Nonlinear relationship between systolic blood pressure and diabetic retinopathy in type 2 diabetic patients in southern China: a single-center observational study

Background Blood pressure (BP) control has been shown in clinical trials to reduce the risk of diabetic retinopathy (DR). To some extent, systolic blood pressure (SBP) has been shown to have a positive correlation with DR. However, there are no studies that have standardized SBP thresholds for DR prevention. Our goal was to use threshold analysis to further investigate the relationship between SBP and DR, identifying safe levels of SBP that contribute to DR prevention Methods We analyzed data from a cross-sectional study (December 2017-November 2018, n = 426, mean age 59.15{+/-}13.68) from the Department of Endocrinology, Guangdong Provincial People's Hospital in patients with type 2 diabetes mellitus. DR severity was assessed by retinal photographs. The International Clinical Diabetic Retinopathy and Diabetic Macular Edema Disease Severity Scale was used to classify DR severity and divide it into two groups: with DR and without DR. SBP was analyzed as a continuous variable. Multivariate logistic regression models, smoothed curve fitting, threshold analysis, and interaction tests were used to assess the relationship between BP and DR. Results Prevalence of DR in the study population was 39.20%. After adjusting for age, sex, DM duration, HbA1c, BUN, HDL, LDL, TRIG, CHOL, TP, DBP, PP, eGFR, and CKD stage, the association between SBP and DR appears as a threshold effect, with a inflection point of 132mmHg. The risk of DR did not change significantly when SBP [≤]132mmHg (OR: 0.86; 95% CI: 0.63 to 1.17, p=0.3400). When SBP [≥]132mmHg, each 10 mmHg increase in SBP raise the risk of developing DR by 28% (OR:1.28; 95% CI:1.07 to 1.54, P=0.0081). In addition, a stronger association of SBP with DR in patients with TP[≤]60g/L (OR=1.58, 95% CI: 1.19- 2.08, P=0.001) compared to those with TP>60g/L (OR=1.15, 95% CI: 1.03, 1.27, P=0.012) was discovered, with a P value for interaction=0.023. Conclusion In Chinese patients with type 2 diabetes, SBP was significantly associated with DR when SBP [≥]132 mmHg. Further longitudinal studies are needed to confirm our findings, especially in patients with TP[≤]60g/L.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted October 21, 2022. ; https://doi.org/10.1101/2022.10.18.22281231 doi: medRxiv preprint CKD stage, the association between SBP and DR appears as a threshold effect, with a  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint information on these patients was gathered in a non-selective and continuous manner.

107
The following were the inclusion criteria: (1) patients with type 2 diabetes (according  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

126
The target exposure variable was SBP, which was recorded as a continuous variable.

127
The BP detection procedure was as follows: after five minutes of rest, an Omron 128 electronic BP monitor was applied to the right arm of the seated participants. Two 129 blood pressure (BP) measurements were taken five minutes apart. If the difference 130 was greater than 10 mm for SBP and 5 mm for DBP, a third measurement was taken.

131
The mean of the two closest readings was used as the BP value [14]. The difference 132 between SBP and DBP (SBP-DBP) was used to calculate pulse pressure (PP). The covariates included in this study were: sex, age, diabetes mellitus duration (DM is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted October 21, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint  212 We discovered a nonlinear relationship between SBP and DR using smoothed curve 213 fitting and a generalized summation model. After adjusting for age, sex, DM duration, 214 HbA1c, BUN, HDL, LDL, TRIG, CHOL, TP, DBP, PP, eGFR, and CKD stage, 215 the findings (Fig 1) show that the association between SBP and DR appears as a 216 threshold effect. We calculated the SBP inflection point to be 132 mm Hg using a  (Table 3). is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

Non-linear association of SBP with DR
The copyright holder for this this version posted October 21, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted October 21, 2022.  The association between SBP and DR was investigated in this retrospective cohort 262 study. We discovered a threshold effect relationship between SBP and DR by . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint However, due to the small number of patients in our study with TP ≤60, the effect of 268 SBP on DR in this group could not be accurately elaborated, which will be one of the 269 priority groups for future studies on the relationship between SBP and DR.  However, in the fully adjusted model (Adjust II model), we discovered that the 277 relationship between SBP and DR was not linear (supplementary information Fig 1). 278 We hypothesize that some residual confounding factors may obscure the true is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The pathophysiology of the relationship between SBP and DR has not been fully is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint   is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint In conclusion, our study demonstrated that SBP≥132 mm Hg was significantly 328 associated with the risk of DR in patients with type 2 diabetes in southern China (OR 329 between SBP/10 and DR: 1.28, 95% CI: 1.07 to 1.54), which suggests that to prevent 330 the risk of DR in diabetic patients, their SBP should be controlled at <132 mm Hg.   is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint