The misuse of insulin by males with Type 1 Diabetes Mellitus for weight and/or shape control: a systematic scoping review

Background Insulin restriction and/or omission in Type 1 Diabetes Mellitus (T1DM) is a risky disordered eating behaviour increasingly reported in the literature. Most existing literature has focused on females with T1DM. Previous research, however, suggests disordered eating behaviours are over-represented in males with T1DM versus males without T1DM. This systematic scoping review sought to summarise the existing literature to contribute to the development of an understanding of males’ misuse of insulin for weight and/or shape control. Methods A systematic scoping review methodology was used. We searched six electronic databases. Eligible articles were quantitative, qualitative, or mixed methods empirical studies with primary data on the restriction and/or omission of insulin for weight and/or shape control among males, written in English from database inception to December 2021. Results A final sample of 56 articles was included (48 quantitative, six qualitative and two mixed methods). Where it was reported to occur, estimates of the behaviour ranged from 5 to 75% in females and 1.4% to 76% in males. In 12 studies examining sex differences, only three reported higher prevalence in females. Understandings about the behavioural context of the behaviour were hampered by lack of qualitative research (n = 4 studies). Conclusion Despite common preconceptions, this disordered eating behaviour appears to occur similarly in males and females. The findings shed light on areas for future research, including aetiology in males and the need for longitudinal and qualitative research. Clinical recommendations include being alert to the possibility of this behaviour in males with Type 1 Diabetes Mellitus and the need for health professionals to use open-ended questions to explore current and past deliberate restriction and/or omission of insulin among their patients. Supplementary Information The online version contains supplementary material available at 10.1007/s40200-022-01151-8.


Introduction
Disordered eating and eating disorders appear alarmingly high in individuals with Type 1 Diabetes Mellitus (T1DM) [1][2][3]. This is concerning given that disordered eating behaviours in individuals with T1DM increase the risk of morbidity and mortality [3][4][5]. A unique, non-dietary, purging behaviour available only to individuals with T1DM or Type 2 Diabetes Mellitus (T2DM) dependent on insulin therapy is the deliberate restriction and/or omission of insulin, inducing hyperglycaemia and weight loss [6][7][8][9].
While a clear diagnostic term and diagnostic criteria are currently lacking [10], individuals deliberately restricting and/or omitting insulin for weight and/or shape control have adopted the term "diabulimia" to describe their behaviour and report a distinct 'diabulimic' identity separate from other eating disorder labels [11]. While it is important to explore and respect terminology of those with lived experience, controversy around the term 'diabulimia' exists as it is often used differently between studies creating confusion regarding its meaning. Some researchers use the term to describe insulin restriction and/or omission for weight and/or shape control, while others use it to describe eating disorders comorbid with T1DM more broadly [12]. It has also been argued that the term is too narrow, predominantly represents bulimic symptomology and does not capture the broad spectrum of disturbed eating behaviour seen in diabetes [12][13][14]. Therefore, in the absence of a diagnostic term, we refer to the deliberate restriction and/or omission of insulin to influence weight and/or shape.
While insulin restriction and/or omission can occur for various reasons beyond weight and/or shape control (e.g., forgetfulness, avoidance of painful injections etc.), insulin restriction for weight and/or shape control" is suggested to occur in 4 to 58% of people with T1DM [14]. Highlighting the seriousness of deliberately restricting and/or omitting insulin, research has found that when compared to other weight control methods, it led to the highest rates of retinopathy and nephropathy [4,15], increased the risk of mortality by 3.2 times and reduced average life span by 14 years [4].
Despite its growing interest to researchers and its clinical significance, deliberate insulin restriction and/or omission for weight and/or shape control remains an understudied problem with females being consistently over-represented in existing research [14]. As a result, much less is known about males' disordered eating behaviours. A lower incidence of eating disorders reported in males generally may contribute to the paucity of research investigating disordered eating behaviours in males with T1DM [16]. Although recent prevalence rates for eating disorders in males are likely a gross underestimate [17], research has demonstrated that males are at risk of eating disorders and with comparable levels of distress and disability to females [18,19]. Furthermore, the prevalence of eating disorders among males has increased dramatically over the last two decades [20], and emerging evidence suggests that disordered eating behaviour in males may be increasing faster than in females [19].
Importantly, research has demonstrated that compared to males without diabetes, disordered eating behaviours are still overrepresented in males with T1DM [7,21]. Research has also reported high levels of body dissatisfaction and drive for thinness among males with T1DM [21][22][23]. Given differences in prescribed sociocultural body ideals for males and females [24], which have implications for the associated maladaptive behaviours individuals may undertake to achieve such ideals [25,26], it is reasonable to assume differences in disordered eating behaviours may exist among males and females with T1DM.
A preliminary search of MEDLINE, the Cochrane Database of Systematic Reviews and Johanna Briggs Institute Evidence Synthesis confirmed that no published systematic or scoping reviews concerning the deliberate misuse of insulin among males with T1DM for weight and/or shape control existed. A manual search of grey literature yielded one systematic review exploring insulin restriction and/or omission for weight and/or shape control in males across n = 16 studies; this was part of an unpublished dissertation [27].
We aimed to identify and summarise existing knowledge about the misuse of insulin among males with T1DM for weight and/or shape control to answer the following primary review questions: 1. What is the nature of males' misuse of insulin for weight and/or shape control, including population characteristics? 2. What influences males' misuse of insulin for weight and/ or shape control? 3. What are the impacts of males' misuse of insulin for weight and/or shape control? 4. What gaps exist in the current literature about insulin misuse for weight and/or shape control among males?

