DIABETES TECHNOLOGY & THERAPEUTICS

Volume 26, Number 5, 2024

Mary Ann Liebert, Inc.
DOI: 10.1089/dia.2023.0491

Open camera or QR reader and

scan code to access this article
and other resources online.

REVIEW ARTICLE

Relationship Between Lipohypertrophy,

Glycemic Control, and Insulin Dosing:
A Systematic Meta-Analysis
Julia K. Mader, MD,1 Ricardo Fornengo, MD,2 Ahmed Hassoun, MD,3 Lutz Heinemann, PhD,4
Bernhard Kulzer, PhD,5 Magdalena Monica, MS,6,7 Trung Nguyen, PharmD,8
Jochen Sieber, PhD,9 Eric Renard, PhD,10 Yves Reznik, MD,11 Przemysław Ryś, PhD,6
Anita Stoz_ ek-Tutro, MS,6,7 and Emma G. Wilmot, PhD12

Abstract
Background: Lipohypertrophy is a common complication in patients with diabetes receiving insulin therapy.
There is a lack of consensus regarding how much lipohypertrophy affects diabetes management. Our study
aimed to assess the potential correlation between lipohypertrophy and glycemic control, as well as insulin
dosing in patients with diabetes.
Methods: We performed a systematic review followed by a meta-analysis to collect data about glycemic control and
insulin dosing in diabetic patients with and without lipohypertrophy. To identify relevant studies published in
English, we searched medical databases (MEDLINE/PubMed, Embase, and CENTRAL) from 1990 to January 20,
2023. An additional hand-search of references was performed to retrieve publications not indexed in medical
databases. Results of meta-analyses were presented either as prevalence odds ratios (pORs) or mean differences
(MDs) with 95% confidence intervals (95% CIs). This study was registered on PROSPERO (CRD42023393103).
Results: Of the 5540 records and 240 full-text articles screened, 37 studies fulfilled the prespecified inclusion
criteria. Performed meta-analyses showed that patients with lipohypertrophy compared with those without
lipohypertrophy were more likely to experience unexplained hypoglycemia (pOR [95% CI] = 6.98 [3.30–
14.77]), overall hypoglycemia (pOR [95% CI] = 6.65 [1.37–32.36]), and glycemic variability (pOR [95%
1
Division of Endocrinology and Diabetology, Department of Internal Medicine, Medical University of Graz, Graz, Austria.
2
Dipartimento di Area Medica, ASL TO4 S.S.D. di Diabetologia, Chivasso, Italy.
3
Department of Medicine, Fakeeh University Hospital, Dubai, United Arab Emirates.
4
Science Consulting in Diabetes GmbH, Kaarst, Germany.
5
Research Institute Diabetes Academy Bad Mergentheim (FIDAM), Diabetes Center Bad Mergentheim, Bad Mergentheim, Germany.
6
HTA Consulting, Cracow, Poland.
7
Doctoral School of Medical and Health Sciences, Jagiellonian University Medical College, Kraków, Poland.
8
embecta, Eysins, Switzerland.
9
embecta, Heidelberg, Germany.
10
Montpellier University Hospital and University of Montpellier, Montpellier, France.
11
Endocrinology and Diabetes Department, CHU Côte de Nacre, Caen Cedex, France.
12
Department of Translational Medical Sciences, University of Nottingham, Nottingham, United Kingdom.

ª Julia K. Mader, et al., 2024; Published by Mary Ann Liebert, Inc. This Open Access article is distributed under the terms of the
Creative Commons License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly credited.

351
352 MADER ET AL.

CI] = 5.24 [2.68–10.23]). Patients with lipohypertrophy also had higher HbA1c (MD [95% CI] = 0.55 [0.23–
0.87] %), and increased daily insulin consumption (MD [95% CI] = 7.68 IU [5.31–10.06]).
Conclusions: These results suggest that overall glycemic control is worse in patients with lipohypertrophy than
in those without this condition.

Keywords: Diabetes, Glycemic Control, Insulin-injection technique, Lipohypertrophy, Meta-analysis,

Systematic review.

