Breastfeeding among parous women offered home-visit by a midwife after early discharge following planned cesarean section: Secondary analysis of a randomized controlled trial.

INTRODUCTION: Early discharge holds several advantages and seems safe after planned cesarean section among low-risk women. However, breastfeeding rates are lower after cesarean section. Thus, concern has been raised that early discharge among these women may affect breastfeeding even further. Therefore, we aimed to assess the effect of early discharge the day after planned cesarean section on breastfeeding, among parous women when a home-visit by a midwife was provided the day after discharge. METHODS: We conducted a secondary analysis of a randomized trial. Parous women (n=143) planned for cesarean section were allocated to either discharge within 28 hours after planned cesarean section followed by a home visit the day after (early discharge) or discharge at least 48 hours after planned cesarean section (standard care). The participants filled in questionnaires approximately 2 weeks before delivery and 1 week, 4 weeks, and 6 months postpartum. RESULTS: The proportions of women initiating breastfeeding were 84% versus 87% (early discharge vs standard care). After 6 months, 23% versus 21% were exclusively breastfeeding, while 29% versus 42% were partially breastfeeding. The mean duration of exclusive breastfeeding was 3.4 months (SD=2.3) in both groups. None of these differences was statistically significant. In both groups, the women's breastfeeding self-efficacy score before cesarean section correlated with the duration of breastfeeding. After 4 weeks, low-score rates were 28% versus 30%. CONCLUSIONS: Early discharge with follow-up home visits by a midwife after planned cesarean section in parous women is feasible without compromising breastfeeding.

Effect of early discharge after planned cesarean section on recovery and parental sense of security. A randomized clinical trial.

INTRODUCTION: In some European countries, discharge the day after planned cesarean section has become an accepted procedure. However, little is known about the patients' perception of early discharge. The aim of this study was to compare early discharge with standard care in relation to parental sense of security. Further, we evaluated postoperative pain, mobilization, and readmission. MATERIAL AND METHODS: We performed a randomized clinical trial including parous, singleton pregnant women with a planned cesarean section at term. The women were allocated to either discharge within 28 hours (intervention group) or after 48 hours (standard care group) following the cesarean section. Women discharged within 28 hours after cesarean section were offered a home visit by a midwife the following day. The primary outcome was the postnatal sense of security, which was reported by the woman and her partner in the "Parents' Postnatal Sense of Security" questionnaire 1 week postpartum. Secondary outcomes were pain score, use of analgesics, mobilization, readmission, and contacts with the healthcare system in the postoperative period. RESULTS: We included 143 women, of whom 72 were allocated to the intervention group and 71 were allocated to the standard care group. There were no differences in baseline characteristics. The two groups did not differ concerning the postnatal sense of security for the women (P = .98) or the postnatal sense of security for the partners (P = .38). We found no difference in pain scores, step count, use of analgesics, or number of contacts with the health-care system between the groups. CONCLUSIONS: Parental postnatal sense of security is not compromised by discharge within 28 hours followed by a home visit compared with discharge after 48 hours after planned cesarean section among parous women.

Length of hospital stay after cesarean section in Denmark from 2004 to 2016: A national register-based study.

INTRODUCTION: Length of hospital stay after birth has decreased during the last decades, but nationwide data on length of hospital stay after cesarean section are lacking. Elements of Enhanced Recovery Programs were reported to reduce the length of hospital stay. The aim of this nationwide study was to describe the length of hospital stay after cesarean section in Denmark from 2004 to 2016 taking into account birth- and health-related factors as well as demographic changes and, further, to assess potential differences between the five Danish regions. MATERIAL AND METHODS: Length of hospital stay was assessed in 164 209 deliveries by cesarean section in Denmark from 2004 to 2016. Data were obtained from the Danish National Patient Register. All deliveries by cesarean section at gestational age <22 weeks were excluded. Median length of hospital stay was reported based on crude and adjusted analyses. RESULTS: The median length of hospital stay was significantly reduced by 39 hours (95% confidence interval [CI] 37.9-40.1), from 97 hours (4.0 days) in 2004 to 58 hours (2.4 days) in 2016. Reductions were observed among both planned and emergency cesarean sections. When birth- and health-related factors as well as demographic changes were accounted for, median length of hospital stay was reduced by 30 hours (95% CI 29.3-30.8) in the period. The decrease in length of hospital stay from 2004 to 2016 varied between the five Danish regions, with adjusted reductions between 19 and 46 hours. CONCLUSIONS: A nationwide decrease in length of hospital stay after cesarean section was observed from 2004 to 2016 across all five regions but with significant regional variations. Further studies on the optimal length of hospital stay are needed, especially with regard to implementation of enhanced recovery programs.