Design and search strategy
A systematic scoping review methodology was employed using the the Joanna Briggs Institute [28] guidelines and the Preferred Reporting Items in Systematic Reviews and Meta-analyses guidelines (PRISMA-ScR; [29]; see Supplementary Table 1). The priori protocol was registered on Open Science Framework (removed for blind review). The search strategy aimed to locate published and unpublished studies. Six databases (Psychinfo, PubMed, Embase, Web of Science, CINAHL and Scopus) were searched for relevant studies from database inception to December 2021. Initially, preliminary searches were undertaken across the databases to identify relevant subject headings and keywords. The final search strategies were finessed in collaboration with an experienced research librarian (See Supplementary Table 2) and database searches were supplemented with a manual search of reference lists of included studies.

Eligibility criteria and study selection
We used the Population, Concept, and Context (PCC) framework [28] to conceptualise the research questions as follows.
Participants: Studies that include male participants, of all ages, with T1DM deliberately restricting and/or omitting insulin. Concept: Deliberate restriction and/or omission of insulin for weight and/or shape control. Context: Any geographic setting.
Inclusion criteria were quantitative, qualitative, mixed methods and case studies published from database inception to December 2021. Studies were included if they: (i) investigated insulin misuse for weight and/or shape control or within the context of disordered eating, (ii) investigated males with T1DM (studies investigating males and  females were eligible for inclusion if data about males  was reported separately; studies investigating males with  T1DM and Type 2 Diabetes Mellitus (T2DM) were eligible for inclusion if data about males with T1DM was reported separately), (iii) presented primary data, and (iv) were written in English. This scoping review considered experimental and quasi-experimental study designs, analytical observational studies (prospective and retrospective cohort studies, case-control studies and analytical crosssectional studies) and descriptive observational designs (descriptive cross-sectional studies, individual case reports and case series) for inclusion. Qualitative studies using designs such as phenomenology, grounded theory, ethnography, thematic analysis, and qualitative description were also eligible. Relevant unpublished studies or grey literature such as reports, dissertations and theses were also considered.
Excluded studies were those not written in English, abstracts, editorials or opinion pieces, discussion or review articles not reporting primary data, and studies using a comparator (e.g., females) that did not present male data separately. Studies were also excluded if they investigated insulin restriction and/or omission for reasons other than for weight and/or shape control (e.g., forgetfulness or avoidance of pain).
The database searches identified 3,756 potentially eligible studies. A further 14 studies were sourced manually from the reference lists of database-identified articles, resulting in 3,770 studies for consideration (see Fig. 1). All identified references and abstracts were pooled into EndNote X9 software to remove duplicates, screen for titles and abstracts and full-text in detail based on the predetermined eligibility criteria. The first and second authors co-screened a random subset of 10% of potentially eligible studies (interrater agreement 99%, K = 0.97, p < 0.05), with any discrepancies resolved by consensus discussion.
The findings of the included articles were extracted by the first author and cross-checked by the second author. Extracted data included research setting/country, year of publication, study design, sample size and characteristics, key findings on males' experiences of insulin restriction and/or omission and any relevant limitations or biases.