Introduction Previous studies reported an increased risk of uncontrolled

glycemia, glycemic variability, and episodes of unexplained
hypoglycemia in patients with lipohypertrophy.2,4,6 For exam-
L ipohypertrophy is a common complication in patients
with diabetes treated with insulin therapy.1 Several risk
factors for developing lipohypertrophy among insulin-
ple, Gentile et al. found that 46.2% of patients with lipohy-
pertrophy experienced one or more episodes of hypoglycemia
injecting patients have been considered, including lack of compared with 6.8% of patients without lipohypertrophy.16 In
systemic rotation,2–4 needle reuse,2,4,5 needle length,3,6,7 and contrast, other studies found no such association.21–23 Ac-
number of daily injections.8 The condition is primarily cording to Kamrul-Hasan et al., the prevalence of hypoglyce-
characterized by the enlargement of adipocytes that mani- mia was comparable among patients with and without
fests by nodular swelling and the induration of fat tissue lipohypertrophy.18 Hence, clarifying the possible link between
around the injection sites.9 the presence of lipohypertrophy and glycemic control is needed.
In clinical practice, lipohypertrophy is usually diagnosed Our research aimed to critically evaluate and explain the
by physical examination, that is, visually and by palpation, potential relationship between lipohypertrophy and outcomes
and the most common presentation of lipohypertrophic related to glycemic control (e.g., hypoglycemia events,
nodules are those of large visible and esthetically displeasing HbA1c, and glycemic variability) and insulin dosing. We
mounds. However, increasing evidence suggests lipohyper- performed a systematic literature review followed by a meta-
trophic nodules exist in various forms, many of which are not analysis to synthesize the current knowledge of this important
easily visible or detectable by palpation.10 clinical issue. We hope our results will provide clinicians with
As physical examination methods vary between countries, additional evidence-based information for the best manage-
as evident in the lack of uniformity in the approaches to visual ment of diabetic patients and help identify critical knowledge
and palpation examination methods, new methods of lipohy- gaps and further directions for research in this area.
pertrophy detection have emerged, including ultrasonographic
skin scanning.11,12 A recent meta-analysis, based on data from Materials and Methods
26,865 patients, showed that lipohypertrophy is a common
Search strategy and selection criteria
health problem with a worldwide prevalence of 41.8% (95%
CI: 35.9%–47.6%) among patients with diabetes.13 However, This systematic review with a meta-analysis followed the
when considering studies that specifically utilized ultrasound Preferred Reporting Items for Systematic Review and Meta-
sonography, these figures can rise as high as 86.5%, suggesting Analyses (PRISMA) 2020 guidelines.24 We searched for
an underappreciated prevalence of this complication.14,15 studies reporting glycemic control (HbA1c, glycemic vari-
Published studies suggest that many insulin-treated patients ability, uncontrolled glycemia, or continuous glycemia
have significant deficiencies in their injection technique. They monitoring data), episodes of hypoglycemia (symptomatic,
often fail to ensure proper site rotation and show a preference asymptomatic, severe, unexplained, and overall), hypergly-
for injecting insulin into lipohypertrophic nodules, as these ar- cemia, and daily insulin doses in diabetic patients with li-
eas are less sensitive to pain.11 Available data indicate that pohypertrophy (LH+) and without lipohypertrophy (LH-)
insulin injections into lipohypertrophic areas may occur in up to who were treated with subcutaneous antihyperglycemic
95.3% of patients with diabetes receiving insulin therapy.16–18 therapy administered by pens or syringes.
Lack of thorough understanding of the possible conse- As therapy with glucagon-like peptide-1 (GLP-1) receptor
quences of lipohypertrophy may have an unaware impact on agonists has become more prevalent in recent years, we also
the efficacy of antihyperglycemic therapy in individual pa- sought to retrieve data for the group using these antidiabetic
tients. Driven by the high number of people living with dia- agents in addition to insulin-treated patients. We defined
betes and the high prevalence of lipohypertrophy, this glycemic variability as blood glucose oscillations <60 to
represents an unnoticed global health problem. >250 mg/dL at least three times a week or more than two
Aside from the apparent esthetic influence on patients’ well- unexplained glycemic fluctuations per week. Unexplained
being and self-image, pharmacological studies suggest that hypoglycemia was determined as hypoglycemic episodes
different structural properties of lipohypertrophic lesions may without a definable precipitating event, such as a change in
affect insulin absorption and metabolism.19,20 The insulin re- medication, diet, or activity. Uncontrolled glycemia included
lease from lipohypertrophic tissue is considered slower and the proportion of patients with HbA1c >7.0%.
more unpredictable than from normal fat tissue, which may We included randomized, observational, and cross-
result in excessive insulin dosing to achieve a pharmacological sectional studies published in English since 1990. We justi-
effect.9 However, available clinical evidence regarding the fied the publication date limitation because the standard of
possible relationship between the presence of lipohypertrophy diabetes care and insulin therapy had changed in the past
and glycemic control is contradictory. decades, which could have influenced the meta-analysis
LIPOHYPERTROPHY AND GLYCEMIC CONTROL META-ANALYSIS 353

results. Studies were excluded if they were conducted in event or as mean differences (MDs) for outcomes expressed
patients administering insulin by nonautomated pumps or by means and standard deviations. All results were given with
other sensor-augmented devices for continuous subcutaneous 95% confidence intervals (95% CIs). We used a random
infusion. Other exclusion criteria included studies describing model (DerSimonian and Laird) for data cumulation if sig-
patients treated mainly (‡80%) with animal insulins, studies nificant between-study heterogeneity was observed (P-value
published only as conference proceedings, and studies with for Cochrane Q test <0.10 and I2 > 50%). In other cases, a
data presentations unsuitable for cumulation. fixed model was chosen. If available, we also extracted
The systematic search was performed in MEDLINE P-values for comparisons reported by authors of the indi-
(through PubMed), Embase, and CENTRAL (through The vidual studies.
Cochrane Library) databases on January 20, 2023, using the We performed subgroup analyses to explore the effect of
keywords ‘‘diabetes’’ and ‘‘lipohypertrophy.’’ Detailed diabetes mellitus type, geographic region, duration of insulin
search strategies are provided in the Supplementary Appen- therapy, and a type of lipohypertrophy measurement on meta-
dix SA1 (Supplementary Tables S1–S3). A hand-search of analyses results. We also conducted sensitivity analyses,
references of the included studies was performed to retrieve including only studies published in the past 10 years, to de-
other relevant publications not indexed in searched medical termine if the publication date impacted meta-analyses re-
databases. sults. Subgroup and sensitivity analyses were performed only
If the same research group published more than one study, for outcomes, including at least 10 studies in the primary
we contacted the corresponding authors to ensure that studies meta-analyses. The risk of publication bias for meta-analysis
published by the same authors do not duplicate data for the of at least 10 studies was assessed by Eggers plots. For all
same patients. Two independent reviewers (A.S.-T. and statistical analyses, Sophie ver. 1.5.0 software was used
M.M.) selected the studies according to the protocol and (validated with STATA ver. 10.0).
predefined eligibility criteria (Table 1). Any disagreements The study was registered on the PROSPERO database
between reviewers on the full-text stage selection process (CRD42023393103).
were resolved by consensus.
Results
Data analysis
Of the 5540 records identified during databases and ref-
Two reviewers (A.S.-T. and M.M.) performed data ex- erences search, 240 full-text articles were assessed for eli-
traction independently. All discrepancies between reviewers gibility, of which 200 were excluded. Finally, 37 studies
were discussed and resolved. Extracted items included the described in 40 articles were included in the systematic meta-
design of studies, baseline population characteristics, details analysis (Table 2 and Supplementary Fig. S1). All excluded
of antihyperglycemic therapy, analyzed outcomes (HbA1c, studies with reasons are provided in the Supplementary Ap-
glycemic variability, uncontrolled glycemia, continuous glu- pendix SA1 (Supplementary Table S4).
cose monitoring data, hypo/hyperglycemia, and daily insulin Most of the included studies were cross-sectional
doses), and their definitions. The risk of bias was assessed (35 studies) and single-center (24 studies). Only two stud-
using Joanna Briggs Institute ( JBI) tools for cross-sectional25 ies17,55 were prospective and quasi-experimental. Five studies
and quasi-experimental26 studies. included only individuals with type 1 diabetes (T1DM), 11
We conducted meta-analyses comparing data for LH+ and with type 2 diabetes (T2DM), patients with either type of
LH- only if two or more studies reported the same outcome. diabetes participated in 20 studies, and 1 study did not report
Results of meta-analyses were presented either as prevalence information about diabetes type.46 Three studies2,21,47 focused
odds ratios (pORs) for the proportion of patients with an on the pediatric population.

