Model suitable for PCOS crowd to predict MetS, construction method and application

Abstract

The invention discloses a model suitable for predicting metabolic syndrome (MetS) of people with polycystic ovarian syndrome (PCOS), a construction method and application, the model is characterized in that values are respectively assigned on the basis of indexes of a plurality of predictive variables, the predictive variables of the model are calculated, and the predictive variables of the predictive variables of the model are calculated according to the predictive variables of the predictive variables of the model, so that the predictive variables of the predictive variables of the model can be calculated according to the predictive variables of the model. The predictive variable is selected from sex hormone binding globulin, acne, maternal family menoxenia incidence history, age, apolipoprotein-B / apolipoprotein-A, fasting insulin and body mass index; and fitting the PCOS with MetS prediction model by adopting Logistic regression, and obtaining the morbidity according to the prediction model. According to the scheme, by optimizing model design, the requirement for the number of indexes is lowered, so that the application difficulty is lowered while the prediction precision is improved, and the prediction efficiency is remarkably improved.

Description

Technical Field

[0001] The present invention belongs to the field of medical technology, and particularly relates to a model, a construction method and an application for predicting MetS in PCOS population, in particular to a prediction model, a construction method and an application for predicting the probability of occurrence of metabolic syndrome (MetS) in the future period in polycystic ovary syndrome (PCOS) population, so as to provide effective information for clinical prevention, auxiliary judgment, etc. Background Art

[0002] MetS is a group of clinical syndromes characterized by the aggregation of obesity, hyperglycemia, hyperlipidemia and hypertension. As a complex of multiple diseases, its mortality rate is much higher than that of suffering from only one of these diseases alone. Once MetS appears, its harm to the human body cannot be ignored. It is estimated that about 1 / 4 to 1 / 3 of adults globally are troubled by MetS, and the adult prevalence rate is about 20%. MetS has become a global health problem. In recent years, with the aggravation of population aging, the transformation of lifestyle and the rapid development of social economy, the prevalence rate of MetS shows a gradually increasing trend, which in turn leads to a significant increase in the disease burden such as the prevalence rate, disability rate and mortality rate of its proximal outcomes, type 2 diabetes and cardiovascular and cerebrovascular diseases. Due to the unique environmental climate, eating habits and lifestyle behaviors in different regions, the risk factors of MetS may be different. At present, there is still a lack of systematic research on predicting the risk of MetS in regional populations, and it is urgent to establish a MetS risk prediction model to fill this gap.

[0003] Therefore, in view of the above technical problems, it is necessary to provide a model, a construction method and an application for predicting MetS in PCOS population.

[0005] The information disclosed in this background art section is only intended to enhance the understanding of the overall background of the present invention, and should not be regarded as an admission or any form of suggestion that this information constitutes the prior art already known to those of ordinary skill in the art. Summary of the Invention

[0006] The purpose of the present invention is to provide a model, a construction method and an application for predicting MetS in PCOS population.

[0007] In order to achieve the above purpose, the technical solution provided by a specific embodiment of the present invention is as follows:

[0008] A model for predicting MetS in PCOS population,

[0009] Indices based on several predictor variables are assigned values respectively, and the predictor variables are selected from: sex hormone-binding globulin (SHBG), acne, family history of oligomenorrhea in the maternal line (FO), age, apolipoprotein B / apolipoprotein A (BA), fasting insulin (FINS), body mass index (BMI).

[0010] According to the PCOS with MetS prediction model fitted by Logistic regression, the incidence rate can be obtained based on the prediction model.