Included
Full-text articles assessed for eligibility (n = 14) Records excluded, with reasons (n = 9) Investigated diabulimia, however, separate data on males unable to be obtained (authors contacted; n = 1 data not available, n = 4 no response) Investigated males with T1DM and disordered eating, however, separate data on diabulimia unable to be obtained (authors contacted; n = 4 no response) [29] demonstrating the article selection and screening process

Description of included studies
The final sample comprised 53 peer-reviewed studies (published between 1982 and 2021) and three unpublished dissertations, representing 46 samples; 48 studies reported quantitative data, six studies reported qualitative data, and two studies were mixed methods (Table 1). Most studies originated from high-income countries (United States of America (n = 10), the United Kingdom (n = 9) or Norway (n = 9)).

Design
Most studies (n = 41) employed cross-sectional designs. Other designs included case studies (n = 4 studies), case-control (n = 4), longitudinal cohort (n = 3) and phenomenological designs (n = 2). One study employed a retrospective chart design [30], and another was a randomised feasibility trial [31]. Quantitative studies most commonly used a mixture of self-report questionnaires, variations of semi-structured, structured or diagnostic interviews and/or information about participants' physical health (e.g., glycosylated haemoglobin [HbA1c], BMI etc.) collected from medical records or within clinic appointments (n = 46). Qualitative studies were predominantly case studies that described clinical case material [31][32][33][34]. The two mixed methods studies used self-report questionnaires [8,31].

Participant characteristics
All studies except for one [23] recruited male and female participants. Where possible, characteristics of the male participants were extracted separately or requested from authors (See Table 2).The total sample size of participants with T1DM across 45 studies was n = 12,450. Sample sizes varied widely from one (case studies) to 2,837 (M = 276.67, SD = 471.67). Across 43 studies, the total sample size of males with T1DM was 5,351 (M = 124.44, SD = 236). The mean age of males with T1DM across 21 studies was 19.67 years (SD = 2.87; n = 14) (see Supplementary Table 3 for the characteristics of all participants).

Measures
Various measures were used to assess insulin restriction and/or omission in the context of weight and/or shape control, most of which were not psychometrically validated (See Table 1). Across the included studies, 24 studies used a measure that assessed insulin restriction and/or omission for weight and/or shape control [8, 21, 27, 35, 39, 40, 42, 43, 47-49, 51, 55-66]. The remaining studies asked about insulin restriction and/or omission broadly without identifying underlying reasons (n = 19; 22, 31, 36-38, 45, 50, 52-54, 67-75) or did not provide sufficient information to determine if reasons were accurately identified (n = 4; 23, [76][77][78]. The measurement of insulin restriction and/or omission was unclear in nine studies [14, 32-34, 41, 44, 46, 79, 80]. Most commonly, studies measured engagement in the behaviour across the previous 28 days (n = 26 studies). The rating period across all included studies, however, varied from the previous seven days to ever, and in many studies (n = 16) the rating period was unclear.

Quantitative, qualitative and mixed-methods findings
Across the included studies, the results for males with T1DM restricting and/or omitting insulin for weight and/or shape control were often combined with those of females. Where possible, data separated by sex were extracted and are presented below.

