Baseline levels of circulating galectin-1 associated with radiographic hand but not radiographic knee osteoarthritis at a two-year follow-up

Objective

We tested the potential of circulating galectin-1, interleukin (IL)-1 beta, IL-6, and tumour necrosis factor alpha (TNF alpha) levels at baseline in individuals with knee pain as biomarkers for development of radiographic knee and/or hand osteoarthritis (OA).

Design

This study comprised 212 individuals with knee pain from the Halland osteoarthritis cohort (HALLOA). Clinical characteristics and serum/plasma levels of galectin-1, IL-1 beta, IL-6, and TNF alpha were measured at baseline, and knee and hand radiographs were obtained at a two-year follow-up. The predictive value of circulating inflammatory markers and clinical variables at baseline was assessed using multinominal logistic regression for those who developed radiographic OA in knees only (n ​= ​25), in hands only (n ​= ​40), and in both knees and hands (n ​= ​43); the group who did not develop OA (n ​= ​104) was used as reference. Correlations were assessed using Spearman's correlation coefficients.

Results

As expected, age was identified as a risk factor for having radiographic knee and/or hand OA at the two-year follow-up. Baseline circulating galectin-1 levels did not associate with developing radiographic knee OA but associated with developing radiographic hand OA (odds ratio (OR) for a 20% increased risk: 1.14, 95% confidence interval (CI) 1.01-1.29) and both radiographic knee and hand OA (OR for a 20% increased risk: 1.18, 95% CI 1.05-1.30). However, baseline IL-1 beta, IL-6, and TNF alpha did not associate with developing radiographic knee and/or hand OA.

Conclusion

Non-age adjusted circulating galectin-1 is superior to IL-6, IL-1 beta, and TNF alpha in predicting radiographic hand but not knee OA.

Comparison of Iopromide versus Iohexol in Aortobifemoral Arteriography

A double-blind randomized, clinical trial was conducted in 9 hospitals comparing the use of non-ionic contrast media (CM) iopromide 300 (Ultravist) and iohexol 300 (Omnipaque) during peripheral arteriography in a total of 446 patients. After premedication with morphine-scopolamine each patient was given two consecutive injections of 50 ml CM at a rate of 12 ml/s above the aortic bifurcation. Both CM were well tolerated. There were no differences between the two substances as far as general tolerance, pulse rate, blood pressure, sensation of heat or pain after CM injection were concerned.

Decreased permeability surface area for glucose in obese women with postprandial hyperglycemia: no effect of phosphodiesterase-5 (PDE-5) inhibition

Insulin-mediated microvascular recruitment is recognized as a potential mechanism contributing to insulin resistance. In this study, we compared a marker of microvascular function, the permeability surface area for glucose (PS(glu)), and forearm glucose uptake after an OGTT in obese women with impaired glucose metabolism and healthy lean nondiabetic women, with the aim to characterize whether decreased permeability surface area for glucose or decreased glucose uptake may contribute to postprandial hyperglycemia in the obese group. In addition, we evaluated whether the phosphodiesterase-5 (PDE-5) inhibitor tadalafil, in a randomized double blind placebo controlled design, might attenuate postprandial glucose levels in obese women. For these purposes, intramuscular microdialysis, blood sampling from arterial and venous blood of the forearm, and measurements of forearm blood flow were performed. The results showed an impaired permeability surface area for glucose (IAUC PS(glu) 31±13 vs. 124±31; p<0.05) in obese when compared with lean participants, but no differences in forearm glucose uptake appeared between the groups. Furthermore, a single dose of tadalafil 10 mg showed no improvement of the permeability surface area for glucose, glucose uptake, or circulating glucose levels in obese participants. In conclusion, the postprandial PS(glu) response was impaired in obese women showing postprandial hyperglycemia, indicating a compromised microcirculation. However, we were unable to demonstrate any acute effect on either vascular function or glucose uptake of the phosphodiesterase-5 (PDE-5) inhibitor tadalafil.

Impaired delivery of insulin to adipose tissue and skeletal muscle in obese women with postprandial hyperglycemia

CONTEXT

An impaired transfer of insulin from the circulation to the interstitial fluid has been suggested to contribute to insulin resistance.

OBJECTIVE

The objective of the study was to address whether the delivery of insulin from the circulation to adipose tissue and skeletal muscle is impaired in obese women with postprandial hyperglycemia compared with lean healthy controls.