[0011] In one or more embodiments of the present invention, the assignment includes:

[0012] For sex hormone-binding globulin, assign a value of 0 or 1 according to its index, where: "0" represents < 30.25 nmol / L, and "1" represents ≥ 30.25 nmol / L;

[0013] For acne, assign a value of 0 or 1 according to its index, where: "0" represents no acne, and "1" represents having acne;

[0014] For the family history of oligomenorrhea in the maternal line, assign a value of 0 or 1 according to its index, where: "0" represents no family history of oligomenorrhea in the maternal line, and "1" represents having a family history of oligomenorrhea in the maternal line;

[0015] For apolipoprotein B / apolipoprotein A, assign a value of 0 or 1 according to its index, where: "0" represents < 0.70, and "1" represents ≥ 0.70;

[0016] For fasting insulin, assign a value of 0 or 1 according to its index, where: "0" represents < 13.25 mIU / mL, and "1" represents ≥ 13.25 mIU / mL;

[0017] For body mass index, assign a value of 0 or 1 or 2 according to its index, where: "0" represents < 23 kg / m 2 ,"1" represents 23 kg / m 2 ≦BMI < 25 kg / m 2 ;"2" represents ≥ 25 kg / m 2 .

[0018] In one or more embodiments of the present invention, the nomogram curve fitting formula of the PCOS with MetS clinical prediction model is: P = 1 / (1 + exp(–(0.066 × age + 0.596 × 23 ≦ BMI < 25 + 1.531 × BMI ≥ 25 – 0.248 × SHBG ≥ 30.25 + 1.280 × BA ≥ 0.700 + 0.560 × maternal oligomenorrhea history + 0.370 × acne + 1.338 × FINS ≥ 13.25 – 5.4147))).

[0019] In one or more embodiments of the present invention, a method for constructing a model for predicting MetS in the PCOS population includes

[0020] Screening volunteers who meet the diagnostic standard indicators;

[0021] Collecting the information of the volunteers and performing variable data conversion based on the information, where the information is selected from: the menstrual conditions of the patients, family history (including at least the history of hypertension and diabetes in first-degree relatives, the history of early-onset alopecia in the paternal family, and the history of oligomenorrhea in the maternal family), height, weight, WC, HC, SBP, DBP, mF-G score (hirsutism score), acne score, acanthosis nigricans score, BMI, WHR, FSH, LH, T, DHEAS, AND, SHBG, FPG, FINS, TG, HDL-C, LDL-C, Apo-A, Apo-B;

[0022] Based on R language, using Lasso regression and univariate regression to screen the predictive variables, and combining the screened predictive variables with clinical experience to perform multiple Logistic regression to construct a PCOS with Mets prediction model.

[0023] In one or more embodiments of the present invention, it further includes using the "regplot" package to draw a nomogram of the PCOS with Mets prediction model.

[0024] In one or more embodiments of the present invention, it further includes model verification, which is selected from: ROC curve evaluation, calibration curve evaluation, clinical decision curve verification, and clinical impact curve verification.

[0025] In one or more embodiments of the present invention, the ROC curve evaluation includes: drawing an ROC curve, calculating the area under the ROC curve (AUC) to obtain the C-statistic. When the C-statistic > 0.5, the prediction model has predictive ability; when the C-statistic is closer to 1, the discrimination of the model is better.

[0026] In one or more embodiments of the present invention, the calibration curve evaluation includes: drawing a calibration curve, calculating the Brier score of the calibration curve. When Brier < 0.25, it proves that the model has predictive ability, and the closer the score is to 0, the better the calibration of the model. The scatter distribution of the calibration curve of the model is distributed along the 45-degree diagonal line, indicating that the calibration of the model is good.

[0027] In one or more embodiments of the present invention, the application of the model for predicting MetS in the PCOS population in the risk assessment of the likelihood of developing MetS in the future period.

[0028] Compared with the prior art, the model, construction method and application for predicting MetS in PCOS population of the present invention optimize the model design, reduce the requirement for the number of indicators, thereby improving the prediction accuracy while reducing the application difficulty, significantly improving the prediction efficiency, and aiming at common deterministic risk factors, and establishing a risk analysis model accordingly. The prediction results provide a reference basis for the prevention and treatment strategies of MetS. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments recorded in the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0030] Figure 1 It is the construction and verification roadmap of the model for predicting MetS in PCOS population in an embodiment of the present invention;

[0031] Figure 2 It is the schematic diagram of using the LASSO regression model for texture feature selection in the construction of the model for predicting MetS in PCOS population in an embodiment of the present invention: (A) In the lasso model, 10-fold cross-validation is selected and the lowest standard is adopted. (B) The LASSO regression algorithm is used to screen out 9 features with non-zero coefficients from 22 features;