Table 1. Inclusion and Exclusion Criteria

PICO Inclusion criteria Exclusion criteria
Population Patients with diabetes treated with insulin or GLP-1 Participants using insulin pumps exclusively
analogs administered by pens or syringes Most of the participants (‡80%) used insulins
other than humans and analogs
Intervention Presence of lipohypertrophy (lipodystrophy) Not applicable
(exposure)
Comparator Lack of lipohypertrophy (lipodystrophy) Not applicable
(control)
Outcomes Glycemic control (glycemic variability, HbA1c, and Other than defined
CGM data), episodes of hypoglycemia and Data presentation unsuitable for cumulation
hyperglycemia (symptomatic, asymptomatic, (e.g., continuous endpoints presented as
severe, and unexplained), and daily insulin dosage medians)
Study type Randomized clinical trials and observational studies Studies published only as conference abstracts or
(cohort, case–control, and cross-sectional) posters
published in English Studies published before 1990
Studies published in languages other than
English
CGM, Continuous glucose monitoring; GLP-1, glucagon-like peptide-1.
 Table 2. Summary of the Included Studies
Number of patients Duration of Duration of
Study Study design (location) (LH +/LH -) LH measurement Diabetes type Population diabetes insulin therapy
Abujbara et al.7 Cross-sectional, single-center 851 (477/374) Visual inspection and T1DM, T2DM Unknown 12.3 (8.1) 7.0 (6.2)
( Jordan) palpation
Al Ajlouni et al.28 Cross-sectional, single-center 1090 (407/683) Visual inspection and Only T2DM Only adults 13.5 (9–20)a 4.6 (5.0)
( Jordan) palpation
Al Hayek et al.2 Cross-sectional, single-center 174 (83/91) Visual inspection and Only T1DM Only children 6.1 (4.5) Unknown
(Saudi Arabia) palpation
AlJaber et al.29 Cross-sectional, multicenter 202 (80/122) Visual inspection and Only T2DM Only adults 16.9 (8.5) 8.5 (5.8)
(Saudi-Arabia) palpation
Arora et al.30 Cross-sectional, single-center 500 (290/210) Visual inspection and T1DM, T2DM Unknown Unknown 3.0 (2.5–5.0)a
(India) palpation, USG in all
patients
Barola et al.3 Cross-sectional, single-center 372 (231/141) Visual inspection and Only T1DM Children, adults 5.6 (5.3) Unknown
(India) palpation
Baruah et al.5 Cross-sectional, single-center 748 (94/654) Visual inspection and Only T2DM Children, adults 12.2 (7.6) 3.4 (4.2)
(India) palpation
Blanco et al.4 Cross-sectional, multicenter 430 (277/153) Visual inspection and T1DM, T2DM Children, adults NA (6–15)b NA (1–5)b
(Spain) palpation, USG in some
patients
Bochanen et al.17 Prospective, quasi- 146 (92/54) Visual inspection and T1DM, T2DM Only adults Unknown Unknown
experimental, multicenter palpation

354
(Belgium)
Cunningham and Cross-sectional, multicenter 55 (28/27) Visual inspection and T1DM, T2DM Unknown Unknown 15.0 (12.6)
McKenna31 (Ireland) palpation
Frid et al.32–34 Cross-sectional, multicenter 13,289 (3855/9344) Visual inspection and T1DM, T2DM Children, adults 13.2 (9.7) 8.7 (8.9)
(Worldwide) palpation
Gentile et al.35 Cross-sectional, multicenter 296 (169/127) Visual inspection and T1DM, T2DM Only adults 7.0 (2.0) 3.0 (1.0)
(Italy) palpation
Gentile et al.36 Cross-sectional, multicenter 1227 (718/509) Visual inspection and Only T2DM Only adults 10.6 (7.9) 7.6 (6.0)
(Italy) palpation, USG in all
patients
Gentile et al.6 Cross-sectional, multicenter 780 (360/420) Visual inspection and T1DM, T2DM Only adults 18.0 (11.0) 10.1 (2.1)
(Italy) palpation, USG in all
patients
Gentile et al.16 Prospective, quasi- 1160 (487/673) Visual inspection and Only T2DM Only adults 15.8 (7.6) 7.6 (2.2)
experimental, multicenter palpation, USG in some
(Italy) patients
Gunhan et al.37 Cross-sectional, single-center 345 (98/247) Visual inspection and Only T2DM Only adults 37.0 (8.5) 8.9 (5.7)
(Turkey) palpation
38
Gupta et al. Cross-sectional, single-center 139 (97/42) Visual inspection and Only T2DM Only adults 8.7 (7.5) Unknown
(India) palpation
39
Hajheydari et al. Cross-sectional, single-center 230 (35/185) Visual inspection and T1DM, T2DM Unknown 14.0 (8.5) 5.4 (6.0)
(Iran) palpation
Hauner et al.40 Cross-sectional, single-center 279 (66/213) Visual inspection and T1DM, T2DM Unknown 14.1 (9.5) Unknown
(Germany) palpation
(continued)
 Table 2. (Continued)
Number of patients Duration of Duration of
Study Study design (location) (LH +/LH -) LH measurement Diabetes type Population diabetes insulin therapy
Ji and Lou41 Cross-sectional, multicenter 380 (134/246) Visual inspection and Only T2DM Only adults Unknown 3.6 (4.1)
(China) palpation
Ji et al.23,42 Cross-sectional, multicenter 401 (213/188) Visual inspection and T1DM, T2DM Only adults 11.8 (7.3) 5.8 (4.5)
(China) palpation
Kamrul-Hasan et al.18 Cross-sectional, multicenter 847 (78/769) Unknown T1DM, T2DM Unknown 9.8 (7.0) 3.8 (4.1)
(China)
Korkmaz et al.43 Cross-sectional, single-center 136 (119/17) Only USG T1DM, T2DM Only adults 15.8 (9.2) 11.4 (8.3)
(Turkey)
Kumar et al.44 Cross-sectional, single-center 88 (60/28) Visual inspection and T1DM, T2DM Only adults 26.4 (5.1) 6.5 (6.6)
(India) palpation, USG in all
patients
Lin et al.8 Cross-sectional, single-center 120 (83/37) Visual inspection and Only T2DM Only adults Unknown 6.6 (4.3)
(China) palpation, USG in all
patients
Luo et al.45 Cross-sectional, single-center 316 (270/46) Visual inspection and T1DM, T2DM Only adults 12.8 (16.6–19.2)a 6.2 (2.9–10.4)a
(China) palpation, USG in some
patients
Nawaz et al.46 Cross-sectional, single-center 363 (83/280) Unknown Unknown Children, adults 7.9 (4.5) 6.5 (3.7)
(Pakistan)
Omar et al.47 Cross-sectional, single-center 119 (62/51) Visual inspection and Only T1DM Only children Unknown Unknown

