Clinical Characteristics and Comorbidities Associated With Testosterone Prescribing in Men

Aditi Sharma; Zia Ul-Haq; Emad Sindi; Ahmed Al-Sharefi; Tahereh Kamalati; Waljit S. Dhillo; Suks Minhas; Channa N. Jayasena

Disclosures

Clin Endocrinol. 2022;96(2):227-235.

Abstract and Introduction

Abstract

Objective: Testosterone replacement therapy (TRT) is recommended for the treatment of symptomatic hypogonadism in men. Data on prescription behaviours are, however, limited and conflicting. The objective of this study was to investigate clinical characteristics associated with the likelihood of being prescribed TRT by general practitioners (GP) in North-West London (NWL).

Design: Retrospective cohort study using Discover database of GP-registered patients in NWL between 2015 and 2019.

Patients: We identified 20,299 men aged ≥18 years with serum total testosterone measurement (TT) and without prior TRT prescription records.

Measurements: We determined whether TRT was subsequently commenced, while analysing clinical characteristics related to hypogonadism.

Results: Of all men having TT measurement, 19,583 (96.4%) were not commenced on TRT (Group A) and 716 (3.5%) men were commenced on TRT (Group B). Men prescribed TRT (Group B) had higher mean age, body mass index (BMI) and higher risks of hypertension, depression type 2 diabetes and ischaemic heart disease; conversely, men in Group B had lower mean pretreatment TT and were less likely to have prostate cancer. Four-hundred and thirty-six men (24.3%) with TT < 8 nmol/L and symptoms of low libido were not prescribed TRT.

Conclusions: Our study highlights several factors which may influence the decisions made by clinicians when initiating TRT in primary care. Clearer guidance for clinicians may help to improve the consistency of treatment of men with hypogonadism.

Introduction

Testosterone is an essential hormone for sexual function, muscle growth and bone mineralisation, and has important behavioural effects in men. Hypogonadism (symptomatic low testosterone) results in significant morbidity and adversely affects the quality of life. Incidence of hypogonadism increases with advancing male age such that levels of circulating testosterone decline annually by approximately 1% from the age of 40 years according to the European Male Ageing Study.^[1] This has contributed to a worldwide increase in the prescription of testosterone therapy.^[2]

The benefits of testosterone replacement therapy (TRT) including increased libido, improved bone density, strength and muscle mass are well-documented. Some studies have reported an increased risk of myocardial infarction and stroke associated with testosterone use.^[3,4] In contrast, other studies have observed that TRT was associated with either lower or unchanged risk of cardiovascular disease in men with hypogonadism.^[5] Furthermore, there exist several clinical guidelines for TRT^[6–9] each with differing advice which may create a lack of clarity for practitioners treating men with hypogonadism. Taken together, clinicians face a myriad of conflicting considerations when making the decision whether to initiate TRT in men. It is likely that there exists substantial variability in the prescribing habits of clinicians with regard to TRT for men with symptoms of hypogonadism.^[10]

To better understand the factors associated with prescribing (or not prescribing) TRT, we analysed a primary healthcare database to profile the clinical characteristics and comorbidities of men who had at least one serum total testosterone level (TT) measured from a population of over 2.5 million patients registered with a general practitioner (GP) in North-West London (NWL), UK.