Quantitative Studies
Rates of insulin restriction and/or omission in males with T1DM. Of the 33 quantitative papers, (25 samples), where insulin restriction and/or omission was identified among males with T1DM [21-23, 27, 35-39, 42, 43, 45, 50, 52-56, 58, 61, 65-70, 77], interpretable point estimate rates ranged from 1.4% [21,35] to 76% [52][53][54]75]. Fourteen quantitative studies reported no males restricting and/or omitting insulin for weight and/or shape control [30-32, 40, 41, 44, 46-49, 51, 59, 60, 64, 78]. One study reported only the percentage of males engaging in insulin restriction and/or omission who scored 20 or more on the Diabetes Eating Problems Survey -Revised (DEPS-R [73]; e.g., 37.5% and 25%, respectively [71]). One study reported suspicion of insulin restriction and/or omission in 2.4% of males, as determined by an expert panel [81]. and approximately 75% [30]. Of these, differences between males and females were assessed in 12 samples. In Powers et al.'s study [45], females tended to report skipping insulin doses significantly more frequently than males. Females were also found to misuse insulin for weight and/or shape control more than males in Matthews [27] dissertation and in the study by Peducci et al. [43], however these differences were only marginally significant. In contrast, rates of insulin reduction or insulin omission between males and females with T1DM were not significant in studies by Troncone [68,69]. The frequency of insulin omission and intentional reduction of insulin was also not significant [69]. Baechle et al. [67] similarly reported no significant difference between males and females when measuring insulin restriction at a frequency of at least three times per week and more than five times per week. Neither insulin restriction (or insulin restriction for weight and/or shape control was associated with sex in Beam et al. [66]. d'Emden et al. [39] also reported no difference when examining the presence of insulin manipulation or omission in males and females. Markowitz et al. [73] reported no difference between males and females regarding clinician-suspected insulin restriction or omission rates. While Bächle [sic] et al. [37] noted there were significantly more females than males who reported disordered eating behaviours using the SCOFF eating disorders screening questionnaire [82], insulin restriction did not significantly differ between males and females. In Baechle et al. [70], disordered eating symptom frequencies were consistently higher among females than males, except for insulin restriction. In Altinok et al. [71], while a greater proportion of females scored above the DEPS-R [73] cut-off, more males than females reported insulin restriction and insulin omission in this group. Finally, in the studies by Wisting et al. [52][53][54]75], using the DEPS-R [73], males reported insulin restriction and insulin omission significantly more than females.

Frequency of insulin restriction and/or omission in males with T1DM.
Araia et al. [22] reported that across 14 days, approximately 14.4% of males misused insulin 1-3 days, and 3.8% of males misused insulin four or more days. Similarly, Baechle et al. [67] reported that 18.5% of males restricted insulin at least three times per week, with 6% restricting more than five times per week. In Beam et al.'s [66] study, males who endorsed insulin restriction for weight control reported this occurring 'rarely'. Insulin restriction occurring more than five times a week was used as a proxy for  [37] reported that the mean number of carbohydrate units without insulin coverage per week for those who were SCOFF negative was 13.2 (SD = 10.6) and 18.8 (SD = 18.6) for those who were SCOFF positive. Troncone et al. [69], who used restriction or omitting at least one insulin shot per week as a proxy for intentional insulin restriction and/or omission, reported that 31.5% restricted insulin and 11.5% of males omitted insulin. Furthermore, Pursey et al. [61] reported that the mean frequency of insulin restriction in males across the past-28 days was 2.90 (SD = 7.245) compared to 1.67 (SD = 3.890) in females; the difference was not statistically significant. Males also had a larger range of days insulin was restricted across the past 28 days (0-28 days) than females (0-14 days) [61], indicating restriction and/or omission of insulin for weight control daily in one or more of the males. Similarly, while the mean score of insulin restriction using the DEPS-R ( [73]; Item 4) ranged between "Rarely" and "Sometimes" for males [36, 52-54, 72, 75], answers ranged between 1 "Never" to 5 "Always" indicating some males are potentially engaging in the behaviour frequently. This was similar for insulin omission [36, 52-54, 72, 75], and ranged from 0 "Never" to 4 "Usually".
Measures of insulin restriction and/or omission among males with T1DM. Insulin restriction and/or omission in males was identified more often when self-report measures were used compared to clinician-administered interviews. Phillipi et al. [44] was the only study using a self-report questionnaire that did not identify males misusing insulin, although exactly how insulin misuse was measured was unclear. Of the studies that used a clinician-administered interview, only two studies identified males misusing insulin (total of three males; 43,55,56). The remaining studies found no male participants misusing insulin [30-32, 47-49, 51, 59, 60, 64]. Similarly, researchers who used a clinician-administered interview and a self-report questionnaire either found no males reported insulin manipulation [40,41,46,57] or found discrepancies in males' reports [23]. Interestingly, different rates of insulin restriction and/or omission were reported across two studies that shared the same sample of males [22,36].

