Professional Documents
Culture Documents
Lima 2016
Lima 2016
Abstract
Background: Diabetic retinopathy (DR) is the major cause of blindness among working age adults. The aim of the
study was to investigate risk factors for development of DR.
Methods: A case–control study was performed based on data from 240 individuals (80 cases and 160 controls)
attending the Outpatient Specialty Clinic of the University of South Santa Catarina (UNISUL), between Mar/2010 and
May/2014. Data collection occurred through review of medical charts for presence or absence of DR, determined by
an ophthalmologist. Study protocol included demographic characteristics, metabolic control, diabetes mellitus (DM)
regression analysis to select independent variables (SPSS®18.0 software). Odds ratio (OR) was used as measure of
profile and comorbidities. Statistical analysis used Chi square test for qualitative variables and multivariate logistic
© 2016 The Author(s). This article is distributed under the terms of the Creative Commons Attribution 4.0 International License
(http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium,
provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license,
and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/
publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Lima et al. Int J Retin Vitr (2016) 2:21 Page 2 of 7
with DM without DR compared to controls. Neverthe- ophthalmologist according to clinical findings confirmed
less, the clinical significance of these outcomes should be by direct and indirect ophthalmoscopy following the clas-
investigated and whether they represent a target in diabe- sification of the international clinical diabetic retinopathy
tes treatment to prevent DR [12, 13]. disease severity scale [25]. This study included records of
Another key point for the development of vascular all diabetic patients attended in the study period and they
lesions is the capillary basement membrane thickening were all examined by an ophthalmologist. Patients with
as a result of hyperglycemia, increased synthesis of base- lens opacities and those in which the pupil dilation could
ment membrane components and other factors. This not be performed were excluded (pregnant, uncontrolled
thickening of the basement membrane is related to excess hypertension and suspected of narrow-angle glaucoma).
vascular permeability that leads to leaky vessels, compro- All information was obtained from clinical charts and
mised tight-junctions and increased vesicular transport finally analyzed for the study variables.
[14]. Additionally, the role of genetic factors and their Mean age was used as cutoff point, since the sam-
relationship with the pathogenesis of DR are being exten- ple showed a normal distribution. It was established
sively studied, including the epigenetic mechanisms such the target for glycemic control the glycated hemoglobin
as DNA methylation, histone modifications in chromatin (HbA1c) ≤7.0 % [26]. Regarding comorbidities, diagnosis
and noncoding RNAs [15]. was entirely based on clinical charts, considering previ-
Although major risk factors—hyperglycemia and ous clinical history and laboratory tests. Hypertension
hypertension—are extensively studied and show a strong was set as blood pressure ≥140/90 mmHg (in distinct
association with DR [16, 17], there are variations in con- measures during the outpatient care) or use of antihy-
sistency and pattern of these factors [2]. Some studies pertensive therapy [26]; dyslipidemia set as triglycerides
indicate that prolonged DM duration is indeed a well- ≥150 mg/dL and total cholesterol ≥200 mg/dL or use of
established risk factor for DR [18–20]. Regarding micro- statins for dyslipidemia [16]; diabetic nephropathy set
albuminuria, studies report its association as a marker as microalbuminuria 30–300 mg/L, macroalbuminuria
of microvascular dysfunction and DR [21, 22], however >300 mg/L or previous diagnosis of disease [26].
further studies are needed to confirm this relationship Statistical analysis was performed by a commercially
[21]. Other risk factors such as body mass index (BMI) available statistical software package (SPSS for Windows,
show controversies in relation to its association with this version 18.0, SPSS, Chicago, IL, USA) and involved con-
disease [23]. The aim of this study was to investigate risk struction of frequency distributions, verifying the pres-
factors for DR. ence of association between variables and outcome, using
Chi square test for qualitative variables. Multivariate
Methods analysis was performed using logistic regression and the
A case–control study was performed at the Outpatient model included variables that showed effect with p < 0.20
Specialty Clinic of the University of South Santa Catarina in bivariate analysis, except for gender and age that are
(UNISUL) in Tubarão city, located in southern Brazil. classically part of models. The significance was calculated
Medical records of diabetic patients seen between March using Wald test and Odds ratio (OR) was used as the
2010 and May 2014 were analyzed. Cases and controls measure of association (p < 0.05). This study was based
corresponded to recordings of patients with DR and on the Resolution 466/12 of the National Health Council
a chance of 3.83 (95 % CI 1.57–9.37) for DR (Table 1). hypertension and dyslipidemia, did not reach the statisti-
Similarly, diabetes duration was significantly associated cal significance level established (p < 0.05).