[0032] Figure 3 It is the Nomogram of the model for predicting MetS in PCOS population in an embodiment of the present invention;

[0033] Figure 4 It is the verification comparison diagram of the model for predicting MetS in PCOS population in an embodiment of the present invention: (A) The receiver operating characteristic (ROC) curve of the model. (B) The calibration curve of the model. (C) The bootstrap calibration curve of the model after 100 self-samplings;

[0034] Figure 5 It is the prediction effect evaluation comparison diagram of the model for predicting MetS in PCOS population in an embodiment of the present invention: (A) Clinical decision curve / DCA curve. (B) Clinical impact curve. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0035] To enable those skilled in the art to better understand the technical solutions in the present invention, the technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without making creative efforts shall fall within the scope of protection of the present invention.

[0036] Example 1

[0037] The participating volunteers were sourced from January 2015 to January 2022, aged 19 - 40 years old, with a previous history of good health. They were PCOS patients who visited the gynecology outpatient department of the First Affiliated Hospital of Heilongjiang University of Chinese Medicine due to diseases such as ovulatory dysfunction, abnormal uterine bleeding, amenorrhea, infertility, etc., and met the inclusion and exclusion criteria of this study. They were recorded in the integrated clinical research information system, totaling 967 cases.

[0038] The research data was sourced from the initial diagnosis information of PCOS patients in the integrated clinical research information system, including clinical data such as medical history, family history, physical examination, etc.

[0039] Diagnostic criteria

[0040] (1) Diagnostic criteria for polycystic ovary syndrome

[0041] Referring to the 2011 Chinese PCOS diagnostic criteria: Suspected PCOS ① oligomenorrhea or amenorrhea or irregular uterine bleeding; ② clinical manifestations of hyperandrogenism and / or hyperandrogenemia; ③ pelvic ultrasound of the gynecology department suggesting polycystic ovarian changes: the number of follicles with a diameter of 2 - 9 mm in one or both ovaries ≥ 12 and / or ovarian volume ≥ 10 ml. Confirmed PCOS: ① is a necessary condition, and additionally add ② and / or ③, and exclude hyperandrogenemia and ovulatory dysfunction caused by other diseases.

[0042] (2) Diagnostic criteria for metabolic syndrome

[0043] Refer to the diagnostic criteria for MetS recommended by the Third Adult Treatment Panel of the National Cholesterol Education Program (NCEP-ATPⅢ) in 2005: ① Central obesity (refer to the Asian abdominal obesity criteria): WC ≥ 80 cm for women; ② Triglyceride (TG) > 1.7 mol / L, or has received this special treatment; ③ High density lipoprotein cholesterol (HDL-C) < 1.3 mmol / L, or has received this special treatment; ④ Elevated blood pressure, systolic blood pressure ≥ 130 mmHg and / or diastolic blood pressure ≥ 85 mmHg, or has been diagnosed with hypertension and is receiving antihypertensive treatment; ⑤ Fasting Plasma Glucose (FPG) ≥ 5.6 mmol / L, or has been diagnosed with diabetes and is receiving hypoglycemic treatment. If 3 or more of the above 5 criteria are met, MetS can be diagnosed.

[0044] Inclusion and exclusion criteria

[0045] (1) Inclusion criteria

[0046] ① Aged 19 to 40 years old;

[0047] ② Meeting the 2011 Chinese diagnostic criteria for PCOS;

[0048] ③ Signing the informed consent form.

[0049] (2) Exclusion criteria

[0050] ① Taking hormonal drugs, hypoglycemic, lipid-lowering, or blood pressure-lowering drugs in the past 3 months, or taking drugs that affect laboratory test results in the past 1 month;

[0051] ② Abnormal liver and kidney functions, and other abnormal diseases in the endocrine, nervous, mental, hematological and other systems;

[0052] ③ Pregnant or lactating women.

[0053] Clinical general information

[0054] Professional medical staff inquired about the medical history and measured the baseline physical conditions. Asked the patients about their menstrual conditions, family history (hypertension and diabetes history of first-degree relatives, early-onset alopecia history in the paternal family, oligomenorrhea history in the maternal family).