Outpatient vs inpatient total laparoscopic hysterectomy: A randomized controlled trial.

INTRODUCTION: The objective of the study was to investigate whether outpatient total laparoscopic hysterectomy (TLH) could be performed as a routine without compromising patient satisfaction. The main outcomes were patient satisfaction with length of hospital stay, quality of life, complications and readmissions, and time to return to work. MATERIAL AND METHODS: A non-blinded prospective randomized controlled trial (Canadian Task Force classification I) performed in a single-center teaching hospital in Denmark. A total of 204 women scheduled for TLH on benign indication were randomized to same-day discharge or overnight stay after TLH. Visual analogue scales (VAS), a validated questionnaire EQ-5D, and a diary were filled in pre- and postoperatively. VAS scores on satisfaction with length of hospital stay and pain were administered taken with the EQ-5D and the diary during a follow up over 4 weeks. Student's t test, Chi-square and non-parametric statistics were used for analysis. The study was registered with ClinicalTrials.gov #NCT02933047. RESULTS: A total of 204 women gave informed consent and 203 underwent surgery (101 outpatient and 102 inpatient women). Complete data were available for 76 women in the outpatient group and 86 women in the inpatient group. There were no differences in baseline characteristics. No group differences were found in satisfaction with length of hospital stay (P = 0.35). The EQ-5D revealed no difference in patient satisfaction. However, one-third in the outpatient group chose overnight stay without medical indication. The groups were comparable in clinical outcomes. Sick leave was longer in the outpatient group regardless of the actual treatment (P = 0.015). CONCLUSIONS: Routine outpatient TLH implies that one-third of the patients stay overnight if this option is available without medical indication. Within this context the procedure can be performed with high patient satisfaction, but may lengthen the time to return to work.

Glycemic Control during Pregnancy-A Predictor of Vitamin C Status at Labor in Type 1 Diabetic Women?

Several experimental studies have suggested that vitamin C (vitC) deficiency during pregnancy may be detrimental to fetal development, and observational studies have shown that vitC status is lower during pregnancy and in people with diabetes. A cross-sectional study in pregnant type 1 diabetic women found that poor maternal vitC status was a significant predictor for obstetric complications of pregnancy when measured within four weeks before labor. The plasma vitC concentration was significantly negatively correlated to HbA1c, the biomarker of glycemic control well-known to be associated with the outcome of the diabetic pregnancy. Here, we evaluated HbA1c during pregnancy in relation to the measured vitC levels in late pregnancy based on data from 46 women from the same cohort. Regression analysis showed that HbA1c of first trimester, the combined mean HbA1c of first and second trimester, mean HbA1c of the whole pregnancy (first, second and third trimester combined), and HbA1c of third trimester alone were all associated with vitC in late pregnancy (p = 0.03, n = 45; p = 0.034, n = 43; p = 0.017, n = 42; and p = 0.008, n = 46, respectively). In third trimester, when adjusted for creatinine clearance, the association between vitC and HbA1c persisted (p = 0.029). Women in third trimester with HbA1c above 7.0% had an increased risk of having poor vitC status compared to women with HbA1c below this level (11 out of 21 vs. 2 out of 25 women, p < 0.001). The results suggest that high HbA1c is associated with poor maternal vitC status and potentially inadequate supply of vitC for the neonate. HbA1c may thus be a relevant substitute biomarker for identifying pregnant women who might benefit from vitC supplementation.

Preeclampsia is associated with increased ambulatory arterial stiffness index in type 1 diabetes mellitus.