with development of DR. It was observed a positive rela-
tionship between duration of DM and DR, with higher Discussion
chances in 11–15 years of disease (OR 7.52, 95 % CI Descriptive analysis of demographic characteristics
3.03–18.68) and >15 years (OR 9.01, 95 % CI 3.58–22.66) in both groups presented a similar mean age (59.5 vs
(Table 1). Regarding type of diabetes, our results demon- 59.3 years, cases and controls respectively) with a slight
strated a chance of 2.37 (95 % CI 0.52–10.68) for DR in predominance of females. In qualitative analysis, the var-
type 1 DM, but this association did not reach statistical iables age and gender were not associated with a greater
significance (p = 0.261) (Table 1). chance of DR. Similarly, in several studies, no associa-
In bivariate analysis, it was found association between tion was found between DR and gender, age or age group
DR and HbA1c >7 %, duration of DM ≥11 years and [27–31]. However, some reports have found an inde-
insulin use. After logistic regression, insulin use lost asso- pendent association in males [18, 32, 33], while a his-
ciation, while other variables (glycemic control and dura- torical cohort of Kajiwara et al. [34], presented greater
tion of DM) remained associated with DR, regardless chance of DR among females. Yet, in disagreement with
other variables included in the model (Table 1). the present study, Al-Sarraf et al. [35] observed a greater
Multivariate analysis of comorbidities showed a chance chance of DR in individuals aged 50–59 years (OR 2.2,
to DR in patients with diabetic nephropathy of 3.32 (95 % 95 % CI 1.1–5.2) and ≥60 years (OR 4.6, 95 % CI 11.0–
CI 1.62–6.79) (Table 2). Other comorbidities studied, 2.0). Present findings show that probably age and gender
Table 1 Cases and controls according to demographic characteristics, metabolic control and diabetes mellitus profile
Total (240) Diabetic retinopathy ORb CIb† ORAdj CIAdj‡
Gender
Female 153 51 33.3 102 66.7 1.00 0.57–1.74 1.13 0.57–2.26
Male 87 29 33.3 58 66.7 1 1
Age (years)
≤60 122 39 32.0 83 68.0 1 1
>60 118 41 34.7 77 65.3 1.13 0.66–1.93 1.28 0.64–2.56
BMI
18.5–24.9 39 13 33.3 26 66.7 1.09 0.51–2.35 – –
25–30 70 26 37.1 44 62.9 1.29 0.70–2.38 – –
>30 131 41 31.3 90 68.7 1 – –
Glycemic control (%)
HbA1c ≤ 7 80 10 12.5 70 87.5 1 1
HbA1c > 7 160 70 43.8 90 56.3 5.43** 2.61–11.36 3.83* 1.57–9.37
Diabetes duration (years)
<5 87 14 16.1 73 83.9 1 1
6–10 67 14 20.9 53 79.1 1.37 0.61–3.13 2.05 0.71–5.92
11–15 26 12 46.2 14 53.8 4.46* 1.71–11.63 7.52** 3.03–18.68
>15 60 40 66.7 20 33.3 10.42** 4.76–22.73 9.01** 3.58–22.66
Type of diabetes
Type 1 11 6 54.5 5 45.5 2.51 0.74–8.50 2.37 0.52–10.68
Type 2 229 74 32.3 155 67.7 1 1
Insuline use
Yes 112 52 46.4 60 53.6 3.09** 1.76–5.41 1.25 0.61–2.57
No 128 28 21.9 100 78.1 1 1
ORb: Odds-ratio Brut, ORAdj: adjusted Odds-ratio
* p < 0.05; ** p < 0.01 (Pearson’s Chi square test and Wald test)
†
95 % of confidence interval Brut; ‡ 95 % adjusted confidence interval
Lima et al. Int J Retin Vitr (2016) 2:21 Page 4 of 7
Hypertension
Yes 204 71 34.8 133 65.2 1.60 0.71–3.59 – –
No 36 9 25.0 27 75.0 1 – –
Dyslipidemia
Yes 176 66 35.8 112 64.2 1.54 0.81–2.90 1.41 (0.61–3.25)
No 64 17 26.6 47 73.4 1 – –
Diabetic nephropathy
Yes 69 39 56.5 30 43.5 4.12** 2.28–7.44 3.32* (1.62–6.79)
No 171 41 24.0 130 76.0 1
ORb: Odds-ratio Brut, ORAdj: adjusted Odds-ratio
* p < 0.05; ** p < 0.01 (Pearson’s Chi square test and Wald test)
†
95 % of confidence interval Brut; ‡ 95 % adjusted confidence interval
are not risk factors for DR. Therefore, regular moni- Duration of DM, an important predictor DR, deter-
