The Impact of Externally Worn Diabetes Technology on Sexual Behavior and Activity, Body Image, and Anxiety in Type 1 Diabetes

Abstract

Background:

We aimed to explore the impact of externally worn diabetes technologies on sexual behavior and activity, body image, and anxiety in adopters and nonadopters of these devices.

Methods:

People with type 1 diabetes aged 16-60 years living in Western Australia were invited to complete an online survey.

Results:

Of the 289 respondents (mean age 34.3 years), 45% used continuous subcutaneous insulin infusion (CSII) and 35% used continuous glucose monitoring (CGM). Approximately half of CSII users stated that the pump interferes with sex. Of these, 75% disconnect their pump during sexual activity to avoid this issue. Comfort during sex influenced the location of the CSII insertion site in 22% of respondents, with the abdomen being preferred. One in four non-CSII users cited sex-related concerns as a factor for not adopting the technology. CGM interfered with sexual activity in 20% of users, but did not commonly affect CGM placement (only 18%). Sexual activity was reported as a factor for not adopting the technology in 10% of non-CGM users. No differences in body dissatisfaction (P = .514) or anxiety (P = .304) between CSII and non-CSII users were observed. No differences in sexual activity and behavior between technology users and nontechnology users were observed.

Conclusion:

Wearable technologies impact upon sexual activity and this influences the decision to adopt the technology. Despite this, technology users are similar in terms of sexual behavior, anxiety, and body image compared to nontechnology users. Where appropriate, these data can be used to identify potential concerns, address strategies to mitigate them, and inform people with diabetes when considering adopting external technologies.

Introduction

Continuous glucose monitoring (CGM) and continuous subcutaneous insulin infusion (CSII) are technologies that facilitate more freedom for people with type 1 diabetes (T1D) regarding their diet and physical activity, control over blood glucose and insulin delivery, and better quality of life.^1-5 Despite this, uptake of these technologies is not universal, even where cost barriers are not a factor limiting utilization.⁶ In Australia less than 50% of children and less than 10% of adults use CSII⁵ despite high rates of private health insurance. Therefore, factors other than cost must influence the decision to adopt externally worn diabetes technologies.

Aside from cost, the common reasons for people with T1D not adopting CSII include not wanting something attached to them⁷ and a perception that CSII is unwieldy, visible, and would interfere with physical activity.⁸ Adult and adolescent CSII users commonly identify that the visibility and difficulty to conceal the technology is a negative aspect to this treatment regimen^9,10 and can lead to increased self-consciousness, particularly in women.¹¹ However, the reasons for not adopting external technologies is not well understood when considering issues of body image and self-consciousness. Currently, there are no data on the perceptions and realities of the impact of external technologies on sexual relationships for those who have adopted external technologies compared to those who have not.

To date, the exploration of body image in people with T1D has focused on comorbid anxiety and eating disorders.^12,13 It is unknown if the proportion of people that choose not to take up CSII are overrepresented with respect to body image concerns. Recent data has found that psychological and social factors influenced adolescents’ decisions to uptake or discontinue CSII⁸; as such we hypothesize that body image, anxiety, and therapeutic choice may be linked, even in adults. Adolescents with T1D are reported to fear sex,¹⁴ and therefore we hypothesize that the prospect of wearing highly visible diabetes technologies could increase this fear and affect therapeutic choice. Those people with T1D that do adopt the technologies are hypothesized to develop solutions to overcome the impracticalities of externally worn technologies during sexual activity, and therefore mitigate any impacts on sexual activity and satisfaction. The objective of this study was to explore the associations between sexual activity, body image, and anxiety in people with T1D who adopt external diabetes technologies and those who do not. We also aimed to explore any concerns regarding the potential of technology to impact on sexual relationships, and practical solutions that technology users utilized to mitigate these.

Methods

Project Design and Methods

The survey was quantitative, using closed and open-ended questions regarding perceptions and realities around person-technology interaction. An online survey tool (Qualtrics) was used as the survey platform. A total of 266 questions were asked; however they included duplicates of questions (altered slightly for gender) that were distributed to male and female participants. The survey was also designed with several “branch, skip, and display” logics, meaning that the survey flow was tailored to each individual participant based on previous answers.