[0055] Physical examinations included height, weight, WC, HC, SBP, DBP, mF-G scoring method, acne scoring, acanthosis nigricans scoring, and calculation of BMI and WHR.

[0056] The hirsutism score is based on the modified Ferriman-Gallwey (mF-G) scoring method. According to large-scale epidemiological surveys in China, a clinical diagnosis of hirsutism is made when the mF-G score is ≥4 points; the acne scoring standard refers to the Pillsbury classification method; the acanthosis nigricans score refers to the 4-level scoring method. A score of ≥1 indicates "presence" of acne and acanthosis nigricans, and a score <1 indicates "absence".

[0057] Laboratory data

[0058] Instruct the patient to collect gonadal hormone indicators at 08:00 - 09:00 on an empty stomach in the early morning on the 3rd to 7th day of the natural menstrual cycle or progesterone withdrawal bleeding: collect FSH, LH, T, DHEAS, AND, SHBG; collect glycolipid indicators: FPG, FINS, TG, HDL-C, LDL-C, Apo-A, Apo-B. Calculate the LH / FSH ratio.

[0059] All laboratory indicators in this example are from the laboratory of the First Affiliated Hospital of Heilongjiang University of Chinese Medicine.

[0060] Candidate variable data transformation

[0061] (1) According to the clinical diagnostic criteria, convert the following continuous variables into categorical variables

[0062] For LH / FSH ≥2 and LH / FSH <2, assign values of "1" and "0" respectively;

[0063] For mF-G ≥4 and mF-G <4, assign values of "1" and "0" respectively;

[0064] BMI: normal: BMI <23 kg / m2, overweight: 23 kg / m 2 ≤BMI <25 kg / m 2 、obese: BMI ≥25 kg / m 2 , assign values of "0", "1", and "2" respectively;

[0065] For acne score ≥1 and acne score <1, assign values of "1" and "0" respectively;

[0066] For acanthosis nigricans score ≥1 and acanthosis nigricans score <1, assign values of "1" and "0" respectively.

[0067] (2) Analyze the relationship between continuous variables and MetS using the ROC curve, intercept the optimal cut-off value, and convert the following continuous variables into categorical variables according to the cut-off value.

[0068] For SHBG <30.25 and SHBG ≥30.25, assign values of "0" and "1" respectively;

[0069] When FINS≥13.23, it is assigned the value of "1"; when FINS<13.23, it is assigned the value of "0".

[0070] When HC≥101.80, it is assigned the value of "1"; when HC<101.80, it is assigned the value of "0".

[0071] When Apo - B / Apo - A≥0.70, it is assigned the value of "1"; when Apo - B / Apo - A<0.70, it is assigned the value of "0".

[0072] (3) Family history (type 2 diabetes, hypertension, maternal oligomenorrhea, paternal early baldness), early baldness, PCOM. "Yes" and "no" are used to determine binary variables, which are respectively assigned the values of "1" and "0".

[0073] Variable screening and construction of the model

[0074] Using the "glmnet", "MASS", and "rms" packages in R language 4.4.1 software, Lasso regression is used to screen for predictive variables. Combining with clinical experience, predictive variables are screened out, and the screened variables are subjected to Logistic regression to construct a PCOS with Mets prediction model. The "regplot" package is used to draw the nomogram of the PCOS with MetS prediction model.

[0075] The "rmda" package is used to draw the decision curve analysis (DCA) and clinical impact curve (CIC) to evaluate the model benefits.

[0076] (1) Model variable screening and construction

[0077] Based on the clinical experience of PCOS diagnosis and treatment, clinical data are collected and used as the predictive variables of the model. The model predictive variables include patients' physical appearance characteristics (hirsutism, acne, seborrhea, early baldness, acanthosis nigricans, BMI, HC), family history (type 2 diabetes, hypertension, maternal oligomenorrhea, paternal early baldness), age, PCOM, gonadal hormone levels (LH / FSH, TT, DHEAS, AND, SHBG), and glycolipid levels (FINS, TC, LDL - C, Apo - B / Apo - A), totaling 22 candidate predictive factors.