INTRODUCTION: Treatment of mild to moderate hypertension might not benefit maternal or fetal outcome. This pessimistic point of view may have come about by using non-validated methods for measuring blood pressure in pregnancy combined with inadequate methodology for diagnosis, treatment, and monitoring effects. AIM: To determine the association between AASI in women with type 1 diabetes mellitus (T1DM) and preeclampsia, and to assess the ability of AASI to diagnose preeclampsia. MATERIAL AND METHODS: Repeated 24-h ambulatory blood pressure recordings were performed three times during pregnancy and once three months postpartum in 151 women with T1DM and 50 control women without diabetes. Circadian rhythm was evaluated as the night day ratio, night blood pressure divided by day blood pressure. RESULTS: Of the T1DM women, 33 developed preeclampsia, which was associated with AASI in the 3rd trimester (p<0.05). The best predictor of preeclampsia in T1DM was an AASI of 0.35. The diurnal blood pressure was significantly higher in all trimesters in women who later had preeclampsia. A flattened circadian rhythm was present in T1DM women with preeclampsia compared to women without preeclampsia (night-day ratio: systole 2nd trimester: 0.94±0.07 vs. 0.91±0.05, women with and without preeclampsia, respectively, p=0.015; diastole 2nd trimester: 0.89±0.07 vs. 0.85±0.07, p=0.003). AASI was higher during pregnancy compared to postpartum in women with T1DM (0.31±0.16, 0.31±0.16 and 0.33±0.18 vs. 0.25±0.17; 1st, 2nd and 3rd trimester vs. postpartum). CONCLUSION: Women with T1DM and preeclampsia demonstrate increased arterial stiffness and had early manifestations in the non-dipping of blood pressure.

Hysterectomy is not associated with de-novo urinary incontinence: A ten-year cohort study.

OBJECTIVE: To determine prevalence, incidence proportion, and changes of urinary incontinence (UI) 10-13 years after hysterectomy compared to two control groups. STUDY DESIGN: A longitudinal cohort study of 661 women with follow-up for ten years. Originally, 866 women answered a questionnaire on continence status preoperatively. Ten years postoperatively the queries were repeated in 371 with a hysterectomy, 89 with laparoscopic cholecystectomy (LC), and 201 with transcervical endometrial resection (TCRE). Significant incontinence was UI at least once a week. The main outcome measures were prevalence and incidence proportions of UI. RESULTS: The overall prevalence of stress UI ten years after surgery was 23% compared to 12% preoperatively. Urge UI was prevalent in 12% compared to 5% preoperatively. Incidence proportions of stress UI were in hysterectomies 21%, in LC 15%, and in TCRE 18%. Similarly, incidence proportions of urge UI were in hysterectomies 11%, in LC 11%, and in TCRE 8%. No significant differences between surgical procedures were found; however, we found substantial amount of changes in continence status from continent to incontinent and vice versa in all three groups. CONCLUSIONS: No significant difference was found after hysterectomy compared to controls in the prevalence or incidence proportions of UI after 10 years follow-up. Hysterectomy is not a risk factor of UI.

Fatigue and physical function after hysterectomy measured by SF-36, ergometer, and dynamometer.

PURPOSE: Information is limited on the early postoperative rehabilitation following hysterectomy. Our purpose was to evaluate the different perioperative modalities of fatigue, pain, quality of life, and physical performance and their time-related. METHODS: A prospective, follow-up study of a cohort of women undergoing abdominal and vaginal hysterectomy at the Gynecology Department at Herning Hospital, Denmark. Data from 108 women with elective hysterectomy were compared pre- and postoperatively. The fatigue level was scored on a visual analogue scale and SF-36. Objective measurements were performed by dynamometer of hand grip, knee extension strength, and postural stability; further, by ergometer cycle work capacity and by impedance lean body mass. Quality of life was assessed using the SF-36 questionnaire. Patients were examined preoperatively and twice postoperatively. RESULTS: Women lost lean body mass 13 and 30 days after their hysterectomy (p < 0.01). Strength in hand (p < 0.05) and knees (p < 0.01) increased compared to preoperative values but no change in postural stability and work capacity was noted. Fatigue resumed to preoperative levels after 30 days. SF-36 revealed that the modality of 'physical functioning' and 'role limitations due to physical problems' remained significantly decreased at the end of the study (p < 0.01) CONCLUSION: Hysterectomy was associated with reduction in physical function assessed by SF-36 30 days after surgery. No impairment of performance was found in physical tests at days 13 and 30 postoperatively.

Recurrence of gestational diabetes in primiparous women.