toring of these patients must be done regardless these mines the exposure time of other risk factors [48]. In the
variables. current study, after 10 years of disease, results showed
Among metabolic control factors, none of BMI catego- high association. Al-Sarraf et al. [35] noticed that patients
ries revealed higher chances to DR. Literature differs in with 10–19 years of disease are twice as likely to have DR
this question and demonstrates a controversy with this and approximately three times more when the exposure
association, especially about type 2 DM [23]. Several is more than 20 years. According to a recent study, con-
studies show no association between BMI and DR [18, ducted in United States of America [18], each year of DM
21, 31, 36–39] and other positive association [34, 40, 41]. represents a 6 % increase in chance of DR. This correla-
Ahmed et al. [28], in a follow-up of involving 977 type 2 tion with duration of diabetes was first demonstrated in
DM during 5, 10 and 15 years, did not find an increased the Wisconsin epidemiologic study of diabetic retinopa-
risk of DR in any of categories studied. Yoshida et al. thy (WESDR) [49], with a higher prevalence in 25, 60 and
[42] presented statistical significance in DR frequencies 80 % for 5, 10 and 15 years of evolution of DM, respec-
in patients with BMI > 23.7 kg/m2 only after 7 years of tively. In addition to aforementioned, several other stud-
follow-up. Our findings appear to corroborate the results ies, show statistical significance for this factor, both in
reported in other studies and BMI does not seem to have type 1 DM and in type 2 DM [27, 29, 30, 36, 37], perhaps
direct relationship with development of DR. the most important independent risk factor for DR [31,
Regarding glycemic control, individuals with HbA1c 50, 51].
>7 % were more likely to develop DR. Chronic hyper- Type of DM was not considered a risk factor for DR.
glycemia is responsible for a chain of events that leads Similarly, in a multicenter study in Kuwait, Al Sarraf et al.
to DR [43] and glycemic control one of the key points [35] found no significant results when comparing same
of primary prevention for this complication [29, 44, 45]. variables. However, WESDR, in 25 years of follow-up of
Strong evidences indicate an adequate glycemic control type 1 DM subjects, demonstrated cumulative incidence
in levels of HbA1c ≤7 % as a reducer of risk for DR, both rates of 97 % for DR [52]. The authors suggest that the
in type 1 DM and in type 2 DM [32, 46, 47]. Accordingly lack of association probably occurred due to small num-
to a longitudinal study [28], there was an association ber of DM type 1 individuals in the study’s sample.
between DR and poor glycemic control showing higher Concerning insulin therapy, an interesting finding was
OR for intervals of 5, 10 and 15 years of 8.34, 3.45 and observed. In bivariate analysis, patients using insulin
4.9, respectively. In the present study, inadequate control presented a highly significant chance of DR. In multi-
of blood glucose levels proved to be an important factor variate analysis, odds for DR lost statistical significance
for DR. After diagnosis of DM, patients must reach the and strength of association. Furthermore, Yang et al.
targets for control of glycemia, reducing the chances of [29] found no statistically significant after multivariate
visual impairment. analysis (OR 1.65, 95 % CI 0.80–3.38). On the other hand,
Lima et al. Int J Retin Vitr (2016) 2:21 Page 5 of 7
some studies have shown significantly higher chances recommended to continue as routine when we are han-
in diabetics treated with insulin, OR 3.23 [21] and OR dling these patients [57].