[0078] LASSO regression is used to screen candidate predictive variables, see Figure 2, combined with clinical experience, finally 7 predictive factors (age, acne, maternal history of oligomenorrhea, apolipoprotein-B / apolipoprotein-A, fasting insulin, body mass index, sex hormone-binding globulin) were screened out, and a clinical prediction model for PCOS with MetS was fitted using Logistic regression (see Table 1). According to the Logistic model parameters, the incidence formula of the clinical prediction model for PCOS with MetS: P = 1 / (1 + exp(–(0.066×age + 0.596×23≤BMI<25 + 1.531×BMI≥25 – 0.248×SHBG≥30.25 + 1.280×BA≥0.700 + 0.560×maternal history of oligomenorrhea + 0.370×acne + 1.338×FINS≥13.25 – 5.4147))).

[0079] As Figure 2 shown: The LASSO regression model was used for texture feature selection. (A) In the lasso model, 10-fold cross-validation was selected and the lowest criterion was adopted. A curve of binomial deviance versus log(λ) was plotted. Based on the minimum criterion and 1 standard error of the minimum criterion, vertical dashed lines were plotted at the optimal λ value, and the optimal λ value was 0.0180. (B) The LASSO regression algorithm was used to screen out 9 features with non-zero coefficients from 22 features.

[0080]

[0081]

[0082] Table 1: Clinical prediction model of Logistic PCOS with MetS. Note: SHBG: sex hormone-binding globulin; acne: acne; FO: maternal family history of oligomenorrhea; age: age; BA: apolipoprotein-B / apolipoprotein-A; FINS: fasting insulin; BMI: body mass index; CI: confidence interval; OR: odds ratio. (2) Visual display of the Nomogram of the clinical prediction model for PCOS with MetS

[0083] According to the Logistic regression fitting of the clinical prediction model for PCOS with MetS, the "regplot" package was used to draw the Nomogram of the prediction model for PCOS with MetS, as shown in Figure 3 .

[0084] As Figure 3Shown as: Nomogram of PCOS with Mets prediction model. Among them: SHBG: sex hormone-binding globulin, "0" represents < 30.25 (nmol / L), "1" represents ≥ 30.25 (nmol / L); acne: acne, "0" represents no acne, "1" represents having acne; FO: maternal family history of oligomenorrhea, "0" represents no maternal family history of oligomenorrhea, "1" represents having maternal family history of oligomenorrhea; age: age; BA: apolipoprotein-B / apolipoprotein-A, "0" represents < 0.70, "1" represents ≥ 0.70; FINS: fasting insulin, "0" represents < 13.25 (mIU / mL), "1" represents ≥ 13.25 (mIU / mL); BMI: body mass index, "0" represents < 23 (kg / m 2 ) and "1" represents ≥ 23 (kg / m 2 ) < 25 (kg / m 2 ) and "2" represents ≥ 25 (kg / m 2 ).

[0085] Model validation

[0086] Apply the "pROC" and "regplot" packages to draw the ROC curve, calculate the area under the ROC curve (AUC) to obtain the C-statistic. When the C-statistic > 0.5, it proves that the prediction model has predictive ability, and the closer the statistic is to 1, the better the discrimination of the model.

[0087] Apply the "regplot" package to draw the calibration curve, calculate the Brier score of the calibration curve. When Brier < 0.25, it proves that the model has predictive ability, and the closer the score is to 0, the better the calibration of the model. The scatter distribution of the calibration curve of the model is along the 45-degree diagonal line, indicating good calibration of the model.

[0088] The ROC curve of the PCOS with MetS clinical prediction model shows that the AUC is 0.857, that is, the C statistic is 0.857 (95% CI: 0.833 - 0.884), the optimal cut-off point P = 0.377, the sensitivity is 70.7%, and the specificity is 86.6%, indicating good discrimination of the prediction model. The calibration curve of the model shows that the calibration scatter is distributed along the 45-degree diagonal line, and the Brier score is 0.141, indicating good calibration of the model, as Figure 4 shown in A and B below.