INTRODUCTION: Gestational diabetes mellitus (GDM) increases the risk for diabetes in the next pregnancy and later in life. Thus, estimating the risk of GDM in further pregnancies provides a time frame for possible preventive measures. We aimed to calculate the recurrence rate of GDM in primiparous women and evaluate the factors involved such as age, body mass index, weight gain, time between pregnancy and postpartum OGTT results. MATERIAL AND METHODS: We established a prospective cohort during a 5-year period at the Department of Obstetrics at Kolding Hospital. Women with diet-treated GDM in their first pregnancy and a subsequent pregnancy constituted our study population. Multiparity and insulin-treated GDM were exclusion criteria. RESULTS: Among 15 735 deliveries, 535 women were diagnosed with GDM (3.4%). Of these, 209 (39.1%) were nulliparous women, treated with diet only. Seventy-two of these women had a subsequent pregnancy and they all attended the recommended screening procedure, a 75-g oral glucose tolerance test at 14-20 (early) or 27-30 (late) weeks' gestation. The recurrence rate of GDM was 47.2%. The risk of recurrence was less in women who lost weight between the first and the subsequent pregnancy. CONCLUSIONS: Recurrence of diet-treated GDM was 47.2% in primiparous women with previous GDM and the recurrence was associated with weight gain between pregnancies.

Impact of metformin on anti-Müllerian hormone in women with polycystic ovary syndrome: a secondary analysis of a randomized controlled trial.

Conclusions on the effect of metformin on circulating anti-Müllerian hormone (AMH) levels in women with polycystic ovary syndrome (PCOS) are ambiguous. We performed a secondary analysis of a randomized, double-blind, placebo-controlled cross-over trial. Fifty-six women with hyperandrogenemic PCOS were included. Each woman served as her own control receiving a daily dose of either 1700 mg metformin or placebo for 6 months. After a 3-month wash-out period they received the opposite treatment. The decrease in AMH from a median of 49.5 to 46.9 pmol/L after 6 months on metformin was overall not significant (p = 0.81), nor were changes in obese women (from 49.5 to 38.2 pmol/L; p = 0.53). Comparing individual metformin/placebo AMH values, a small absolute decrease of 9.3 pmol/L (p = 0.03) was observed in obese women after 6 months relative to baseline, suggesting a trend towards decreasing values after metformin treatment, mainly in obese women.

Less extensive surgery compared to extensive surgery: survival seems similar in young women with adult ovarian granulosa cell tumor.

OBJECTIVE: To describe the outcome of adult granulosa cell tumor (AGCT) with respect to initial clinical findings, methods of surgery, and perioperative treatment. STUDY DESIGN: Retrospective follow-up study. SETTING: All hospitals in Jutland. SAMPLE: 163 women diagnosed with AGCT. METHODS: Follow-up by hospital data files, general practitioner, death certificate, and autopsy report. Revision of histopathology by a single pathologist. MAIN OUTCOME MEASURES: Survival and relapse by clinical data, stage, and type of surgery. RESULTS: The incidence of AGCT was 1.37 per year per 100,000 women (95% CI: 1.08, 1.68). The median follow-up time was 15 years and for the 79 surviving women 22 years. Stage I was found in 94% of cases. Relapse occurred in 24% of women in stage I and 100% of the other stages. Survival in stage I was 95%, 89% and 84% after 5, 10 and 20 years respectively. Increased survival of stage I in postmenopausal women was associated with surgery including hysterectomy and bilateral oophorectomy (p<0.001). In women younger than 40 years no difference in survival was found due to type of surgery. Endometrial carcinoma was found 138 times (95% CI: 48, 275) more prevalent than the expected rate. CONCLUSION: The survival of women was better in AGCT than in epithelial ovarian tumor. Age and type of surgery, besides stage, influenced survival. Total abdominal hysterectomy and bilateral salpingo-oophorectomy is the recommended treatment with advancing age. At younger age less extensive surgery was associated with similar survival compared to extensive surgery, but with advancing age conservative surgery increased the risk of relapse and death.

Ovarian granulosa cell tumor and increased risk of breast cancer.

Granulosa cell tumor of the ovary (GCT) is a rare neoplasm. The tumor often secretes estrogens and then presents at an earlier stage due to hormone-related symptoms. GCT women are at increased risk of endometrial carcinoma, but there is only limited information about GCTs and potential association to other hormone-related neoplasms such as breast cancer. We conducted a retrospective follow-up study on 163 women with GCT. Medical records and histological sections were reviewed and a search in the pathology registry performed. Eight [95% confidence interval (CI); 3.4-15.8] GCT women were diagnosed with a breast neoplasm; one with Paget's disease of the nipple and seven with breast carcinoma. Based on calculations using incidence rates on breast cancer among Danish women, we would have expected 2.5 cases of breast cancer. The odds ratio was 3.3 (95% CI, 1.6-6.6), suggesting an increased risk of breast cancer in GCT women.