355
(Egypt) palpation
Pahuja et al.48 Cross-sectional, single-center 96 (65/31) Visual inspection and Only T2DM Only adults 19.8 (NA) 6.8 (NA)
(India) palpation
Pozzuoli et al.22 Cross-sectional, single-center 352 (151/201) Visual inspection and T1DM, T2DM Unknown 20.4 (9.9) 9.1 (8.6)
(Italy) palpation
Saeed et al.49 Cross-sectional, single-center 360 (157/203) Visual inspection and T1DM, T2DM Unknown 14.7 (7.6) 8.5 (6.1)
(Pakistan) palpation
Saez de Ibarra and Cross-sectional, single-center 150 (78/72) Visual inspection and T1DM, T2DM Unknown 13.0 (9.0) 11.4 (7.9)
Gallego50 (Spain) palpation
Singha et al.51 Cross-sectional, single-center 95 (46/45) Visual inspection and Only T1DM Unknown Unknown Unknown
(India) palpation, USG in all
patients
Strollo et al.52 Cross-sectional, multicenter 387 (298/98) Visual inspection and T1DM, T2DM Only adults 13.0 (9.0) 10.0 (9.0)
(Italy) palpation
Arda Sürücü and Cross-sectional, single-center 436 (191/245) Visual inspection and Only T2DM Only adults Unknown Unknown
OKurArslan53 (Turkey) palpation
Thewjitcharoen et al.54 Cross-sectional, single-center 400 (149/251) Visual inspection and T1DM, T2DM Unknown 23.0 (10.2) 11.4 (8.7)
(Thailand) palpation, USG in all
patients
Tsadik et al.21 Cross-sectional, single-center 176 (103/73) Visual inspection and Only T1DM Only children Unknown Unknown
(Ethiopia) palpation
Continuous data are given as mean (SD) unless otherwise stated.
a
Median (IQR).
b
Median (range).
IQR, interquartile range; LH, lipohypertrophy; NA, not available; SD, standard deviation; T1DM, type 1 diabetes; T2DM, type 2 diabetes; USG, ultrasonography.
356 MADER ET AL.

FIG. 1. Forest plot for unexplained hypoglycemia.

The size of the population in the included studies varied mon among the lipohypertrophy group (pOR [95% CI] = 2.77
from 5531 to 13,2899 participants. The quality of the research [1.62–4.73]; Fig. 4).
was diverse based on JBI scales (4–8/8 points for cross- The presence of lipohypertrophy was also associated with
sectional studies and 5–6/9 points for quasi-experimental), a higher prevalence of glycemic variability among patients
although no study was excluded from meta-analyses due to with diabetes (pOR [95% CI] = 5.24 [2.68–10.23]; Fig. 5).
the high risk of bias. Detailed characteristics of the included Mean values of glycemic variability based on only two
studies are presented in the Supplementary Appendix SA1 studies35,36 were higher in the lipohypertrophy group com-
(Supplementary Tables S5–S17). pared with the no lipohypertrophy group (MD [95%
The primary analysis showed that patients with lipohy- CI] = 100.20 [93.70–106.69] mg/dL).
pertrophy were more likely to experience unexplained hypo- Patients with lipohypertrophy were treated with higher
glycemia (pOR [95% CI] = 6.98 [3.30–14.77]; Fig. 1) and insulin doses compared with those without lipohypertrophy
overall hypoglycemia (pOR [95% CI] = 6.65 [1.37–32.36]; (MD [95% CI] = 7.68 IU [5.31–10.06]; Fig. 6). The differ-
Fig. 2) compared with patients without lipohypertrophy. No ence remained significant even if insulin doses were adjusted
between-group difference was found regarding symptomatic to the individuals’ body weight (MD [95% CI] = 0.06 [0.01–
and severe hypoglycemia. Data for other endpoints related to 0.12] IU/kg). Only a few identified studies reported data for
hypoglycemia were presented only in individual studies, and hyperglycemia and continuous glucose monitoring. Results
performing meta-analyses for these outcomes was impossible. for these endpoints are given in the Supplementary Appendix
Patients with lipohypertrophy also had significantly higher SA1 (Supplementary Tables S35 and S36).
values of HbA1c than those without lipohypertrophy (MD Performed sensitivity analyses indicate that the publica-
[95% CI] = 0.55 [0.23–0.87] %; Fig. 3). Uncontrolled gly- tion date did not significantly impact meta-analysis results for
cemia, defined as HbA1c values >7%, was also more com- primary outcomes, including unexplained hypoglycemia,

FIG. 2. Forest plot for overall hypoglycemia.

LIPOHYPERTROPHY AND GLYCEMIC CONTROL META-ANALYSIS 357

FIG. 3. Forest plot for HbA1c.

HbA1c, and the total daily insulin dose (Supplementary were noticed for unexplained hypoglycemia and the total
Figs. S35–S37). However, only a few identified studies were daily insulin dose. Similarly, inconsistent results were re-
published before 2014. ported for the impact of the diabetes duration. All subgroup
Based on subgroup analyses, we noticed that the impact of analyses are presented in the Supplementary Appendix SA1
lipohypertrophy on the prevalence of unexplained hypogly- (Supplementary Figs. S14–S34).
cemia and uncontrolled glycemia was slightly more pro- No publication bias was identified for analyzed outcomes
nounced in individuals with T2DM than in those with T1DM; (Supplementary Figs. S38–S39).
however, no significant interaction between subgroups was
observed (P > 0.05). In contrast, we reported a significant
Discussion
interaction for a greater impact of lipohypertrophy on the
total daily insulin dose in the T2DM subgroup compared with Our systematic review identified 37 studies comparing
T1DM (P = 0.013). glycemic control parameters and insulin dosing in patients
We also found that in studies reporting lipohypertrophy with and without lipohypertrophy. We performed meta-
measured by ultrasonography, the association with unex- analyses with a satisfactory number of studies only for four
plained hypoglycemia, HbA1c values, and total daily insulin outcomes (HbA1c, uncontrolled glycemia, unexplained hy-
dose was more substantial than in those with only clinical poglycemia, and total daily insulin dose). Other endpoints
assessment of lipohypertrophy (P < 0.05). Regarding the stated in the protocol (e.g., hyperglycemia and continuous
geographical region, random interactions resulting from the glucose monitoring) were available in only a few studies,
imbalance in the number of studies in particular subgroups limiting the possibility of conducting a reliable meta-
358 MADER ET AL.