3.51 [8], but only included individuals aged above 40 and Microvascular complications such as diabetic neuropa-
67 years, respectively. One of the most important studies thy and DR can be correlated and they usually follow
on the subject, the United Kingdom Prospective Diabe- with the onset of diabetic nephropathy, which suggests a
tes Study [47], suggested that adequate and continuous strong association between them [23, 58]. Jost et al. [27],
glycemic control inhibits development of DR, regardless after logistic regression revealed a chance of 3.24 (95 % CI
the method of treatment. Other major study, the Dia- 1.04–10.04) for DR in patients with diabetic nephropathy.
betes Control and Complications Trial (DCCT), found Rajalakshmi et al. [37] observed a significant presence
that intensive insulin treatment reduces the risk of devel- of microalbuminuria in both type 1 DM and type 2 DM
opment of DR and slows the progression of clinically individuals with early diagnosis of DM (p = 0.001 and
important retinopathy. In patients without retinopathy at p = 0.009, respectively). Studies indicate that the damage
enrollment, the 3-year risk of developing retinopathy was to the small blood vessels share the same pathogenesis,
reduced by 75 % in the intensive insulin treatment group part of a systemic microvascular dysfunction [59, 60].
compared to the conventional treatment group [53]. The strengths of this study allowed the validation of
In the present study, the authors suggest that prob- results obtained. First, information sources showed an
ably the absence of significance between insulin use and excellent quality because patient records were complete
DR occurred as a reflection of reverse causality bias. In a and properly filled. Second, it is an outpatient specialty
similar way, the WESDR observed no higher relative risk clinic with adequate infrastructure for care and for
for DR with insulin use among patients >30 years [49]. monitoring of patients that maximizes the formation of
Yoshida et al. [42], found that change of therapy method representative groups of cases and controls. Third, we
modifies incidence rate of DR and argued that individu- undertook methodological efforts to enable scientific
als under insulin therapy experience a more severe met- comparability using multivariate data analysis in order
abolic deterioration, not being probably insulin use an to reduce confusing factors. Fourth, it highlights a sig-
isolated risk factor for this diabetic complication. nificant population problem and there is no major epide-
About blood pressure, even in absence of established miological study in Brazil about this topic. Therefore, our
hypertension, changes in homeostasis of blood pres- study amplifies the comparability of data regarding the
sure levels appear to be associated with DR [23]. In a reality of DR in Brazil.
study with diabetic over 67 years [21], the association In addition to its strengths, there are some limita-
between higher levels of systolic pressure and DR at any tions in the present study. It was not possible to perform
stage occurred, but in agreement with the present study, more accurate tests, such as fluorescein angiography or
hypertension as a comorbidity was not an independent SD-OCT, for all cases and controls. Consequently, some
risk factor for DR, as well as in other studies [29, 31, 36, patients with DR could have been inadvertently excluded
54]. Nevertheless, Zheng [54] concluded that the rela- from this study. Moreover, since it is a chart review cri-
tionship between these two variables cannot be excluded, teria of time of disease may have been inaccurate, how-
since isolated pressure measurements may not accurately ever considering the fact that these patients had a longer
represent continued effect of high pressure. In agree- time of disease before the diagnoses it only generates an
ment, some studies indicate hypertension as one of the underestimation of data. Lastly, we could not estimate
most important risk factors in the development of DR specific OR for each variable of individuals with type 1
[50, 51, 55, 56]. and type 2 DM individually, since the sample was mainly
Many researchers have shown that the relationship composed of type 2 DM individuals.
