Greater Than Expected Prevalence of Type 1 Diabetes Mellitus Found in an Urban Gender Program

Abstract

The prevalence of type 1 diabetes mellitus among transgender and gender diverse (TGD) youth is nearly five times higher than in the general pediatric population (9.9 per 1000 people vs. 1.93 per 1000 people). We hypothesize that minority stress experienced by TGD youth may lead to a higher prevalence of diabetes.

Introduction

The Gender Multispecialty Service (GeMS) at Boston Children's Hospital is a program focused on assessment and medical management of transgender and gender diverse (TGD) children and adolescents. For the past several years, clinicians have anecdotally noted an increased occurrence of type 1 diabetes mellitus (T1DM) in the clinic population.

There have been two prior reports of an association between T1DM and gender dysphoria; the first by Defreyne et al. reported a 2.3 times higher prevalence of T1DM in their population of transgender adults in Belgium when compared with the Belgian population as a whole (9 per 1000 vs. 4 per 1000) and the second, a report in TGD youth in the United States, found a 9.4-fold higher prevalence of T1DM in TGD youth compared with the prevalence in patients without TGD (24.77 per 1000 vs. 2.68 per 1000).^1,2 In the United States, the prevalence of T1DM in the general pediatric population is estimated to be 1.93 per 1000.³ A similarly low prevalence of gender dysphoria among children and adolescents in the United States, estimated to be between 5 and 7 per 1000, would suggest a rare co-occurrence given such a low combined prevalence.^4,5

Methods

Patients with T1DM and gender dysphoria seen in the GeMS program between January 1, 2007 and December 31st, 2018 were identified by International Classification of Diseases (ICD)-9 and ICD-10 code. Prevalence of T1DM was calculated and compared with previously published rates.³ Z-scores for anthropometric measures were determined based on the Centers for Disease Control (CDC) growth data and are presented for both sex designated at birth and affirmed gender.⁶ This study was approved by the Institutional Review Board.

Mean and standard deviation were used to summarize normally distributed variables, median and range were used to summarize non-normally distributed variables, and comparisons were made through Student's t-test and the Mann–Whitney test, respectively. Proportions were compared using Fisher's exact test. A p-value of <0.05 was considered significant, with no adjustment for multiple testing for this descriptive analysis. Stata Statistical Software: Release 16 (College Station, TX) was used for calculations.

Results

There were 1114 patients seen in GeMS during the study period, of which 11 had T1DM (Table 1). Of the TGD youth with T1DM 8 (72%) were white, 2 (18%) were African American, and 1 (9%) was Hispanic/Latinx. The majority, 9 (82%), were designated female at birth and had a transmasculine gender identity. Nine adolescents had initiated gender-affirming hormones (GAH); eight received testosterone, one received estrogen, and two had received pubertal suppression, of which one also received GAH.

Table 1.

Demographic Characteristics of Gender Diverse Patients With and Without Type 1 Diabetes Mellitus

	TGD (n=1114)	TGD and T1DM (n=11)
Age at first gender clinic visit, years, mean (SD)*	13 (3.7)	16 (2.7)
Race, n (%)
White	806 (72)	8 (72)
Hispanic/Latino	61 (5.5)	1 (9)
African American	30 (2.7)	2 (18)
American Indian or Alaskan Native	6 (0.5)	0
Asian	27 (2.4)	0
Other/unknown	184 (17)	0
Designated sex, n (%)
Male	390 (35)	2 (18)
Female	724 (65)	9 (82)
Affirmed gender, n (%)
Male	663 (60)	8 (73)
Female	353 (32)	2 (18)
Nonbinary	98 (8.8)	1 (0.09)
Gender-affirming therapy received, n (%)
GnRH agonist	177 (16)	2 (18)
Gender-affirming hormones	482 (43)	9 (82)

^*p-value <0.05.

GnRHa, gonadotropin-releasing hormone analog; SD, standard deviation; T1DM, type 1 diabetes mellitus; TGD, transgender/gender diverse.

The TGD patients with T1DM presented to the GeMS clinic at an older age than the overall clinic population (16 vs. 13 years, p=0.007). There was no difference in race, designated sex, affirmed gender, or proportion receiving gender-affirming treatments between the T1DM patients and the broader GeMS population. The mean BMI in the T1DM group was in the normal range (24.6±2.2 kg/m²), with six individuals categorized as normal and five as overweight (Table 2).

Table 2.

Clinical Characteristics of Gender Diverse Patients with Type 1 Diabetes

Subject	Race/Ethnicity	Designated sex	Affirmed gender	Age at T1DM diagnosis (years)	Age at first GeMS visit (years)	Antipancreatic antibody	BMI at first visit (kg/m2)	BMI z-score for sex designated at birth	BMI z-score for affirmed gender	BMI category	Gender-affirming medical therapy	T1DM insulin delivery
1	White	F	M	3	20	?	22.5	0.22	−0.17	Normal	GAH	MDI
2	Black	F	M	9	16	?	23.4	0.45	0.13	Normal	GAH	Pump
3	Hispanic	F	M	3	15	?	24.4	0.99	1.1	Normal	GAH	Pump
4	White/Hispanic	M	F	16	15	+	28.2	1.8	1.7	Overweight	Both	Pump
5	White	F	M	8	15	?	20.4	0.11	0.13	Normal	GAH	MDI
6	White	F	M	6	16	?	26.0	1.2	1.3	Overweight	GAH	MDI
7	Black	M	F	6	10	+	27.2	2.2	2.2	Obese	GnRH	MDI
8	White	F	M	9	16	+	24.8	1.0	1.1	Normal	GAH	MDI
9	White	F	M	11	13	+	23.5	1.2	1.4	Overweight	GAH	Pump
10	White	F	NB	11	18	+	23.0	0.47	0.33	Normal	Neither	MDI
11	White	F	M	11	18	+	27.0	1.6	1.7	Overweight	GAH	MDI

BMI, body mass index; F, female; GAH, gender-affirming hormones; GeMS, Gender Multispecialty Service; M, male; MDI, multiple daily injection; NB, nonbinary; Pump, insulin pump; ?, unknown.

