 2 diabetes risk as part of the programme, one using FINDRISC (‘New life, new you’) and the other relying on blood glucose measurement (Sandwell learning disability diabetes prevention group). Only very limited evidence was found for pharmacological interventions, from two systematic reviews [12, 13], and no relevant studies describing surgical interventions were identified.

Intervention components

Participants were recruited via a combination of self-referral (n = 10), general practitioner (n = 6) and other health professional referral (n = 7). Programmes included a mix of education (n = 12), behavioural approaches (e.g. personal goal setting, motivational interviewing) (n = 8) and experiential learning (e.g. accompanied shopping trips, cook and eat sessions, led walks) (n = 11) and were mostly delivered in a group setting. Personal goals, food diaries and pedometers are examples of the type of monitoring and motivation tools that have been used.

Outcomes

Eleven intervention reports provided follow-up data that enable some assessment of outcomes to be made, although none used a controlled study design. Moreover, most interventions only followed participants to the end of the programme or for a short period following completion, and so in general it is not possible to assess longer-term outcomes.

On the whole, the interventions led to positive outcomes, with many participants achieving their goals. Examples of reported positive outcomes include improvements in diet (Khush Dil, national Well-being Support pilot), physical activity (CASAHA 50 Plus, Fit for life, Khush Dil, ‘New life, new you’, national Well-being Support pilot) and reductions in weight, BMI or waist circumference (10% club, Cromwell weight management clinic, Khush Dil, Lighten Up, both Slimmers' Kitchen interventions, Weight Busters, national and local Well-being Support pilots).

Despite these positive changes, none of the studies reported whether the interventions led to a reduced incidence of Type 2 diabetes. Furthermore, none of the included projects specifically targeted blood glucose and therefore this was not monitored as an outcome. Unpublished data from the ‘New life, new you’ programme show that mean FINDRISC scores decreased in those with elevated baseline scores. A detailed breakdown of the cause of the reduction in FINDRISC scores is not available, but this can only be achieved by a reduction in waist circumference or BMI, or significant improvements in self-reported levels of physical activity or intake of fruits and vegetables [34]. It is also worth noting that, while FINDRISC is a useful tool for identifying people with raised diabetes risk, it is not yet established as a validated tool for monitoring changes in levels of risk.

Feedback from some of the included projects also highlighted positive outcomes in terms of improved knowledge and attitudes towards healthy lifestyle changes that may protect against progression to Type 2 diabetes. For example, participants in a number of projects demonstrated greater awareness of healthy eating and nutrition and/or were encouraged to maintain and improve their fitness levels (Apnee Sehat, 10% club, CASAHA 50 Plus, Khush Dil, Sandwell learning disability diabetes prevention group, both Slimmers' Kitchen programmes).

The literature reviews confirm that positive outcomes associated with reduced diabetes risk can be attained in vulnerable groups, including overweight or obese adults with learning disabilities and severe mental illness. Well-designed lifestyle interventions can be effective in supporting weight loss in both of these groups [9, 14]. There is also some international evidence that pharmacological interventions (metformin in particular) can prevent or attenuate antipsychotic-induced weight gain in severe mental illness, at least in the short term (although combined interventions may be more effective) [12, 13].

Facilitators and barriers to implementation and positive outcomes

Despite the methodological weaknesses of some of the studies and wide diversity in the project aims and activities, many of the interventions share common approaches that have consistently been reported as contributing to successful implementation as well as supporting positive outcomes. Common success factors are listed below. Each of these success factors is mapped to the interventions that reported it in the Supporting Information (Appendix S3).

  1. Outreach delivery in convenient and appropriate local venues, including places of worship, at local community events, plus home visits where required and feasible (n = 17).
  2. Tailored and flexible approaches that are sensitive to the needs, ability and cultural and religious norms of the participants (n = 13).
  3. Social support, including group-based and peer-led activities and opportunities for the involvement of family, friends and carers (n = 11).
  4. Effective inter-agency working to facilitate inward and onward referrals (n = 10).
  5. Accessible and understandable methods of communication, including translations (for adults who do not speak English as a first language) or adaptations (e.g. for people with a learning disability or low levels of literacy) (n = 9).
  6. Long-term and ongoing support, including post-intervention; for example, drop-in sessions, follow-up calls, maintenance packs (n = 8).
  7. Delivery of programmes by dedicated, sensitive and well-trained staff (n = 7).
  8. Realistic, achievable goals (n = 7).
  9. Involvement of the target community in the design, implementation and evaluation of interventions (n = 6).
  10. Free or low-cost activities (n = 4).
  11. Practical and experiential learning; for example, cooking sessions and supervised exercise (n = 4).
  12. Use of incentives and self-monitoring tools; for example, pedometers, shopping and eating diaries, prizes for achievement (n = 4).
  13. Widespread, varied and target publicity (n = 3).
  14. Multiple contacts, including appointment reminder calls, to encourage uptake (n = 2).