Correlates of insulin restriction and/or omission among males with T1DM.
Most included studies' analyses combined male and female data. Of the included quantitative studies, seven reported correlates of insulin misuse. Wisting et al. [50] found that males who restricted insulin had significantly higher HbA1c than males not restricting insulin. However, Wisting et al. [51] found no significant associations between the pattern of eating and age, BMI, self-induced vomiting, binge eating, excessive exercise, intentional insulin omission, or metabolic control among males. Bächle [sic] et al. [37] in contrast reported that HbA1c values were highest across males who restricted insulin compared to males who did not report insulin restriction. Individuals restricting insulin had significantly worse average HbA1c values compared with individuals who were not restricting insulin, with HbA1c differences being greater among adolescent males than adolescent females [57]. One of the males misusing insulin in Svennson et al.'s [23] study had a high score on the Drive for Thinness subscale of the EDI-C [83]. Additionally, Araia et al. [22] found that body dissatisfaction was significantly correlated with eating binges and BMI for both sexes but was only significantly correlated with insulin omission for females. Among males in the study by Beam et al. [66], skipping insulin to lose weight was significantly correlated with HbA1c levels, diabetes self-management, overall difficulty with emotional regulation, difficulties with impulse control, limited access to strategies to regulate negative emotions, lack of clarity about what emotions are being experienced and symptoms of depression.

Qualitative studies
Across the six included qualitative studies, restriction and/or omission of insulin for weight and/or shape control was identified in four studies, with a total of four males [34][35][36][37]. Within these studies, males indicated a strong desire to lose weight [34,62,79,80], and as described by one individual, insulin restriction and/or omission was akin to a "magic pill" ( [62], p.62); He expressed making a conscious decision to make his diabetes "even more of a mess" in pursuit of weight loss ( [62], p. 74) and worth the sacrifice of poor health. Indeed, alongside the easy weight loss came a large physical and mental burden in the form of high blood glucose levels with blood sugars over 25 mmol/L [34], little energy [62], retinopathy [34,62,80], and feelings of regret, as one male described being "half of everything [he] could have and should have been" ( [62], p. 67). One male engaging in insulin restriction reported welcoming feelings of "queasiness" experienced when his blood sugar levels were high, which diminished his appetite [79]. Another also described feeling better when he was a "bit sugary" [80]. He further expressed that he needed higher blood levels than others to feel well, which was also related to a dread of experiencing hypoglycaemia [80].
Interestingly, Morris [62] noted that the feelings of safety, belonging, and desirability that females gained through insulin restriction was not shared by the male participant. For this male, the behaviour was seen as a way of achieving better fitness performance through weight loss [62] and was used intermittently: "It was not an ongoing thing; it was like a switch that I would turn off and turn back on" ([62], p 74). In contrast to the male described by Morris [62], the males described by Szmukler and Russell [80] and Szmukler [79] reported the deliberate restriction and/or omission of insulin among other disordered eating behaviours (e.g., restriction of food intake, purging). McConnell et al.'s [33] case study reported the deliberate omission of insulin by a 9-year-old male in response to "unpleasant situations" rather than for weight and/or shape control, while Sien et al. [63] had no males who restricted and/or omitted insulin for weight and/ or shape control.