All patients had T1DM before presentation to the GeMS program except one, who was diagnosed within 2 weeks of their first appointment. The mean age at T1DM diagnosis for this cohort was 8.45±3.7 years (range 3–16 years), which was significantly younger than the average age at T1DM diagnosis in the United States of 14.5 years.³ The average age at the first visit to GeMS among those patients with T1DM, 16±2.7 years, was older than the average in our clinic, which is 13±3.7 years.

The prevalence of T1DM among children and adolescents in the United States is 1.93 per 1000; thus, the expected number of patients with T1DM among the 1114 TGD youth seen during the study period is 2.15.⁷ We found 11 individuals with T1DM and gender dysphoria, which is a prevalence rate of 9.87 per 1000, or a 5.11-fold higher prevalence rate than would be expected for youth with T1DM and gender dysphoria.

To estimate glycemic control in the patients with T1DM seen in GeMS, the mean hemoglobin A1c (HgbA1c) values 6 months before the first GeMS visit was compared with the mean value the 6 months after the first GeMS clinic visit. We noted a downward trend in HgbA1c from 8.46% to 8.16% among those patients with values available.

Discussion

This retrospective study revealed a significantly greater prevalence of TGD children and adolescents presenting with T1DM than seen in the general pediatric population. These findings contribute to the growing literature surrounding this topic.^1,2 Similar to Logel et al., we found that the prevalence of T1DM among TGD youth is fivefold higher than in the general population.²

Although apparently correlated, a causal association between T1DM and TGD has yet to be demonstrated or explained. One compelling explanation is that the minority stress secondary to TGD status serves as a trigger for T1DM onset. It is well documented that TGD adults and adolescents have increased psychological stress with higher rates of social stigma, poor body image, depression, and suicidality.^8–10 Stress increases the activity of the sympathetic nervous system and release of cortisol and other counter-regulatory hormones that decrease insulin sensitivity. Decreased insulin sensitivity leads to a higher demand for insulin release putting pressure on the pancreatic β cells, which may hasten autoimmune destruction and β cell failure.¹¹ We and others have found that patients with T1DM and TGD present with T1DM younger than the U.S. average, at 8.5 years in our report and 9.9 years in the previous report compared with a national average of 14.5 years.³ The additional minority stress that TGD patients with T1DM experience compared with their cis-gender peers with T1DM may lead to earlier β cell failure and earlier presentation.

Gender-affirming medical treatment has been shown to decrease psychological stress in TGD individuals. We hypothesized that presenting to our gender-affirming clinic may, through a reduction in stress, improve insulin sensitivity and lead to improved glycemic control. A marginal improvement was observed in HgbA1c values, an average measure of glycemic control for the past 3 months, before and after the first gender clinic visit.¹² Although not statistically significant, this downward trend suggests that initiating gender-related care may have a positive effect on glycemic control. We did not have enough data available to investigate the impact of GAH or pubertal blockade on glycemic control.

There are several limitations of our study and previous reports that may explain the link between gender dysphoria and T1DM.^1,2 The available data are retrospective, observational, and done in majority Caucasian populations. T1DM is more prevalent in the Caucasian population, and as such, comparisons between these cohorts and the national prevalence are less reliable. The increased prevalence we observed may simply be a representation of the referral bias to these centers. Patients with T1DM, a chronic condition requiring frequent interactions with physicians and nurses, may develop a greater sense of comfort or familiarity with the medical system. Moreover, as care tends to be followed at larger tertiary centers, which are more likely to have gender clinics and promote trans-friendly environments, gender-affirming care and exploration may be easier to facilitate. One would then expect that patients with T1DM would present to gender-affirming care earlier than the population average; however, in our study the TGD patients with T1DM presented later (16±2.7 years vs. 13±3.7 years, p=0.007) than the average in our clinic population.

In summary, we report in this study a higher prevalence of T1DM in TGD youth than would be expected. Future studies should investigate the impact of gender-affirming care and hormone therapy on diabetes control and the role of minority stress in the genesis of T1DM in this population. All patients had T1DM before presentation to the GeMS program except one, who was diagnosed within 2 weeks of their first appointment.

Abbreviations Used

BMI

body mass index

CDC

Centers for Disease Control

F

female

GAH

gender-affirming hormones

GeMS

Gender Multispecialty Service

GnRHa

gonadotropin-releasing hormone analog

HgbA1c

hemoglobin A1c

ICD

International Classification of Diseases

M

male

MDI

multiple daily injection

NB

nonbinary

Pump

insulin pump

SD

standard deviation

T1DM

type 1 diabetes mellitus

TGD

transgender/gender diverse

?

unknown

Author Disclosure Statement

The authors have no relevant financial conflicts of interest to disclose.

Funding Information

This study was supported by National Institutes of Health grant T32 DK007699 and Doris Duke Charitable Foundation grant 2019119.

Cite this article as: Maru J, Millington K, Carswell J (2021) Greater than expected prevalence of type 1 diabetes mellitus found in an urban gender program, Transgender Health 6:1, 57–60, DOI: 10.1089/trgh.2020.0027.