Prospective participants were identified through the Western Australian division of the National Diabetes Services Scheme database and invited to participate in an online survey via a postal mail-out. This database only keeps a physical address, so a postal mail-out was necessary. Only a single mail-out was used to engage potential participants. At the time of recruitment, there were 2574 people with T1D on this database between the ages of 16 and 60 who were identified and invited to participate. Recruitment occurred between November 2016 and February 2017. The study was approved by the Princess Margaret Hospital Ethics Committee (2016089EP).

Participants responded to the survey anonymously and no identifiable information was requested in the survey. No incentives were provided to participants for completion of the survey. The online questionnaire was only made available to each participant if they completed the online consent form indicating informed consent to participate in the research. Participants who consented were given access to the online questionnaire and could exit the process anytime if they did not wish to continue.

Measures

Trait anxiety was indexed using the trait subscale of the Spielberger State-Trait Anxiety Inventory (STAI-T).¹⁵ In the STAI-T, participants are asked to respond to a series of statements on a Likert scale from “almost never” to “almost always” depending on how they feel generally. Body dissatisfaction was measured with the Stunkard figure rating scales.¹⁶ This scale requires participants to select an approximation of their current perceived body size and their ideal body size. A positive score indicates that a participant idealizes weight gain, and a negative score indicates idealizing weight loss. Absolute scores were used to measure degree of body dissatisfaction. The Stunkard figure rating scale is a validated method for identifying body dissatisfaction, but not body image distortion.¹⁷ Questions on sexual health, activity, and behavior were taken directly from the Second Australian Study of Health and Relationships (ASHR2).¹⁸

Original diabetes-specific questions exploring the relationship between sexual activity and technology were developed by the investigators. These questions were reviewed and refined by adults with T1D. Open-ended questions were used to elaborate on responses to quantitative items. These questions gave participants the opportunity to provide further explanation on their experiences with the technologies and how they impact upon sexual health, activity, and behavior.

Analysis

Simple descriptive statistics were presented for outcome measures, demographic measures, and diabetes technology-use: Means and standard deviations were presented for continuous variables and counts/percentages were presented for dichotomous and categorical variables. T-tests and chi-square analyses were used to compare demographic characteristics based on CSII use.

Regression models were used to examine the association between technology-use and outcomes of interest while adjusting for potential demographic confounders; linear regression was used for continuous outcomes, logistic regression was used for dichotomous outcomes, and ordinal logistic regression was used for ordinal outcomes. For the Stunkard body dissatisfaction scale, nonparametric comparisons were used to explore the differences based on technology-use; this was further stratified across gender and age group. All analyses were conducted in Stata 14.0.¹⁹ A P-value of less than .05 was considered statistically significant.

A thematic analysis using a general inductive approach²⁰ was employed by a single researcher (PS) using Nvivo.²¹ Emerging themes were discussed by three of the researchers (PS, MDB, and CR). Themes and subthemes were then summarized and frequency of responses calculated.

Results

Of the potential 2574 respondents, 319 surveys were initiated. Seven respondents did not consent and a further 27 respondents completed less than 10% of the survey and were excluded from analysis, resulting in a sample size of 285. Over the course of the survey a further 19 participants withdrew; all available data up to the point of withdrawal were included in the analyses.

Of the respondents, 53% were female (n = 152), 46% were male (n = 132), and one respondent identified as “other,” but did not wish to specify. The mean age of the respondents was 34.5 ± 13.3 years (range 16-60), and the majority were born in Australia (71.7%), followed by Europe (15.1%), Africa and the Middle East (5.0%), and Asia (2.5%). Only 1.8% identified as Aboriginal or Torres Strait Islander. With regard to technology use, at the time of the survey 44% (n = 125) were predominantly using CSII and 35% (n = 101) were using CGM. Most participants identified as heterosexual (91.8%), with 4.6% identifying as bisexual, 1.8% as homosexual, 1.1% were “not sure/undecided,” and 0.7% were asexual. Major demographic characteristics of the sample by CSII use are shown in Table 1.

Table 1.

Major Demographic Characteristics: CSII Versus Non-CSII.

		CSII Mean ± SD	Non-CSII Mean ± SD	P-value
Age (years)		30.6 ± 12.7	37.5 ± 13.1	.001
Gender (female)		57.6%	50%	.177
Duration of diabetes (years)		16.7 ± 9.9	15.1 ± 12.2	.239
Education status	Less than secondary	9.8%	9.0%	.057
	Secondary	31.3%	19.4%
	Tertiary	59.8%	71.6%

Abbreviation: CSII, continuous subcutaneous insulin infusion.