1 2 3 4

Next Section

Abstract and Introduction
Materials and Methods
Results
Discussion
References

Table 1. Characteristics of patients in Group A and B
	No. (%) of patients
	Group A (n = 19,583)	Group B (n = 716)	OR^a (95% CI)	p-Value
Age (mean ± SD), year	47 ± 15.86	53 ± 16.29		<.0001
BMI (mean ± SD), kg/m²	27.5 ± 5.56	29.35 ± 6.07		<.0001
Total testosterone (mean ± SD), nmol/L	16 ± 7.41	11 ± 10.05		<.0001
Haemoglobin (mean ± SD) g/L	147.58 ± 14.12	147.28 ± 15.16		.58
Haematocrit (mean ± SD), L/L	0.48 ± 1.14	0.51 ± 1.55		.50
Glucocorticoids	1560 (7.97%)	134 (18.72%)	2.66 (2.19–3.22)	<.0001
Opioids	1022 (5.22%)	96 (13.41%)	2.81 (2.24–3.52)	<.0001
Antipsychotics	632 (3.23%)	44 (6.15%)	1.96 (1.44–2.67)	.0001
Antiresorptives	191 (0.98%)	16 (2.23%)	2.32 (1.37–3.84)	.0034
Smoking	8331 (42.54%)	331 (46.23%)	1.16 (1.00–1.35)	.05
Low libido	5359 (27.37%)	133 (18.58%)	0.61 (0.50–0.73)	<.0001
HTN	3479 (17.77%)	184 (25.70%)	1.60 (1.35–1.90)	<.0001
Depression	2730 (13.94%)	151 (21.09%)	1.65 (1.37–1.98)	<.0001
Type 2 diabetes	2335 (11.92%)	177 (24.72%)	2.43 (2.06–2.89)	<.0001
IHD	1043 (5.33%)	58 (8.10%)	1.57 (1.19–2.06)	.0024
Prostate disease	1070 (5.46%)	43 (6.01%)	1.11 (0.81–1.51)	.50
Fractures	925 (4.72%)	36 (5.03%)	1.07 (0.76–1.50)	.72
Anaemia	748 (3.82%)	40 (5.59%)	1.49 (1.08–2.06)	.0228
Sleep apnoea	494 (2.52%)	59 (8.24%)	3.47 (2.62–4.59)	<.0001
Prostate cancer	394 (2.01%)	8 (1.12%)	0.55 (0.23–1.09)	.10
Pituitary disease	264 (1.35%)	89 (12.43%)	10.39 (8.04–13.38)	<.0001
Erectile dysfunction	375 (1.91%)	8 (1.12%)	0.58 (0.28–1.15)	.16
DVT/PE	248 (1.27%)	20 (2.79%)	2.24 (1.42–3.54)	.0021
Stroke	228 (1.16%)	15 (2.09%)	1.82 (1.08–3.05)	.0337
Osteoporosis	165 (0.84%)	20 (2.79%)	3.38 (2.11–5.35)	<.0001
Sweating	132 (0.67%)	10 (1.40%)	2.09 (1.10–3.94)	.0351
Testicular cancer	34 (0.17%)	8 (1.12%)	6.50 (3.05–13.84)	<.0001
Delayed puberty	13 (0.07%)	4 (0.56%)	8.46 (2.89–24.70)	.0025
Cryptorchidism	7 (0.04%)	3 (0.42%)	11.77 (3.30–41.70)	.0044
Mumps	6 (0.03%)	0 (0.00%)	b	.64

Note: SI conversion factors: to convert testosterone from nmol/L to ng/dl, divide by 0.0347.
Abbreviations: BMI, body mass index; CI, confidence interval; DVT/PE, deep venous thrombosis/pulmonary embolism; HTN, hypertension; IHD, ischaemic heart disease; OR, odds ratio; SD, standard deviation.
^aOR unadjusted odds ratio calculated as odds group B/odds group A, CI confidence interval. Continuous variables such as age, BMI, total testosterone presented as mean ± SD.
^bUnable to calculate odds ratios due to no occurrences in one of the two groups.

Table 2. Characteristics of patients with serum testosterone below 8 nmol/L in both groups
	No. (%) of patients
	Group A (n = 1796)	Group B (n = 344)	OR^a (95% CI)	p-Value
Haemoglobin (mean ± SD), g/L	141.56 ± 16.61	145.67 ± 15.07		.0001
Haematocrit (mean ± SD), L/L	0.46 ± 0.97	0.57 ± 2.24		.14
Glucocorticoids	241 (13.42%)	67 (19.48%)	1.56 (0.48–0.86)	.0043
Opioids	163 (9.08%)	47 (13.66%)	1.59 (1.13–2.25)	.013
Antipsychotics	97 (5.40%)	23 (6.69%)	1.26 (0.79–1.99)	.37
Antiresorptives	45 (2.51%)	7 (2.03%)	0.81 (0.35–1.78)	.71
Smoking	803 (44.71%)	147 (42.73%)	0.92 (0.73–1.16)	.52
Low libido	436 (24.28%)	59(17.15%)	0.65 (0.48–0.87)	.004
HTN	544 (30.29%)	92 (26.74%)	0.84 (0.65–1.09)	.20
Depression	280 (15.59%)	80 (23.26%)	1.64 (1.24–2.16)	.0009
Type 2 diabetes	451 (25.11%)	94 (27.33%)	1.12 (0.87–1.46)	.38
IHD	172 (9.58%)	31 (9.01%)	0.94 (0.62–1.39)	.84
Prostate disease	127 (7.07%)	17 (4.94%)	0.68 (0.38–1.12)	.16
Fractures	118 (6.57%)	15 (4.36%)	0.65 (0.38–1.12)	.14
Anaemia	145 (8.07%)	21 (6.10%)	0.74 (0.45–1.19)	.23
Sleep apnoea	87 (4.84%)	30 (8.72%)	1.88 (1.22–2.87)	.006
Prostate cancer	145 (8.07%)	3 (0.87%)	0.10 (0.03–0.30)	<.0001
Pituitary disease	70 (3.90%)	48 (13.95%)	4.00 (2.72–5.86)	<.0001
Erectile dysfunction	27 (1.50%)	6 (1.74%)	1.16 (0.49–2.77)	.74
DVT/PE	47 (2.62%)	11 (3.20%)	1.23 (0.64–2.32)	.59
Stroke	30 (1.67%)	7 (2.03%)	1.22 (0.51–2.71)	.65
Osteoporosis	39 (2.17%)	8 (2.33%)	1.07 (0.51–2.25)	.84
Sweating	14 (0.78%)	9 (2.62%)	3.42 (1.41–7.70)	.007
Testicular cancer	6 (0.33%)	3 (0.87%)	2.63 (0.72–9.63)	.16
Delayed puberty	8 (0.45%)	3 (0.87%)	1.97 (0.56–6.54)	.40
Cryptorchidism	2 (0.11%)	3 (0.87%)	7.89 (1.57–infinity)	.004
Mumps	2 (0.11%)	0 (0.0%)	b	.54