[0089] The Bootstrap method is used for internal model validation. The "regplot" package is used to draw the Bootstrap internal validation calibration curve. The model data is resampled 100 times, and the adjusted C statistic is calculated to be 0.849 and the Brier score is 0.145. The internal validation shows that the model performs well, asFigure 4 as shown in C.

[0090] As Figure 4 Shown is the validation of the clinical prediction model for PCOS with MetS. (A) Receiver operating characteristic (ROC) curve of the model. The ROC curve was used to evaluate the ability of the model to predict MetS (MetS) in PCOS patients. The cut-off value was 0.377 (0.757 0.829). (B) Calibration curve of the model. Calibration of the model between predicted risk and actual probability. The calibration curve depicts the consistency of the model in terms of calibration, i.e., the predicted risk of MetS and the actual incidence. The vertical axis represents the actual incidence of MetS. The horizontal axis represents the predicted MetS risk. The diagonal dashed line represents the perfect prediction of an ideal model. (C) Bootstrap calibration curve of the model after 100 self-samplings. The horizontal axis represents the predicted MetS risk. The vertical axis represents the actual confirmed MetS cases. The diagonal dashed line represents the perfect prediction of an ideal model. The solid line represents the performance of the model, and the closer it is to the diagonal dashed line, the better the prediction effect.

[0091] The "rmda" and "ggplot2" packages were used to plot the clinical decision curve (DCA) to evaluate the clinical usability and benefits of the model. The horizontal dashed line in the DCA curve indicates the extreme case where all patients predicted by the model are disease-free, and at this time the net benefit is 0. The slanted dashed line indicates the extreme case where all patients predicted by the model are diseased, and at this time the slope of the net benefit is negative. According to the clinical prediction model, if the drawn DCA curve is above the horizontal dashed line and the slanted dashed line, it indicates that the model has clinical practicability. If the curve is closer to the dashed line, the clinical value is lower; if it overlaps with or is below the dashed line, it has no clinical value. To evaluate the clinical benefits of the model, the DCA curve can be directly observed. The model with the highest net benefit (Net Benefit, NB) within the decision threshold range is the optimal model. The abscissa is the threshold probability: the probability (Pi) that PCOS patient i has MetS. When Pi reaches a certain threshold (Pt), it is marked as positive, and at this time the patient will have the benefit of treatment and also the loss of untreated patients (false negatives). Subtracting the loss from the benefit gives the ordinate, i.e., NB. As Figure 5 shown in A, the Pt range of the clinical prediction model for PCOS with MetS is 2% - 83%, and the net benefit of the model is above that of no intervention and full intervention. In this range, the missed diagnosis rate and over-intervention rate can be reduced, indicating good clinical decision-making value of the model.

[0092] The "rmda" and "ggplot2" packages were used to plot the clinical impact curve (CIC). The X-axis is the risk threshold, and the Y-axis is the number of people at risk per thousand. The blue solid line represents the number of people predicted by the model to have the outcome event, and the blue dashed line represents the actual number of people. The closer the two curves are, the higher the degree of match between the model prediction and the actual occurrence, and the higher the clinical efficiency. As Figure 5As shown in B, when the threshold is greater than 0.6, the clinical effective rate is higher.

[0093] According to the incidence formula of the PCOS with MetS clinical prediction model: P = 1 / (1 + exp(–(0.066×age + 0.596×(23 ≤ BMI < 25) + 1.531×(BMI ≥ 25) – 0.248×(SHBG ≥ 30.25) + 1.280×(BA ≥ 0.700) + 0.560×maternal history of oligomenorrhea + 0.370×acne + 1.338×(FINS ≥ 13.25) – 5.4147))), calculate the incidence rate of MetS in the PCOS population.

[0094] Example 1: A 32-year-old PCOS patient, with no maternal family history of oligomenorrhea, no facial acne, BMI < 23, SHBG > 30.25, FINS < 13.25, and Apo-B / ApoA < 0.70.

[0095] According to the incidence formula of the PCOS with MetS clinical prediction model, calculate the incidence rate P of this patient: P = 1 / (1 + exp(-(0.066×32 + 0.596×0 + 1.531×0 - 0.248×1 + 1.280×0 + 0.560×0 + 0.370×0 + 1.338×0 - 5.4147))) = 0.0003.

[0096] Calculate the probability of disease according to the incidence formula of the PCOS with MetS clinical prediction model. The probability of this PCOS patient having MetS is 0.3%, indicating that the risk of having MetS is very low, and it is not recommended to carry out preventive treatment for abnormal glucose and lipid metabolism clinically.

[0097] Example 2: A 30-year-old PCOS patient, with a maternal family history of oligomenorrhea, no facial acne, BMI = 26.54, SHBG > 30.25, FINS < 13.25, and Apo-B / ApoA < 0.70.

[0098] According to the incidence formula of the PCOS with MetS clinical prediction model, calculate the incidence rate P of this patient: P = 1 / (1 + exp(-(0.066×30 + 0.596×0 + 1.531×2 - 0.248×1 + 1.280×0 + 0.560×1 + 0.370×0 + 1.338×0 - 5.4147))) = 0.465.

[0099] Calculate the probability of disease according to the incidence formula of the PCOS with MetS clinical prediction model. The probability of this PCOS patient having MetS is 46.5%, indicating that the risk of having MetS is relatively low, and it is not recommended to carry out preventive treatment for abnormal glucose and lipid metabolism clinically.

[0100] Example 3. A 26-year-old PCOS patient with a history of oligomenorrhea in the maternal family, acne on the face, BMI = 23.17, SHBG > 30.25, FINS > 13.25, and Apo-B / ApoA > 0.70.

[0101] According to the incidence formula of the PCOS with MetS clinical prediction model, the incidence of this patient is calculated as P = 1 / (1 + exp(-(0.066×26 + 0.596×1 + 1.531×0 - 0.248×1 + 1.280×1 + 0.560×1 + 0.370×1 + 1.338×1 - 5.4147))) = 0.612.

[0102] According to the incidence formula of the PCOS with MetS clinical prediction model to calculate the probability of disease, the probability of this PCOS patient having MetS is 61.2%, indicating a relatively high risk of having MetS. It is recommended to conduct preventive treatment for abnormal glucose and lipid metabolism clinically.

[0103] Example 4. A 22-year-old PCOS patient with a history of oligomenorrhea in the maternal family, acne on the face, BMI = 20.15, SHBG < 30.25, FINS > 13.25, and Apo-B / ApoA > 0.70.

[0104] According to the incidence formula of the PCOS with MetS clinical prediction model, the incidence of this patient is calculated as P = 1 / (1 + exp(-(0.066×22 + 0.596×0 + 1.531×0 - 0.248×0 + 1.280×1 + 0.560×1 + 0.370×1 + 1.338×1 - 5.4147))) = 0.278

[0105] According to the incidence formula of the PCOS with MetS clinical prediction model to calculate the probability of disease, the probability of this PCOS patient having MetS is 27.8%, indicating a relatively low risk of having MetS. It is not recommended to conduct preventive treatment for abnormal glucose and lipid metabolism clinically.

[0106] Example 5. A 25-year-old PCOS patient with a history of oligomenorrhea in the maternal family, acne on the face, BMI = 30.11, SHBG < 30.25, FINS > 13.25, and Apo-B / ApoA > 0.70.

[0107] According to the incidence formula of the PCOS with MetS clinical prediction model, the incidence of this patient is calculated as P = 1 / (1 + exp(-(0.066×25 + 0.596×0 + 1.531×2 - 0.248×0 + 1.280×1 + 0.560×1 + 0.370×1 + 1.338×1 - 5.4147))) = 0.999

[0108] Calculate the probability of disease according to the incidence formula of the clinical prediction model for PCOS with MetS. The probability of this PCOS patient having MetS is 99.9%, indicating a relatively high risk of having MetS. It is recommended to carry out preventive treatment for abnormal glucose and lipid metabolism clinically.

[0109] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be embraced within the present invention.

[0110] In addition, it should be understood that although this specification is described according to embodiments, not every embodiment only contains an independent technical solution. This narrative way of the specification is only for clarity. Those skilled in the art should regard the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A model for predicting MetS in PCOS population, Assigning values ​​based on indicators of several predictive variables, wherein the predictive variables are selected from: sex hormone binding globulin, acne, maternal family history of oligomenorrhea, age, apolipoprotein-B / apolipoprotein-A, fasting insulin, and body mass index; According to the use of Logistic regression to fit the PCOS with MetS prediction model, the incidence rate can be obtained according to the prediction model.

2. The model for predicting MetS in PCOS population according to claim 1, characterized in that: The assignment includes: Sex hormone binding globulin, assigned a value of 0 or 1 according to its index, where "1" represents ≥30.25nmol / L and "0" represents <30.25nmol / L; Acne, assigned a value of 0 or 1 according to its index, where "0" represents no acne and "1" represents acne; Maternal family history of oligomenorrhea, with a value of 0 or 1, where "0" represents no maternal family history of oligomenorrhea and "1" represents a maternal family history of oligomenorrhea; Apolipoprotein-B / Apolipoprotein-A, assigned a value of 0 or 1 according to its index, where "0" represents <0.70 and "1" represents ≥0.70; Fasting insulin, assigned a value of 0 or 1 according to its index, where "0" represents <13.25mIU / mL and "1" represents ≥13.25mIU / mL; Body mass index, which is assigned a value of 0, 1, or 2, where "0" represents <23 kg / m 2 , "1" represents 23kg / m 2 ≦BMI<25kg / m 2 ; "2" represents ≥25kg / m 2 .

3. The model for predicting MetS in PCOS population according to claim 1, characterized in that: The fitting formula of the clinical prediction model for PCOS with MetS is: P=1 / (1+exp(–(0.066×age+0.596×23≦BMI<25+1.531×BMI≥25–0.248×SHBG≥30.25+1.280×BA≥0.700+0.560×maternal history of oligomenorrhea+0.370×acne+1.338×FINS≥13.25–5.4147))).

4. The method for constructing a model suitable for predicting MetS in PCOS population according to any one of claims 1 to 3, comprising: Screen volunteers who meet the diagnostic criteria; Collect the information of volunteers and perform variable data conversion based on the information, the information is selected from: patient's menstrual status, family history, height, weight, WC, HC, SBP, DBP, mF-G score, hirsutism score, acne score, BMI, WHR, FSH, LH, T, DHEAS, AND, SHBG, FPG, FINS, TG, HDL-C, LDL-C, Apo-A, Apo-B; Based on R language, Lasso regression and univariate regression were used to screen predictive variables. Multivariate Logistic regression was performed on the screened predictive variables combined with clinical experience to construct a prediction model for PCOS with MetS.

5. The method for constructing a model suitable for predicting MetS in PCOS population according to claim 4, characterized in that: It also includes the application of the "regplot" package to draw a nomogram for the prediction model of PCOS with MetS.

6. The method for constructing a model suitable for predicting MetS in PCOS population according to claim 4 or 5, characterized in that: Also included is model validation, which is selected from: ROC curve evaluation, calibration curve evaluation, clinical decision curve validation, and clinical impact curve validation.

7. The method for constructing a model suitable for predicting MetS in PCOS population according to claim 6, characterized in that: The ROC curve evaluation includes: drawing the ROC curve, calculating the area under the ROC curve to obtain the C-statistic, when the C-statistic>0.5, the prediction model has predictive ability; when the C-statistic is closer to 1, the discrimination of the model is better.

8. The method for constructing a model suitable for predicting MetS in PCOS population according to claim 6, characterized in that: The calibration curve evaluation includes: drawing the calibration curve, calculating the Brier score of the calibration curve, when Brier<0.25, it proves that the model has predictive ability, and the closer the score is to 0, the better the calibration of the model.

9. The method for constructing a model suitable for predicting MetS in PCOS population according to claim 4, characterized in that: The family history is selected from the history of hypertension and diabetes in first-degree relatives, the history of premature baldness in the paternal family, and the history of oligomenorrhea in the maternal family.

10. Use of the model for predicting MetS in PCOS population according to any one of claims 1 to 3 in risk assessment of the possibility of developing MetS in the future.

Images