FIG. 4. Forest plot for glycemic variability.

analysis. Nonetheless, we were able to collect sufficient data poor glycemic control, and higher doses of insulin. However,
to determine the possible relationship between lipohyper- available data from studies evaluating the impact of educa-
trophy and glycemic control. tional programs on proper insulin injection techniques, with
Our results showed that all primary outcomes regarding avoidance of injections into lipohypertrophy areas, indicate
glycemic control were significantly worse in patients with the direct involvement of lipohypertrophy in worsening
lipohypertrophy than those without lipohypertrophy. Epi- glycemic control and excessive insulin dosing.
sodes of unexplained hypoglycemia, uncontrolled glycemia, Wang et al.56 reported that a 3-month intensive training on
and glycemic variability were more prevalent in patients with insulin injection technique in patients with lipohypertrophy
lipohypertrophy than in a control group. In addition, those resulted in a significant and clinically relevant decrease of
with confirmed lipohypertrophy also used higher insulin mean HbA1c by 0.60%, fasting plasma glucose by
doses. Although these results suggest that lipohypertrophy is 1.20 mmol/L, 2h postprandial plasma glucose by 1.70 mmol/L
associated with poorer glycemic control and higher insulin without increasing the insulin dosage. Indicators of glycemic
doses, we cannot draw an unequivocal causal conclusion. variability, hyperglycemic and hypoglycemic events were
Nearly all of the included studies were designed as cross- also markedly decreased. These meaningful results were
sectional without any follow-up. further confirmed by a randomized controlled trial conducted
Therefore, we cannot rule out that there are other causal by the AMD-OSDI Study Group,57 in which 318 patients with
factors affecting both the development of lipohypertrophy, lipohypertrophy were assigned either to the intervention

FIG. 5. Forest plot for uncontrolled glycemia (HbA1c >7%).

LIPOHYPERTROPHY AND GLYCEMIC CONTROL META-ANALYSIS 359

FIG. 6. Forest plot for daily insulin dose.

group receiving appropriate injection technique education or reported outcomes. In many studies, comprehensive infor-
to the control group without education. mation on the research methodology and statistical analysis
After a 6-month follow-up, HbA1c values, glycemic var- assumptions was not provided. Errors in reporting patient
iability, and episodes of severe and symptomatic hypogly- numbers and events also occurred,28,52 which made it im-
cemia were significantly decreased in the intervention group possible to cumulate these data for specific endpoints. An-
compared with the control group. In addition, insulin doses in other issue observed was a high heterogeneity for almost all
the intervention group decreased by nearly 21%, suggesting analyzed outcomes (I2 > 80%).
that improvement in injection technique and its impact on We could not establish the source of heterogeneity since
lipohypertrophy allows for a reduction in insulin consump- the studies included in the meta-analysis were diverse in
tion. Notably, the benefits associated with educating patients many factors simultaneously. At the same time, based on
with lipohypertrophy on injection technique are not only aggregated data, we could only investigate one factor in
limited to improving health outcomes but also result in cost subgroup analyses. According to the protocol, we aimed to
savings in diabetic care, including insulin costs and treatment obtain a result on the effect of lipohypertrophy on glycemic
of diabetes complications.55 Obtained results indicate the control, regardless of the type of diabetes, patients’ age,
significance of the issue of lipohypertrophy and the necessity treatment history, and different diabetic management stan-
to adhere to FITTER guidelines regarding proper injection dards of care in various geographical regions.
technique, including the importance of avoiding injections in In addition, differences in the definition and diagnosis of
areas affected by lipohypertrophy, proper site rotation, and lipohypertrophy and analyzed outcomes between studies
needle single-use.9 could have influenced the high heterogeneity of results. For
Our research has some limitations that cannot be over- example, in some studies, the definition of lipohypertrophy
come, which result mainly from the low reliability of data also included patients with lipoatrophy, and the diagnosis
pooled in meta-analyses. The studies included in this meta- could be based on either ultrasonographic examination or
analysis exhibited varying risks of bias and quality of solely visually and by palpation. To eliminate this potential
360 MADER ET AL.