Two interventions in particular highlighted the value of adopting a social marketing approach, which involved the design of tailored interventions and communications based on a detailed understanding of the needs of the target population (Happy Hearts, ‘New life, new you’).

In addition to these generic facilitators, universal leisure cards acceptable across local authority boundaries were suggested as a potentially useful tool to support continued lifestyle change in highly mobile populations (e.g. gypsy and traveller communities). Successful approaches to engaging with gypsy and traveller communities, as highlighted by the Keep Well programme, also included relaxation of programme eligibility criteria (e.g. screening of adults outside the normal age range) and adopting flexibility in the type of referrals made, i.e. to services in demand but out of the Keep Well remit (e.g. for sight and eye tests). The importance of providing crèche facilities was also highlighted in the context of supporting participation in physical activity interventions targeted at South Asian Muslim women [16], but may also apply for women in other target groups.

The risk identification and preventative interventions also highlight similar barriers to implementation, engagement and programme impact, as described below. Each of these barriers is mapped to the interventions that reported it in the Supporting Information (Appendix S4).

Barriers to implementation:

  1. Poor inter-agency working (n = 3).
  2. Literacy and language difficulties of programme participants (n = 3).
  3. Resource implications of costly programme adaptations (n = 3).
  4. Highly mobile and transient populations (e.g. travellers, homeless people, prisoners), which pose challenges for continuity and sustainability of the interventions (n = 3).
  5. Limited access to long-term funding sources (n = 4).

Barriers to achieving positive outcomes:

  1. Weak or unsupportive social networks (n = 5).
  2. Cultural and religious norms and codes of conduct, including health beliefs (n = 4).
  3. Low prioritization of diabetes prevention relative to other, more pressing issues (e.g. financial difficulties, social problems, family crises, existing illness and medication) (n = 4).
  4. Structural barriers, including poor access to facilities to support ongoing lifestyle changes (n = 4).

Additional barriers for adults with severe mental illness and learning disabilities are created by a culture of ‘therapeutic nihilism’ in treating obesity and overweight in these patients, combined with ‘diagnostic overshadowing’ (a process whereby physical symptoms are incorrectly attributed to the person's mental illness) [35]. A recent literature review suggests that implementing primary care health checks for people with learning disabilities may help to overcome some of these attitudes and help to educate general practitioners and practice nurses in the preventative health needs of this group [10].

Discussion

Despite numerous challenges and potential barriers to success, the evidence in this review demonstrates that it is possible to engage successfully with high-risk adults from disadvantaged and vulnerable communities to achieve positive health outcomes relevant to the prevention of Type 2 diabetes. However, the effects are often small and evidence on longer-term outcomes is scarce. Further, no evidence was identified on the cost-effectiveness of these interventions.

As highlighted in the European IMAGE guideline for the prevention of Type 2 diabetes (IMAGE is shorthand for ‘Development and Implementation of a European Guideline and Training Standards for Diabetes Prevention’), there is a growing consensus about the specific components of effective approaches to identifying Type 2 diabetes risk and preventing or delaying progression to disease in high-risk individuals in the general population [36]. These components include staged case-finding procedures, followed by a combination of lifestyle and pharmacological preventative therapies [36, 37]. However, the heterogeneous nature of the included interventions and the quality of the available evidence prevents any firm conclusions being drawn about the most effective approaches to risk identification and prevention in the target groups of the current review.