Mixed methods studies
Falcao and Francisco [8] found that 11 participants with T1DM (20.8%) reported intentionally omitting insulin; of these, only three were males, and no males reported doing it intentionally to lose weight. Therefore, the qualitative data concerning insulin restriction and/or omission for weight and/or shape control did not contain any males with T1DM. Of the 32 adolescents screened for inclusion by Boggiss et al. [31], 15 males (46.9% of the total sample) responded to at least one item about omitting or restricting insulin (either item 4 'when I overeat, I don't take enough insulin to cover the food' or item 13 'after I overeat I skip my next insulin dose) with a 1 (rarely) or higher versus 13 females.

Main findings and implications
Our scoping review, the first of its kind, suggests that the deliberate restriction and/or omission of insulin is a unique disordered eating behaviour that may occur at comparable rates in males with T1DM to that of females with T1DM. Where it was reported to occur, estimates of the behaviour ranged from 5% [42] to 75% [30] in females and 1.4% [21,35] to 76% in males [52][53][54]75]. Across 12 samples [27, 37, 39, 43, 45, 52, 54, 66-71, 73, 75] where sex differences were examined, higher prevalence in females was reported in only three [27,43,45], suggesting sex differences may not be as great as earlier thought. Furthermore, our synthesis summarises what is currently known about the behaviour among males with T1DM, including correlates, and provides insights into limitations and directions for future research, as well as clinical recommendations. Our review also identifies several areas of concern about how researchers conduct and report results on insulin restriction and/or omission in males with T1DM.
Consistently, there is a marginalisation of males with T1DM within research exploring this dangerous behaviour and a paucity of research examining the nature and relationships of insulin restriction and/or omission for weight and/or shape control in males with T1DM. It is plausible that research in this area may be somewhat stunted by the idea that disordered eating behaviours occur much less often in males with T1DM and males generally than in females [84,85]. Nonetheless, our review demonstrates sufficient evidence exists to suggest that this dangerous behaviour may be a significant problem for males and that it warrants attention in the literature especially given the severe consequences it poses to their morbidity and mortality [4,15].
Of the included studies, many also did not provide data on the relationships of insulin restriction and/or omission with other variables among males. This omission made it difficult to understand why males may or may not deliberately restrict and/or omit insulin for weight and/or shape control and how it impacts males. While current evidence is limited, the driving motivators for insulin restriction and/ or omission for weight and/or shape control may be qualitatively different for males than females. Findings suggest that reasons beyond body dissatisfaction may drive the behaviour among males with T1DM, including a desire to improve fitness performance [11] or in response to negative affect and difficulty regulating emotion [66]. Both negative affect generally and negative diabetes-related affect have previously been associated with increased odds of restricting insulin [86]. However, our results suggest differences exist in how emotion dysregulation relates to insulin restriction and/or omission for weight and/or shape control between males and females. Health professionals may wish to explore motivators for the behaviour as males and females with T1DM may be motivated by different factors. There have been previous calls for researchers to identify the motivation underlying insulin restriction more carefully [87], as individuals with T1DM may restrict and/or omit insulin due to fear of injections, fear of hypoglycaemia, denial of disease and avoidance of injections in social settings [32].
Diverse and generally poor assessment of insulin restriction and/or omission for weight and/or shape control also limited our review. Our findings identified the need for a standardised measure of insulin restriction and/or omission for weight and/or shape control while considering differences in males' and females' prescribed body ideals. Researchers' use of customised or adapted questionnaires to measure insulin misuse is also highlighted and may be due to the limited data on effective screening tools. Standard questionnaires for disordered eating behaviours and eating disorders for individuals with T1DM have been criticised for under-and over-estimating disordered eating behaviours [3, 1 3 73]. While currently the most validated tool for adolescents and adults with T1DM, the DEPS-R [73,88] does not directly enquire about insulin misuse for weight and/or shape control. The DEPS-R [73] and other existing tools for measuring body dissatisfaction, body image concerns and disordered eating behaviours are also often female-centric and less geared towards concerns more commonly held by males [84]. Furthermore, there are often high levels of secrecy and shame around disordered eating behaviours and eating disorders including insulin misuse [11,27,[89][90][91], and individuals may be secretive about their engagement in disordered eating behaviours for fear of stigmatisation [92,93] or due to low