bias, if possible, we excluded data for patients with lipoa- aware that lipohypertrophy is not only a cosmetic issue but
trophy from the analysis. In other cases, patients with li- also a clinically relevant topic. Routine screening for lipo-
poatrophy constituted a small proportion of the entire hypertrophy and intensive patient education on the proper
lipohypertrophy group, so including their data should not insulin injection technique, including site rotation and needle
have significantly affected the meta-analysis results. single-use, may have a beneficial effect on better diabetes
We also performed subgroup analyses regarding the type control, insulin dosing, and prevention of long-term com-
of lipohypertrophy measurement. Interestingly, our results plications of the disease.
showed that the negative impact of lipohypertrophy on gly-
cemic control was markedly higher in those with lipohyper- Authors’ Contributions
trophy confirmed by ultrasound imaging compared with
A.S.-T. and M.M. performed a systematic search, data
those with clinical assessment alone. This result may suggest
extraction, and statistical analyses. All authors were involved
that patients with subclinical lipohypertrophy, often unaware
in protocol preparation, data interpretation, article writing,
of their condition, are particularly vulnerable to glycemic
and editing. All authors had full access to all the data in the
fluctuations due to insulin injections into lipohypertrophy
study and had final responsibility for the decision to submit
areas. At the same time, adequately educated patients with
for publication.
visible lipohypertrophy may avoid administering insulin into
lipohypertrophic nodules.
Author Disclosure Statement
Unfortunately, due to insufficient reporting in the studies,
no analysis considering the proportion of diabetic patients J.K.M. is a member of the advisory board of Abbott Dia-
with lipohypertrophy who injected insulin into affected areas betes Care, Becton-Dickinson/embecta, Boehringer In-
could be performed. Therefore, the observed heterogeneity in gelheim, Eli Lilly, embecta, Medtronic, Novo Nordisk A/S,
our meta-analyses may also result from the inability to con- Roche Diabetes Care, Sanofi-Aventis, and Viatris, and re-
sider the actual percentage of patients injecting insulin into ceived speaker honoraria from A. Menarini Diagnostics,
lipohypertrophy. Nonetheless, in light of these findings, it Abbott Diabetes Care, AstraZeneca, Becton-Dickinson/
may be worth considering the introduction of routine ultra- embecta, Boehringer Ingelheim, Dexcom, Eli Lilly, Med-
sonographic assessments in both clinical practice and trials as trust, MSD, Novo Nordisk A/S, Roche Diabetes Care, Sanofi,
a more sensitive diagnostic method. Servier, and Ypsomed. B.K. is a member of the advisory
Other subgroup analyses indicate that the impact of lipo- board of embecta. E.R. is a member of the advisory board of
hypertrophy on glycemic control was slightly more pro- Eli Lilly, Novo Nordisk, and received speaker honoraria from
nounced in T2DM than in those in T1DM. One possible embecta. J.S. is an employee and stockholder of embecta and
explanation is that patients with T1DM are often better ed- Becton&Dickinson. T.N. is an employee of embecta. L.H. is
ucated in injection technique and the negative consequences a consultant for several companies developing novel diag-
of lipohypertrophy due to the usual longer duration of insulin nostic and therapeutic options for diabetes treatment. He is a
therapy. In contrast, the frequently coexisting obesity in pa- shareholder of the Profil Institut für Stoffwechselforschung
tients with T2DM may affect the overlook of skin changes. GmbH, Neuss, Germany. Y.R. declares consultant/speaker
The pathophysiology of lipohypertrophy in patients with fees from Medtronic, Insulet, embecta, Abbott, Novo Nor-
diabetes has not been fully elucidated, but some authors disk, Eli-Lilly, Sanofi, and Air Liquide Santé International.
suggest that it may result not only from the lipogenic prop- A.H., A.S.-T., M.M., P.R., and R.F. declare no conflict of
erties of insulin, promoting the growth of fat cells but also interest. EGW is a member of the advisory board of Abbott
from mechanical damage to subcutaneous tissue through Diabetes Care, Eli Lilly, embecta, Insulet, Medtronic, Novo
repeated and improper injections in the same location.9 The Nordisk, Roche Diabetes Care, and Sanofi-Aventis. Research
rationale for including glucagon-like peptide 1 receptor support from Abbott Diabetes Care, embecta, Insulet, Novo
agonists (GLP-1-RAs) in the meta-analysis was based on the Nordisk, and Sanofi-Aventis, and has received speaker hon-
observation that some studies reported the occurrence of oraria from Abbott Diabetes Care, AstraZeneca, Dexcom, Eli
lipohypertrophy in patients using GLP-1-RA.58 Lilly, embecta, Glooko, Insulet, Medtronic, Novo Nordisk,
Despite a comprehensive literature review, we found no Sanofi, and Ypsomed.
scientific evidence regarding the potential link between li-
pohypertrophy and glycemic control in patients receiving Funding Information
GLP-1 analogs. Only one study included in the systematic
The study was sponsored by embecta. The sponsor had a
review9 involved patients using GLP-1 receptor agonists, but
role in the study design, interpreting data, writing the report,
they accounted for <2% of all study participants, and no sub-
and in the decision to submit the article for publication.
group analysis was available for them. Thus, there is a sig-
nificant evidence gap for this patient group, and further
Supplementary Material
research should focus on assessing the occurrence of lipohy-
pertrophy and its consequences in patients treated with other Supplementary Appendix SA1
than insulin subcutaneous antihyperglycemic medications.
References
Conclusions
1. Abu Ghazaleh H, Hashem R, Forbes A, et al. A systematic
The meta-analysis results indicate that lipohypertrophy is review of ultrasound-detected lipohypertrophy in insulin-
associated with poorer glycemic control and higher insulin exposed people with diabetes. Diabetes Ther 2018;9(5):
consumption. Clinicians and health care providers should be 1741–1756; doi: 10.1007/s13300-018-0472-7
LIPOHYPERTROPHY AND GLYCEMIC CONTROL META-ANALYSIS 361