CSII users were younger than non-CSII users, but there was no statistically significant difference in gender, duration of diabetes, or educational status. Of the CSII group, 45.6% also used CGM (and of those 42.1% reported using it all the time), while 26.9% of those on injections used CGM (and of those 55.8% reported using it all the time). Of non-CSII users, 6.9% had previously trialed CSII.

Anxiety

The mean STAI-T score for the sample was 42.2 (±11.9). In the linear regression model adjusting for age group, gender, and highest educational level, neither CSII use nor CGM use was statistically significant predictor of STAI-T score (b = −1.4 [95% confidence interval −4.4, 1.6], P = .352 and b = 0.5 [−2.5, 3.5], P = .721, respectively). A model including an interaction term for CSII use by CGM use indicated no significant interaction effect (b = −0.5, P = .863).

Body Image

Median absolute difference for non-CSII users was 1 (interquartile range [IQR]: 1, 2), CSII users was 1 (IQR: 1, 2), non-CGM users was 1 (IQR: 1, 2), and CGM users was 1 (IQR: 1, 2). There was no statistically significant effect of CSII status (z = 0.10, P = .918) or CGM use (z = −0.68, P = .498). The pattern of results was consistent when stratifying by gender and age group and when examining the effect of CGM within each CSII group.

Sexual Activity

Overall, 93.0% of the sample reported being sexually active, including 35 of the 47 adolescent (16-18 years old) respondents. Most (74.9%) of the respondents reported that they told their current sexual partner about their diabetes within one month of knowing them, with the highest proportion (35.7%) of people revealing their diabetes in the first 24 hours.

Sexually active CSII users commonly reported (48.6%) that CSII interferes with sexual activity. The major themes (see Supplemental Table S1) emerging from the qualitative responses were: that it gets in the way, it interrupts the moment, gets tangled and pulls, and causes issues or interference with the sexual partner. These common themes are captured in the following response:

Having to stop in the heat of the moment to disconnect can ruin the atmosphere and also the connection site can be a bit hindering during sex, trying not to lean on it too much or scratch my husband with it.

The majority of CSII users (75.2%) disconnect their insulin pump during sex to avoid these problems, and 20.8% of respondents had used pump disconnection as a signal to their partner that they wanted to engage in sexual activity. The most preferred CSII insertion site with respect to sexual activity was the abdomen (49.1%), with 19.8% reporting no preference; 21.7% of people reported that comfort during sex influenced the location of the site.

Within the CSII group, 19.8% reported that they had concerns about how the technology would impact on developing new intimate relationships prior to commencing CSII. The most common thoughts or concerns mentioned were the idea of being attached, the pump getting in the way or disconnecting, and being a mood killer or unattractive (see Supplemental Table S2). For example, one participant stated:

I was single at the time, so I was concerned about hiding it from a first time intimate encounter with a new partner.

In the non-CSII group, one in four users (25.5%) cited concerns about how the technology would impact on sexual activities as a factor for not adopting the technology, and among this group the most commonly reported concerns were the idea of being attached, the pump disconnecting, being unattractive, getting in the way, and being a reminder of the disease (see Supplemental Table S3).

Of the sexually active CGM users, 20.2% reported that it interfered with sexual activity. The most commonly reported problems were it gets in the way, worrying about it coming off, followed by being unattractive, and/or a turn-off (see Supplemental Table S4). Sexual activity was only reported to influence CGM placement in 18% of respondents.

Prior to starting CGM, 12.6% of CGM users reported having thoughts or concerns about developing new intimate relationships while on CGM. The most commonly cited concern was that it was perceived to be unattractive (see Supplemental Table S5). For example, a respondent stated that she was

Concerned that my husband would find the site bothersome or a turn off. (Thought) that I’d look like a robot.

Of the non-CGM users, 9.9% cited intimacy-related concerns as a factor for not adopting the technology, and the most commonly cited reasons were wanting to feel natural with no attachments during sexual activity and that it was unattractive and would get in the way or catch on things (see Supplemental Table S6).