Abbreviations: CI, confidence interval; DVT/PE (deep venous thrombosis/pulmonary embolism; HTN, hypertension; IHD, ischaemic heart disease; OR, odds ratio; SD, standard deviation.
^aUnadjusted odds ratio calculated as odds group B/odds group A.
^bUnable to calculate odds ratios due to no occurrences in one of the two groups.

Table 3. Changes in comorbidities before and after index testosterone measurement for Group A, and before and after TRT for Group B, in men with total serum testosterone below 8 nmol/L
	No. (%) of patients			No. (%) of patients
	Group A (n = 1796)		p-Value	Group B (n = 344)		p-Value
	Before	After		Before	After
Smoking	598 (33.30%)	580 (32.29%)	.5457	123 (35.76%)	115 (33.43%)	.3033
Low libido	316 (17.59%)	218 (12.14%)	<.0001	47 (13.70%)	19 (5.52%)	.0004
HTN	308 (17.15%)	403 (22.44%)	<.0001	57 (16.57%)	61 (17.73%)	.7617
Depression	141 (7.85%)	183 (10.19%)	.0168	58 (16.86%)	47 (13.66%)	.289
Type 2 diabetes	249 (13.86%)	377 (20.99%)	<.0001	63 (18.31%)	75 (21.80%)	.295
IHD	78 (4.34%)	134 (7.46%)	<.0001	17 (4.94%)	23 (6.69%)	.4156
Prostate disease	62 (3.45%)	80 (4.45%)	.1452	7 (2.03%)	13 (3.78%)	.2561
Fractures	66 (3.67%)	64 (3.56%)	.9289	7 (2.03%)	8 (2.33%)	>.9999
Anaemia	68 (3.79%)	97 (5.4%)	.0254	13 (3.78%)	11 (3.20%)	.8359
Sleep apnoea	35 (1.95%)	70 (3.90%)	.0007	21 (6.10%)	17 (4.94%)	.6171
Prostate cancer	115 (6.4%)	82 (4.57%)	.0156	3 (0.87%)	1 (0.29%)	.6239
Pituitary disease	17 (0.95%)	61 (3.4%)	<.0001	37 (10.76%)	29 (8.43%)	.3649
Erectile dysfunction	17 (0.95%)	12 (0.67%)	.4564	4 (1.16%)	2 (0.58%)	.6861
DVT/PE	25 (1.39%)	29 (1.61%)	.6812	5 (1.45%)	7 (2.03%)	.7724
Stroke	14 (0.78%)	19 (1.06%)	.4848	3 (0.87%)	5 (1.45%)	.725
Osteoporosis	20 (1.11%)	30 (1.67%)	.1995	4 (1.16%)	5 (1.45%)	>.9999
Sweating	9 (0.50%)	5 (0.28%)	.423	6 (1.74%)	4 (1.16%)	.7521
Testicular cancer	5 (0.28%)	2 (0.11%)	.0623	3 (0.87%)	No data	–
Delayed puberty	2 (0.11%)	7 (0.39%)	.1791	4 (1.16%)	No data	–
Cryptorchidism	0 (0.00%)	1 (0.06%)	>.9999	1 (0.29%)	2 (0.58%)	>.9999
Mumps	0 (0.00%)	1 (0.06%)	>.9999	0 (0.00%)	0 (0.00%)	>.9999

Abbreviations: DVT/PE, deep venous thrombosis/pulmonary embolism; HTN, hypertension; IHD, ischaemic heart disease; TRT, testosterone replacement therapy.