2. Al Hayek AA, Robert AA, Braham RB, et al. Frequency of truth. Diabet Med 2022;39(1):e14672; doi: 10.1111/dme
lipohypertrophy and associated risk factors in young pa- .14672
tients with type 1 diabetes: A cross-sectional study. 18. Kamrul-Hasan A, Amin MN, Gaffar MAJ, et al. Insulin
Diabetes Ther 2016;7(2):259–267; doi: 10.1007/s13300- injection practice and injection complications—Results
016-0161-3 from the Bangladesh Insulin Injection Technique Survey.
3. Barola A, Tiwari P, Bhansali A, et al. Insulin-related li- Eur Endocrinol 2020;16(1):41–48; doi: 10.17925/EE.2020
pohypertrophy: Lipogenic action or tissue trauma? Front .16.1.41
Endocrinol 2018;9:638; doi: 10.3389/fendo.2018.00638 19. Johansson U-B, Amsberg S, Hannerz L, et al. Impaired
4. Blanco M, Hernández MT, Strauss KW, et al. Prevalence absorption of insulin aspart from lipohypertrophic injection
and risk factors of lipohypertrophy in insulin-injecting pa- sites. Diabetes Care 2005;28(8):2025–2027; doi: 10.2337/
tients with diabetes. Diabetes Metab 2013;39(5):445–453; diacare.28.8.2025
doi: 10.1016/j.diabet.2013.05.006 20. Famulla S, Hövelmann U, Fischer A, et al. Insulin injection
5. Baruah M, Kalra S, Bose S, et al. An audit of insulin usage into lipohypertrophic tissue: Blunted and more variable
and insulin injection practices in a large Indian cohort. insulin absorption and action and impaired postprandial
Indian J Endocr Metab 2017;21(3):443–452; doi: 10.4103/ glucose control. Diabetes Care 2016;39(9):1486–1492; doi:
ijem.IJEM_548_16 10.2337/dc16-0610
6. Gentile S, Guarino G; AMD-OSDI Study Group, et al. 21. Tsadik AG, Atey TM, Nedi T, et al. Effect of insulin-
Bruising: A neglected, though patient-relevant complica- induced lipodystrophy on glycemic control among children
tion of insulin injections coming to light from a real-life and adolescents with diabetes in Tikur Anbessa Specialized
nationwide survey. Diabetes Ther 2021;12(4):1143–1157; Hospital, Addis Ababa, Ethiopia. J Diabetes Res 2018;
doi: 10.1007/s13300-021-01026-w 2018:1–7; doi: 10.1155/2018/4910962
7. Elsayed S, Soliman AT, De Sanctis V, et al. Insulin- 22. Pozzuoli GM, Laudato M, Barone M, et al. Errors in insulin
induced lipodystrophy and predisposing factors in children treatment management and risk of lipohypertrophy. Acta
and adolescents with type 1 diabetes mellitus (T1DM) in a Diabetol 2018;55(1):67–73; doi: 10.1007/s00592-017-
tertiary care Egyptian center. Acta Biomed 2023;94(3): 1066-y
e2023078; doi: 10.23750/abm.v94i3.14117 23. Ji L, Sun Z, Li Q, et al. Lipohypertrophy in China: Pre-
8. Lin Y, Lin L, Wang W, et al. Insulin-related lipohyper- valence, risk factors, insulin consumption, and clinical
trophy: Ultrasound characteristics, risk factors, and impact impact. Diabetes Technol Ther 2017;19(1):61–67; doi: 10
of glucose fluctuations. Endocrine 2022;75(3):768–775; .1089/dia.2016.0334
doi: 10.1007/s12020-021-02904-w 24. Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA
9. Frid AH, Kreugel G, Grassi G, et al. New insulin delivery 2020 statement: An updated guideline for reporting sys-
recommendations. Mayo Clin Proc 2016;91(9):1231–1255; tematic reviews. BMJ 2021;372:n71; doi: 10.1136/bmj.n71
doi: 10.1016/j.mayocp.2016.06.010 25. Moola S, Munn Z, Tufanaru C, et al. Chapter 7: Systematic
10. Perciun R. Ultrasonographic aspect of subcutaneous tissue Reviews of Etiology and Risk. In: Joanna Briggs Institute
dystrophies as a result of insulin injections. Med Ultrason Reviewer’s Manual. (Aromataris E, Munn Z, eds.) JBI
2010;12(2):104–109. Manual for Evidence Synthesis. The Joanna Briggs In-
11. Xu X-H, Carvalho V, Wang X-H, et al. Lipohypertrophy: stitute: 2017. Available from: https://synthesisma-
Prevalence, clinical consequence, and pathogenesis. Chin nual.jbi.global; doi: 10.46658/JBIMES-20-08
Med J 2021;134(1):47–49; doi: 10.1097/CM9.0000000 26. Tufanaru C, Munn Z, Aromataris E, et al. Chapter 3:
000000970 Systematic Reviews of Etiology and Risk. In: Joanna
12. Gentile G, Strollo F. Factors hindering correct identifica- Briggs Institute Reviewer’s Manual. (Aromataris E,
tion of unapparent lipohypertrophy. J Diabetes Metab Munn Z, eds.) The Joanna Briggs Institute: 2017. Available
Disord Control 2016;3(2):42–47; doi: 10.15406/jdmdc from: https://synthesismanual.jbi.global; doi: 10.46658/
.2016.03.00065 JBIMES-20-04
13. Wang K, Zhang S, Liu C, et al. A meta-analysis and meta- 27. Abujbara M, Khreisat EA, Khader Y, et al. Effect of insulin
regression on the prevalence of lipohypertrophy in diabetic injection techniques on glycemic control among patients
patients on insulin therapy. Therapies 2021;76(6):617–628; with diabetes. Int J Gen Med 2022;15:8593–8602; doi: 10
doi: 10.1016/j.therap.2021.04.002 .2147/IJGM.S393597
14. Volkova NI, Davidenko IY. Clinical significance of lipo- 28. Al Ajlouni M, Abujbara M, Batieha A, et al. Prevalence of
hypertrophy without visual and palpable changes detected lipohypertrophy and associated risk factors in insulin-treated
by ultrasonography of subcutaneous fat. Ter Arkh 2019; patients with type 2 diabetes mellitus. Int J Endocrinol Metab
91(4):62–66; doi: 10.26442/00403660.2019.04.000128 2015;13(2):e20776; doi: 10.5812/ijem.20776
15. Volkova N, Davidenko I, Rudakova J, et al. Insulin-induced 29. AlJaber AN, Sales I, Almigbal TH, et al. The prevalence of
lipohypertrophy diagnostics in diabetic patients: Sub- lipohypertrophy and its associated factors among Saudi
cutaneous fat ultrasonography. Endocrine Abstracts 2015; patients with type 2 diabetes mellitus. J Taibah Univ Med
37 EP351; doi: 10.1530/endoabs.37.EP351 Sci 2020;15(3):224–231; doi: 10.1016/j.jtumed.2020.03
16. Gentile S, Guarino G, Della Corte T, et al. Why do so many .006
people with type 2 diabetes who take insulin have lipohy- 30. Arora S, Agrawal NK, Shanthaiah DM, et al. Early detec-
pertrophy? Fate or educational deficiencies? Diabetes Ther tion of cutaneous complications of insulin therapy in type 1
2023;14(1):179–191; doi: 10.1007/s13300-022-01341-w and type 2 diabetes mellitus. Prim Care Diabetes 2021;
17. Bochanen N, Decochez K, Heleu E, et al. Lipohypertrophy 15(5):859–864; doi: 10.1016/j.pcd.2021.06.004
Monitoring Study (LIMO): Effect of single use of 4 mm 31. Cunningham MT, McKenna M. Lipohypertrophy in
pen needles combined with education on injection site ro- insulin-treated diabetes: Prevalence and associated risk
tation on glycaemic control: Confirmation of an unpleasant factors. J Diabetes Nurs 2013;17(9):340–343.
362 MADER ET AL.