Sexual Behavior

There were no statistically significant differences observed between technology (CGM or CSII) and nontechnology users with respect to any measures of sexual behavior, for example, age of first sexual intercourse (P = .568), number of sexual partners (female P = .642, male P = .271), frequency of sexual intercourse with current partner (P = .916), or ratings of physical satisfaction (P = .304) as shown in Supplemental Figure S1. Technology users did report higher levels of emotional satisfaction with their current sexual partner (P = .026), also shown in Supplemental Figure S1.

Almost all participants (99.2%) reported being 90%-100% honest when answering the questions in the survey. Most did not find the survey embarrassing (72.9% reported that it was not at all embarrassing, 21.4% found it slightly embarrassing).

Discussion

To our knowledge, this is the first detailed study that has explored potential differences in people with T1D who adopt externally worn diabetes technologies and those that do not with respect to anxiety, body image, and sexual activity. We found that there were no differences in body image or trait anxiety between technology users and nonusers. Importantly, approximately one in four participants still using multiple daily injections stated that, when making their decision regarding switching to CSII, concerns about how the wearable technologies would impact on their sexual encounters influenced their decision. Indeed, CSII users reported that CSII can interrupt sex, but by utilizing strategies such as disconnecting their insulin pump, they are able to overcome this problem. There were no observed differences in frequency of sexual intercourse and reports of physical satisfaction during sex between technology users and nontechnology users. This suggests that by adopting practical steps (such as disconnecting the pump) sex is unlikely to be significantly impacted upon by CSII.

The high proportion of technology users citing sex and intimacy-related concerns as part of the decision to adopt technology is an important finding given that in the authors’ clinical experience, discussions regarding sex and technology use are infrequently raised by clinicians. The often-expressed reason for not wishing to adopt externally worn technology, such as the hassle of wearing devices and not liking devices on one’s body,⁷ may indirectly reflect concerns about intimacy and sex. Our data show that this is particularly an issue for CSII possibly due to the size of the device or the presence of the tubing which was reported to pull and get tangled during sexual activities. Therefore, if age and maturity are appropriate, exploring concerns around sexual activity with people considering external technologies is an important area of a clinical discussion. This includes discussions with adolescents, given that many people aged above 16 years are sexually active and are particularly sensitive to concerns about sexual activity.²²

Contrary to our hypothesis, we observed no differences in trait anxiety or body dissatisfaction between technology and nontechnology users. This suggests that while people with T1D who do not want to use CSII often express that they do not want something attached to them, this reaction is unlikely to be driven by underlying anxiety or body image concerns.

CSII users reported developing pragmatic solutions (such as disconnecting the insulin pump) to overcome the commonly cited issues of pump entanglement and dislodgement during sex. One in five participants stated that sex influenced pump insertion site (favoring the abdomen), which suggests that a degree of planning is undertaken by many people. The concerns about intrusion and requirements of planning can impact the spontaneity of intimacy and sexual activity. Despite this, the frequency of sexual intercourse and satisfaction ratings were no different between technology and nontechnology users. These data provide important information for healthcare professionals to consider when discussing the practicalities of sex and wearable technology with people to guide them in making informed decisions.

This study was limited in that the survey had a low response rate (approximately 11% potential respondents) and the resultant sample is not representative of the population with T1D, limiting the generalizability of the findings. The high rate of technology use in the respondents compared to known national data suggests a selection bias toward technology users, and our data may underrepresent issues for nontechnology users. Using an online survey platform may have contributed to this bias.

This study provides novel and important data. The use of mixed quantitative and qualitative questions is an additional strength, allowing us to further explore the issues around wearable technology and sex. The majority of respondents did not find the survey embarrassing to complete, and reported being 90%-100% honest with their answers. Replication is needed to further research in different populations.

Conclusion

In summary, people with T1D who do not adopt externally worn technologies do not show higher rates of anxiety or body image dissatisfaction compared to those who do use these devices. Externally worn diabetes technologies cause concerns for people with T1D with respect to sex, particularly CSII. However, these concerns do not appear to translate into differences in sexual behavior or satisfaction. With the increasing use of diabetes technologies, it is important for healthcare professionals to be aware of potential concerns, address strategies to mitigate them, and reassure their patients where appropriate. Therefore, these data are important for healthcare practitioners to use when counseling patients around externally worn diabetes technologies in the appropriate setting and context, to provide reassurance.