As a result of a range of personal, family and social circumstances, the vulnerable and disadvantaged adults targeted by the interventions in this review are often excluded from ‘mainstream’ services, including diabetes prevention programmes. While there are some specific challenges to meeting the needs of these groups in such programmes, many of the barriers and successful approaches identified in this review are consistent with evidence from interventions targeted more broadly at the general population of adults at high risk of progression to Type 2 diabetes. This is apparent from the findings of a series of recent systematic reviews commissioned to inform the development of the same NICE public health guidance as the current review. For example, the use of familiar clinical settings and reminder calls were identified as important factors in increasing uptake of risk assessment appointments [38]. Similarly, successful lifestyle changes were found to be facilitated by the involvement and support of family and friends, achievable goals and supportive well-trained professionals who provide feedback and ongoing support [39, 40]. Social support and frequent contacts are also recognized in the IMAGE toolkit as key factors contributing to the success of lifestyle interventions [36]. Likewise, a recent expert review emphasized the promotion of realistic lifestyle changes and investment in high-quality staff training in supporting successful implementation of prevention programmes [37]. This same expert review also recommends the use of multimedia approaches and social marketing techniques to ensure messages are relevant to the target population in order to maximize participation in screening and preventative interventions [37].

Reported barriers to effective implementation in the general population also mirror those identified in this review, and include the short-term nature of funding for preventative interventions, as well as a lack of shared aims and poor coordination between partner agencies [38]. The aforementioned expert review also identified intervention cost and availability of sustainable funding as challenges to implementation [37]. Important inhibitors to positive outcomes include the cost of and access to local exercise facilities and healthy food, other commitments taking priority over lifestyle change, plus low levels of literacy among participants [38]. Many of the findings from the current review of interventions targeting disadvantaged and vulnerable adults therefore have resonance across a broader population base.

The transient and highly mobile nature of many of the vulnerable groups included in this current review (especially travellers and homeless people) pose additional challenges to the design and delivery of diabetes risk identification and preventative interventions. Positive outcomes can also be hindered by the cultural norms and health beliefs of certain disadvantaged communities (e.g. travellers, as well as certain ethnic and religious groups). Moreover, therapeutic nihilism may result in the diabetes risk of people with learning disabilities and severe mental illness being overlooked, thus excluding these vulnerable groups from prevention programmes from which they are likely to benefit.

These challenges require a tailored and flexible approach to prevention for adults in vulnerable and disadvantaged communities. Interventions should be informed by a detailed understanding of the specific needs of the target population, while at the same time recognizing that these communities do not constitute a homogenous whole. The evidence identified by this review suggests that consulting and involving the target community in designing and delivering interventions can be effective in overcoming barriers and delivering positive outcomes. Examples of successful programme components include adaptation of materials and content, practical and experiential learning, activities which are free of charge or very low cost, plus ongoing and continued support beyond the lifetime of the intervention.

An important limitation of this current study is that it relied on a focused, and necessarily pragmatic, approach to identifying relevant interventions, projects and programmes. It is highly probable that a number of relevant local, and perhaps national, initiatives have not been identified, despite attempts to adopt an exhaustive and inclusive search strategy. The diverse nature of the included vulnerable and disadvantaged groups and the focus on UK-based interventions also limit the generalizability of the results of this review.

Conclusion

A number of common themes have emerged from this review regarding successful approaches to the delivery of diabetes risk identification and preventative interventions targeted at adults in disadvantaged and vulnerable groups, who are at significantly increased risk of developing Type 2 diabetes. The individual risk of adults within these groups can be particularly challenging to identify or problematic to manage, for a range of different reasons. It is also possible that existing preventative programmes are not adequately meeting these specific needs. The importance of a tailored approach to diabetes risk identification and prevention, which is flexible and sensitive to the needs of adults from a range of different backgrounds and circumstances, cannot be underestimated. However, more robust evidence on long-term outcomes, including incidence of diabetes, is required to ensure that programmes are designed to achieve success with disadvantaged and vulnerable adults, as well as the wider population at risk.

Funding sources

None.

Competing interests

None declared.

Acknowledgements

We would like to acknowledge the support of the Chair (Professor Kamlesh Khunti) and members of the Programme Development Group committee that led the development of NICE Public Health Guidance, ‘Prevention of Type 2 diabetes: risk identification and interventions for individuals at high risk’. We are also extremely grateful to the invaluable input and advice of Tricia Younger, Associate Director at NICE.

Ancillary