32. Frid AH, Hirsch LJ, Menchior AR, et al. Worldwide in- associated factors. BMC Res Notes 2011;4(1):290; doi: 10
jection technique questionnaire study: Injecting complica- .1186/1756-0500-4-290
tions and the role of the professional. Mayo Clin Proc 2016; 48. Pahuja V, Punjot P, Fernandes G, et al. Exploring the
91(9):1224–1230; doi: 10.1016/j.mayocp.2016.06.012 factors associated with lipohypertrophy in insulin-treated
33. Frid AH, Hirsch LJ, Menchior AR, et al. Worldwide in- type 2 diabetes patients in a tertiary care hospital in
jection technique questionnaire study: Population parame- Mumbai, India. Int J Diabetes Dev Ctries 2019;39(3):426–
ters and injection practices. Mayo Clin Proc 2016;91(9): 431; doi: 10.1007/s13410-019-00735-0
1212–1223; doi: 10.1016/j.mayocp.2016.06.011 49. Saeed M, Aslam W, Ahmad M, et al. Frequency and risk
34. Anonymous. ITQ Survey Data Mayo. Available from: factors of lipohypertrophy in insulin treated patients with
https://public.tableau.com/profile/adam.yeung#!/vizhome/ diabetes: An experience from a tertiary care center in
ITQsurveydataMayo/START [Last accessed: March 2, 2020]. Multan. Med Forum Mon 2022;33(8):78–82.
35. Gentile S, Strollo F, Satta E, et al. Insulin-related lipohyper- 50. Saez de Ibarra L, Gallego F. Factors related to lipohyper-
trophy in hemodialyzed diabetic people: A multicenter obser- trophy in insulinltreated diabetic patients: Role of educa-
vational study and a methodological approach. Diabetes Ther tional intervention. Pract Diabetes Int 1998;15(1):9–11.
2019;10(4):1423–1433; doi: 10.1007/s13300-019-0650-2 51. Singha A, Bhattacharjee R, Dalal BS, et al. Associations of
36. Gentile S, Guarino G; the AMD-OSDI Study Group on insulin-induced lipodystrophy in children, adolescents, and
Injection Techniques and Nefrocenter Research and Nyx young adults with type 1 diabetes mellitus using recombinant
Start-Up Study Group, et al. Insulin-induced skin lipohy- human insulin: A cross-sectional study. J Pediatr Endocrinol
pertrophy in type 2 diabetes: A multicenter regional survey Metab 2021;34(4):503–508; doi: 10.1515/jpem-2020-0556
in Southern Italy. Diabetes Ther 2020;11(9):2001–2017; 52. Strollo F, Guarino G, Armentano V. Unexplained hypogly-
doi: 10.1007/s13300-020-00876-0 caemia and large glycaemic variability: Skin lipohypertrophy
37. Gunhan H, Elbasan O, Imre E, et al. Lipodystrophy fre- as a predictive sign. Diabetes Res 2016;2(1):24–32.
quency according to insulin treatment regimen in type 2 53. Arda Sürücü H, OKurArslan H. Lipohypertrophy in indi-
diabetic patients: Is insulin injection frequency matters in viduals with type 2 diabetes: Prevalence and risk factors.
analog insulin era? Acta Endo (Buc) 2022;18(2):202–208; J Caring Sci 2018;7(2):67–74; doi: 10.15171/jcs.2018.011
doi: 10.4183/aeb.2022.202 54. Thewjitcharoen Y, Prasartkaew H, Tongsumrit P, et al.
38. Gupta SS, Gupta KS, Gathe SS, et al. Clinical implications Prevalence, risk factors, and clinical characteristics of li-
of lipohypertrophy among people with type 1 diabetes in podystrophy in insulin-treated patients with diabetes: An
India. Diabetes Technol Ther 2018;20(7):483–491; doi: 10 old problem in a new era of modern insulin. Diabetes
.1089/dia.2018.0074 Metab Syndr Obes 2020;13:4609–4620; doi: 10.2147/
39. Hajheydari Z, Kashi Z, Akha O, et al. Frequency of lipo- DMSO.S282926
dystrophy induced by recombinant human insulin. Eur Rev 55. Gentile S, Guarino G, Della Corte T, et al. The economic
Med Pharmacol Sci 2011;15:1196–1201. burden of insulin injection-induced lipohypertophy. Role of
40. Hauner H, Stockamp B, Haastert B. Prevalence of lipohy- education: The ISTERP-3 study. Adv Ther 2022;39(5):
pertrophy in insulin-treated diabetic patients and predis- 2192–2207; doi: 10.1007/s12325-022-02105-5
posing factors. Exp Clin Endocrinol Diabetes 1996; 56. Wang W, Tong Y, Chen Y, et al. Evaluation of ultrasound
104(02):106–110; doi: 10.1055/s-0029-1211431 examination combined with intensive injection technique
41. Ji J, Lou Q. Insulin pen injection technique survey in pa- education on insulin-induced lipohypertrophy (LH) manage-
tients with type 2 diabetes in mainland China in 2010. Curr ment: A prospective cohort study in China. Int J Diabetes Dev
Med Res Opin 2014;30(6):1087–1093; doi: 10.1185/ Ctries 2021;41(4):669–675; doi: 10.1007/s13410-021-00935-7
03007995.2014.895711 57. Gentile S, Guarino G; AMD-OSDI Study Group on Injec-
42. Ji L, Chandran A, Inocencio TJ, et al. The association be- tion Technique, Nefrocenter Research and Nyx Start-Up,
tween insurance coverage for insulin pen needles and et al. Role of structured education in reducing lypodis-
healthcare resource utilization among insulin-dependent trophy and its metabolic complications in insulin-treated
patients with diabetes in China. BMC Health Serv Res people with type 2 diabetes: A randomized multicenter
2018;18(1):300; doi: 10.1186/s12913-018-3095-9 case–control study. Diabetes Ther 2021;12(5):1379–1398;
43. Korkmaz FN, Gökçay Canpolat A, Güllü S. Determination of doi: 10.1007/s13300-021-01006-0
insulin-related lipohypertrophy frequency and risk factors in 58. Kirac C, Gonullu C, Baldane S, et al. Effect of vitamin D
patients with diabetes. Endocrinol Diabetes Nutr (Engl Ed) deficiency on the frequency of lipohypertrophy occurrence
2022;69(5):354–361; doi: 10.1016/j.endinu.2021.07.002 in patients with type 2 diabetes mellitus under injectable
44. Kumar R, Gupta RD, Shetty S, et al. Lipohypertrophy in insulin treatment. Turk J Endocrinol Metab 2020;24:184–188; doi:
injecting patients with diabetes mellitus: An under-recognized 10.25179/tjem.2020-74469
barrier for glycemic control. Int J Diabetes Dev Ctries 2021;
41(2):329–336; doi: 10.1007/s13410-020-00889-2
45. Luo D, Shi Y, Zhu M, et al. Subclinical lipohypertrophy—
Easily ignored complications of insulin therapy. J Diabetes Address correspondence to:
Complications 2021;35(3):107806; doi: 10.1016/j Julia K. Mader, MD
.jdiacomp.2020.107806 Division of Endocrinology and Diabetology
46. Nawaz A, Hasham MA, Shireen M, et al. Prevalence of Department of Internal Medicine
lipohypertrophy and its associations in insulin-treated dia- Medical University of Graz
betic patients. Pak J Med Sci 2022;39(1):209–213; doi: 10 Graz 8036
.12669/pjms.39.1.6134 Austria
47. Omar MA, El-Kafoury AA, El-Araby RI. Lipohypertrophy
in children and adolescents with type 1 diabetes and the E-mail: julia.mader@medunigraz.at