motivation for change [94]. Against this backdrop, identifying insulin restriction and/or omission generally among people with T1DM is difficult; however, our results suggest that a self-report measure capturing insulin restriction and/or omission may facilitate more honest reporting by individuals and an increased willingness to disclose the behaviour. In the absence of a psychometrically robust measure, we suggest that health professionals use a validated, self-report measure to explore the presence of body dissatisfaction and disordered eating behaviours, followed by open-ended questions to explore current and past deliberate restriction and/or omission of insulin among their patients. Table 3 outlines possible measures for use  among males with T1DM, alongside sensitive open-ended questions to gather information about insulin use and weight and shape concerns.
Our review was limited to studies in English, and we excluded studies examining the behaviour in T2DM due to different aetiology and management aspects compared to T1DM. Also, possibly eligible studies were excluded where male data were not reported separately, and we received no response from the authors when we sought these data. The included studies were heterogeneous and could not be directly compared statistically, which limited interpretation. In addition, as included studies primarily used cross-sectional designs, we could not make inferences about cause and effect. Most included studies focused on individuals aged in early adulthood, which hinders the generalisability of the results to other age groups. The included studies did not provide information on males' sexual or gender identity, and therefore little can be concluded on how this may be related to insulin misuse. Additionally, future research comparing males with and without diabetes and employing longitudinal study designs is needed to explore contributing factors and trajectories of disordered eating in diabetes.
Most studies tended not to include comparison groups of males without T1DM, so it is unclear to what extent the explored factors are relevant to males with T1DM only or also to other male populations. Many participants were recruited from a clinic or specialist centre and therefore may not be representative of the larger population of individuals with T1DM. Also, we cannot rule out possible selection bias. It may be that non-respondents are more likely to engage in insulin restriction and/or omission for weight and/or shape control or have more severe symptoms.
Given that data for males and females with T1DM was most commonly combined, we recommend future research report data separately by sex [96]. Also, more precise terminology should be used when reporting insulin restriction and/or omission among males and females with T1DM, as different definitions reflect different behaviours [4,88,97]. Additionally, as males are poorly represented in qualitative research, with reports from only four men included in our review, we recommend further qualitative research to gain males' experiential perspectives of insulin restriction and/ or omission for weight and/or shape to better understand behavioural aetiology and correlates. •How satisfied are you with your weight and/or shape? •How does your weight and/or shape affect the way you feel about yourself? •How important is it to you to change your weight and/or shape? •How would you want to change your weight and/or shape? •How would you go about changing your weight and/or shape? •What behaviours might you engage in to change your weight and/or shape? Insulin Use Questions •Some people with diabetes find it difficult to manage their insulin treatment. How is this going for you? •Your blood sugar levels have been elevated recently. What do you think may be the reason/s for this increase? •Could you tell me about the reasons you skip or reduce insulin?

Conclusions
Previously thought to occur much more frequently in females with T1DM, our results suggest that males with T1DM engage in insulin restriction and/or omission for weight and/ or shape control at similar rates. Health professionals should be alert to the potential for its presence in both sexes and should employ a self-report measure to explore the presence of body dissatisfaction and disordered eating behaviours, followed by open-ended questions to inquire about the presence of insulin restriction and/or omission for weight and/ or shape control. This vigilance is particularly critical given the increased risk of morbidity and mortality this behaviour poses to individuals, regardless of sex.

Declarations
Ethics The paper reports a scoping review that undertook a secondary analysis of data already collected (from human participants) and does not require ethical approval.

Consent for publication Not applicable.
Public significance statement This review is the first to summarise existing literature on the deliberate restriction and/or omission of insulin as a disordered eating behaviour in males with type 1 diabetes mellitus. The results highlight several concerns in the research thus far, demonstrate the need for further research on the behaviours' aetiology and impact within males and highlights the need for health professionals to be aware of its presence in males and females.

Conflicts of interest
On behalf of all authors, the corresponding author states that there is no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.