Metformin lowers serum cobalamin without changing other markers of cobalamin status: a study on women with polycystic ovary syndrome.

Treatment with the anti-diabetic drug metformin is followed by a decline in plasma cobalamin, but it is unsettled whether this denotes an impaired cobalamin status. This study has explored changes in the markers of cobalamin status in women with Polycystic Ovary Syndrome treated with metformin (1.5-2.5 g per day) (n = 29) or placebo (n = 23) for six months. Serum samples were collected before and after two, four, and six months of treatment. We found serum cobalamin to decline and reach significant lower levels after six months of treatment (p = 0.003). Despite the decline in serum cobalamin, we observed no reductions in the physiological active part of cobalamin bound to transcobalamin (holotranscobalamin), or increase in the metabolic marker of cobalamin status, methylmalonic acid. Instead, the non-functional part of circulating cobalamin bound to haptocorrin declined (p = 0.0009). Our results have two implications: The data questions whether metformin treatment induces an impaired cobalamin status in PCOS patients, and further suggests that serum cobalamin is a futile marker for judging cobalamin status in metformin-treated patients.

Ambulatory arterial stiffness index in type 1 diabetes mellitus: any different during pregnancy?

OBJECTIVE: To analyze the ambulatory arterial stiffness index (AASI) and pulse pressure (PP) during pregnancy and 3 months after delivery in type 1 diabetes mellitus (T1DM) and compare it to healthy pregnant controls. STUDY DESIGN: Prospective, descriptive study of 59 women with T1DM and 42 non-diabetic women. Blood pressure was measured using a portable oscillometry monitor and AASI was calculated as 1 minus the regression slope of diastolic on systolic blood pressure obtained from 24-h monitoring. Main outcome measures were comparisons of the AASI and PP between T1DM women and controls examined during pregnancy, and of the AASI and PP during and after pregnancy in T1DM women. RESULTS: PP and AASI were higher at all times during pregnancy in T1DM compared to postpartum (p<0.01). AASI and PP were significantly associated with albumin excretion rate when adjusting for retinopathy, preeclampsia, duration of diabetes, HbA1c, age, and BMI. The AASI was positively correlated with night-day ratio in the 1st and 3rd trimesters during pregnancy. No difference was found in AASI compared with non-diabetic controls during pregnancy. CONCLUSIONS: AASI and PP increased during diabetic pregnancy and were associated with the women's albuminuria grade.

[Research at regional hospitals in Denmark can be reinforced]. / Forskning på regionshospitaler kan styrkes.

We examined the conditions for health research at regional hospitals in Denmark. The study was conducted as an interview-based case study containing interviews with key persons at the governing level as well as medical doctors at the operating level. The results showed that the settings and opportunities regarding health research vary between university hospitals and non-university regional hospitals. However our findings indicate, that focusing on a long-term strategy to implement health research in regional hospitals will enable these to be more active in health research.

A substantial number of scientific publications originate from non-university hospitals.

INTRODUCTION: As we found no recent published reports on the amount and kind of research published from Danish hospitals without university affiliation, we have found it relevant to conduct a bibliometric survey disclosing these research activities. MATERIAL AND METHODS: We retrieved all scientific papers published in the period 2000-2009 emanating from all seven Danish non-university hospitals in two regions, comprising 1.8 million inhabitants, and which were registered in a minimum of one of the three databases: PubMed MEDLINE, Thomson Reuters Web of Science and Elsevier's Scopus. RESULTS: In 878 of 1,252 papers, the first and/or last author was affiliated to a non-university hospital. Original papers made up 69% of these publications versus 86% of publications with university affiliation on first or last place. Case reports and reviews most frequently had authors from regional hospitals as first and/or last authors. The total number of publications from regional hospitals increased by 48% over the 10-year period. Publications were cited more often if the first or last author was from a university hospital and even more so if they were affiliated to foreign institutions. Cardiology, gynaecology and obstetrics, and environmental medicine were the three specialities with the largest number of regional hospital publications. CONCLUSION: A substantial number of scientific publications originate from non-university hospitals. Almost two thirds of the publications were original research published in international journals. Variations between specialities may reflect local conditions. FUNDING: not relevant. TRIAL REGISTRATION: not relevant.