between serum lipids and DR does not seem consistent
or loses association after adjustment for confounding Conclusions
factors [27–30, 37]. The Beijing Eye Study [36] reported According to our results, diabetic patients after 10 years
no association between increased cholesterol levels and of disease with poor glycemic control and nephropathy
DR. Thus, lipid changes do not appear to be the main have a greater chance of DR. The mechanisms underlying
responsible for microvascular involvement. Even though, DR development are not fully elucidated. Thus, the sci-
in the current study, dyslipidemia have not been associ- entific publication in this area should intensify, especially
ated with DR, the importance of diet quality of patients in Brazil, to improve prevention, identify risk groups and
cannot be underestimated. Moreover, in addition to gly- establish real modifiable factors of this highly prevalent
cemic control, HbA1c and lipid profile investigations are disease.
Lima et al. Int J Retin Vitr (2016) 2:21 Page 6 of 7
hospital in Malaysia. Diab Met Syndr: Clin Res Rev. 2014. doi:10.1016/j. 47. Kohner EM, Aldington SJ, Stratton IM, Manley SE, Holman RR, Matthews
dsx.2014.04.019. DR, et al. United Kingdom Prospective Diabetes Study, 30: diabetic
32. Semeraro F, Parrinello G, Cancarini A, Pasquini L, Zarra E, Cimino A, et al. retinopathy at diagnosis of non-insulin-dependent diabetes mellitus and
Predicting the risk of diabetic retinopathy in type 2 diabetic patients. J associated risk factors. Arch Ophthalmol. 1998;116(3):297–303.
Diabetes Complications. 2011;25(5):292–7. 48. Wu CR, Ma ZZ, Hu LN, Xu ZR, Ren ZH. Analysis of systemic factors associ-
33. Hussain S, Qamar MR, Iqbal MA, Ahmad A, Ullah E. Risk factors of retin- ated with diabetic retinopathy. Int J Ophthalmol. 2007;7:1056–9.
opathy in type 2 diabetes mellitus at a tertiary care hospital, Bahawalpur 49. Klein R, Klein BE, Moss SE, Davis MD, Demets DL. The Wisconsin epidemio-
Pakistan. Pak J Med Sci. 2013;29(2):536–9. logic study of diabetic retinopathy. II. Prevalence and risk of diabetic
34. Kajiwara A, Miyagawaa H, Saruwatari J, Kita A, Sakata M, Kawata Y, retinopathy when age at diagnosis is less than 30 years. Arch Ophthal-
et al. Gender differences in the incidence and progression of diabetic mol. 1984;102:520–6.
retinopathy among Japanese patients with type 2 diabetes mellitus: a 50. Bamashmus MA, Gunaid AA, Khandekar RB. 1. Diabetic retinopathy, visual
clinic-based retrospective longitudinal study. Diabetes Res Clin Pract. impairment and ocular status among patients with diabetes mellitus in
2014;103:7–10. Yemen: a hospital-based study. Indian J Ophthalmol. 2009;57:293–8.
35. Al-Sarraf AA, Al-Bannai SK, Al-Furaih AM, El-Shazly MK. Prevalence and fac- 51. Rani PK, Raman R, Chandrakantan A, Pal SS, Perumal GM, Sharma T. Risk
tors associated with diabetic retinopathy, a multi-centric study in Kuwait. factors for diabetic retinopathy in self-reported rural population with
Bull Alex Fac Med. 2010;46(2):99–108. diabetes. J Postgrad Med. 2009;55:92–6.
36. Xi WX, Xu L, Wang YX, Jonas JB. Prevalence and associated factors of 52. Klein R, Knudtson MD, Lee KE, Gangnon R, Klein BE. The Wisconsin
diabetic retinopathy.The Beijing Eye Study 2006. Graefes Arch Clin Exp Epidemiologic Study of Diabetic Retinopathy XXII. The twenty five year
Ophthalmol. 2008;246:1519–26. progression of retinopathy in persons with type 1 diabetes. Ophthalmol.