Transfusion rate and prevalence of unexpected red blood cell alloantibodies in women undergoing hysterectomy for benign disease.

OBJECTIVE: To determine transfusion rates, risk factors for transfusion and the prevalence of unexpected red blood cell alloantibodies in women undergoing hysterectomy for benign disease. In addition, we aimed to evaluate the necessity of the pretransfusion testing for red blood cell alloantibodies. DESIGN: Retrospective cohort study. SETTING: The Danish Hysterectomy Database and a regional computerized blood bank register. POPULATION: The 4 181 hysterectomies in 2004 reported to the Hysterectomy Database. The blood bank registers 2 603 hysterectomies performed between 1997 and 2005. METHODS: From the hysterectomy database, information about indications for the hysterectomy, surgical procedures, re-operations, number of blood transfusions, and demographic, descriptive and clinical characteristics were extracted. Urgency of the transfusion episodes was evaluated by a retrospective review of the patients' medical records. From the regional blood bank register, results of the screening for red blood cell alloantibodies were extracted. MAIN OUTCOME MEASURES: Transfusion rates, prevalence of unexpected red blood cell alloantibodies. RESULTS: In all, 242 women (5.8%) received blood transfusions, but only 32 of the 4 181 women (0.77%) were urgently transfused. Re-operations were frequently associated with urgent blood transfusions. Nine of the 2 603 women from the regional register (0.35%) had newly detected, clinically significant red blood cell alloantibodies. CONCLUSIONS: The risk of a hemolytic transfusion reaction was estimated to be less than 1 in 17 000 hysterectomies (upper confidence limit) if the routine pretransfusion test were to be omitted. We suggest that reconsideration of the necessity for routine preoperative pretransfusion testing for women undergoing hysterectomy for benign disease is indicated.

Preterm delivery in normoalbuminuric, diabetic women without preeclampsia: the role of metabolic control.

OBJECTIVE: The aim of this study was to examine the importance of glycemic regulation on the risk of preterm delivery in women with normoalbuminuria and no preeclampsia later in pregnancy. STUDY DESIGN AND METHODS: A prospective study of 71 women with type 1 diabetes mellitus where complete data were collected on HbA1c, insulin dose, and albumin excretion rate from week 12 and every second week hereafter. Fundus photography was performed and diurnal blood pressure measured three times during pregnancy. RESULTS: The preterm rate was 23% and women delivering preterm showed higher HbA1c throughout pregnancy. At regression analysis HbA1c was the strongest predictor for preterm delivery from week 6 to 32, also when including insulin dose, BMI, age, duration of diabetes, and diurnal blood pressure. The risk of delivering preterm was more than 40% when HbA1c was above 7.7% in week 8. Diurnal blood pressure was not found associated with preterm delivery. CONCLUSION: The quality of glycemic regulation in the early and mid-pregnancy is a major, independent risk factor for preterm delivery in normoalbuminuric diabetic women without preeclampsia.

Increased serum IGF-I during pregnancy is associated with progression of diabetic retinopathy.

The IGF system has been associated with development and progression of diabetic retinopathy. We examined whether a simple measurement of the IGF system (serum total IGF-I) correlated with progression of diabetic retinopathy in pregnancy in type 1 diabetes. A prospective observational study was performed in 103 pregnant women with type 1 diabetes. Serum IGF-I was measured in maternal serum from week 14, every fourth week until week 30, and every second week until delivery. Twenty-four-hour blood pressure was measured with a portable oscillometry monitor. The women had visual acuity testing and fundus photography before pregnancy, once in each trimester, and 4 months after birth. Each eye was assigned an overall retinopathy grade on a scale from 1 to 6 independently by two experienced graders. During pregnancy, serum IGF-I increased with increasing gestational age until a plateau was reached in week 32. Progression of retinopathy was significantly associated with a higher level of IGF-I (P < 0.01). Serum IGF-I increased with increasing progression of retinopathy. Change of retinopathy was significantly associated with level of IGF-I (P < 0.01). During pregnancy, serum IGF-I increased with increasing birth weight until a plateau was reached in week 32. Birth weight was significantly associated with a higher level of serum IGF-I (P < 0.01).