37. Rajalakshmi R, Amutha A, Ranjani H, Ali MK, Unnikrishnan R, Anjana 2008;115:1859–68.
RM, et al. Prevalence and risk factors for diabetic retinopathy in Asian 53. The DCCT Research Group. The effect of intensive diabetes treatment on
Indians with young onset Type 1 and Type 2. J Diabetes Complications. the development and progression of long-term complications in insulin-
2014;28:291–7. dependent diabetes mellitus: the Diabetes Control and Complications
38. Zheng Y, Lamoureux EL, Lavanya R, Wu R, Ikram MK, Wang JJ. Prevalence Trial. New Engl J Med. 1993;329:977–86.
and risk factors of diabetic retinopathy in migrant indians in an urbanized 54. Zheng W. Factor analysis of diabetic retinopathy in Chinese patients.
society in asia. Ophthalmol. 2012;119(10):2119–24. Diabetes Res Clinical Pract. 2011;92:244–52.
39. Sen D, Gosh S, Roy D. Correlation of C-reactive protein and body mass 55. Steffes M, Cleary P, Goldstein D, Little R, Wiedmeyer HM, Rohlfing C, et al.
index with diabetic retinopathy in Indian population. Diab Met Syndr: The DCCT/EDIC research group. Hemoglobin A1c measurements over
Clin Res Rev; 2014. p. 1–2. nearly two decades: sustaining comparable values throughout the Dia-
40. Cheung N, Wong TY. Obesity and eye diseases. Surv Ophthalmol. betes Control and Complications Trial and the Epidemiology of Diabetes
2007;52:180–95. Interventions and Complications study. Clin Chem. 2005;51:753–8.
41. Dirani M, Xie J, Fenwick E, Benarous R, Rees G, Wong TY, et al. Are obesity 56. Bek T, Lund-Andersen H, Hansen AB, Johnsen KB, Sandbaek A, Lauritzen T.
and anthropometry risk factors for diabetic retinopathy? The diabetes The prevalence of diabetic retinopathy in patients with screen-detected
management project. Invest Ophthalmol Vis Sci. 2011;52:4416–21. type 2 diabetes in Denmark: the ADDITION study. Acta Ophthalmol.
42. Yoshida Y, Hagura R, Hara Y, Sugasawa G, Akanuma Y. Risk factors for 2009;87(3):270–4.
the development of diabetic retinopathy in Japanese type 2 diabetic 57. El-Bab MF, Shawky N, Al-Sisi A, Akhtar M. Retinopathy and risk factors
patients. Diabetes Res Clin Pract. 2001;51:195–203. in diabetic patients from Al-Madinah Al-Munawarah in the Kingdom of
43. Cheung N, Mitchell P, Wong TY. Diabetic retinopathy. Lancet. Saudi Arabia. Clin Ophthalmol. 2012;6:269–76.
2010;376:124–36. 58. Alrawahi AH, Rizvi SG, Al-Riyami D, Al-Anqoodi Z. Prevalence and risk
44. Mohamed Q, Gillies MC, Wong TY. Management of diabetic retinopathy. factors of diabetic nephropathy in omani type 2 diabetics in Al-dakhiliyah
JAMA. 2007;298(8):902–16. region. Oman Med J. 2012;27(3):212–6.
45. Raymond NT, Varadhan L, Reynold DR, Bush K, Sankaranarayanan S, 59. Qiu C, Cotch MF, Sigurdsson S, Garcia M, Klein R, Jonasson F. Retinal and
Bellary S, et al. Higher prevalence of retinopathy in diabetic patients of cerebral microvascular signs and diabetes: the age, gene/environment
South Asian ethnicity compared with White Europeans in the commu- susceptibility—Reykjavik study. Diabetes. 2008;57(6):1645–50.
nity: a cross-sectional study. Diabetes Care. 2009;32:410–5. 60. Cheung N, Wong TY. Diabetic retinopathy and systemic vascular compli-
46. The Diabetes control and Complication Trial Research Group. The effect cations. Prog Retin Eye Res. 2008;27(2):161–76.
of intensive treatment of diabetes on the development and progression
of long-term complications in insulin-dependent diabetes mellitus. N
Engl J Med. 1993;329:977–86.