DESIGN, SETTING, AND PARTICIPANTS

Seven obese nondiabetic women with postprandial hyperglycemia and nine lean healthy women were recruited. The interstitial insulin concentration in adipose tissue and muscle tissue was measured by the microdialysis technique during an oral glucose tolerance test (75 g). In parallel, arterial insulin levels were measured. We used ¹³³Xe clearance and plethysmography to monitor blood flow. Subcutaneous needle biopsies were taken to obtain fat cell size.

RESULTS

After oral glucose ingestion, mean arterial insulin levels were higher in obese women than in the lean group. However, interstitial insulin levels in sc adipose tissue and forearm muscle were similar in both groups. Accordingly, the proportion of circulating insulin being transported across the vascular endothelium to the interstitial fluid in adipose tissue and in muscle tissue was significantly lower in the obese compared with the lean group.

CONCLUSIONS

Obese subjects with postprandial hyperglycemia need higher circulating insulin levels than lean controls to attain similar interstitial insulin levels in adipose tissue and skeletal muscle, indicating an impaired transfer of insulin across the endothelium.

Tadalafil increases muscle capillary recruitment and forearm glucose uptake in women with type 2 diabetes

AIMS/HYPOTHESIS

Recent evidence suggests that reduced synthesis of nitric oxide in endothelial cells, i.e. endothelial dysfunction, contributes to the impaired action of insulin in the vasculature of patients with type 2 diabetes. We investigated whether selective inhibition of phosphodiesterase-5 by tadalafil has beneficial effects on peripheral microcirculation and glucose uptake in these patients.

METHODS

We enrolled seven postmenopausal women with type 2 diabetes and ten age-matched healthy women as controls in a placebo-controlled study to evaluate the acute metabolic effects of tadalafil. We performed microdialysis and blood flow measurements in muscle, and sampled arterial and deep venous blood before and after a single dose of tadalafil 20 mg or placebo. Circulating glucose and insulin levels, muscle capillary recruitment as reflected by permeability surface area for glucose (PS(glu)) and forearm glucose uptake were measured.

RESULTS

In women with type 2 diabetes, but not in the control group, tadalafil induced increases in the incremental AUC for PS(glu) (tadalafil vs placebo 41 +/- 11 vs 4 +/- 2 ml [100 g](-1) min(-1), p < 0.05) and forearm glucose uptake (46 +/- 9 vs 8 +/- 4 micromol [100 g](-1) min(-1), p < 0.05). The variable that best predicted forearm glucose uptake was PS(glu), which explained 70% of its variance. However, fasting glucose and insulin concentrations were similar following treatment with placebo or tadalafil in the two groups.

CONCLUSIONS/INTERPRETATION

This study suggests that tadalafil evokes positive metabolic effects in insulin-resistant women with type 2 diabetes.

Laser-Doppler flowmetry reveals rapid perfusion changes in adipose tissue of lean and obese females

The present study aimed to evaluate adipose tissue blood flow (ATBF) by means of laser-Doppler flowmetry (LDF) in humans. Lower body negative pressure (LBNP) and straining known to affect epidermal blood flow through the autonomic nervous system were performed in 11 lean and 11 obese female volunteers. ATBF changes were compared between both groups and also discriminated from skin blood flow (SBF) responses of the immediate vicinity. Additionally, LDF measurements were compared with flow measurements using (133)xenon washout in 10 lean subjects during whole body cooling. LDF estimations of SBF and ATBF showed a positive correlation to (133)Xe during cooling. SBF and ATBF were reduced to the same extent in both lean and obese subjects during LBNP. Straining induced divergent changes in SBF and ATBF: initially SBF decreased while ATBF increased, but toward the end of straining SBF increased above baseline and ATBF returned down to baseline level. Those changes were similar in both weight groups. Interestingly, only in obese subjects, both LBNP and straining were followed by ATBF augmentation, while SBF levels remained stable. In conclusion, LDF compares with (133)Xe washout in monitoring ATBF during tonic perfusion changes. Its strength, however, lies in the detection of rapid flow alterations within the subcutaneous tissue, allowing the evaluation of reflex responses of the subcutaneous microcirculation. Interestingly, those rapid changes in SBF and ATBF can be both concordant and discordant. With regard to ATBF, vasoconstrictor components of the reflex responses were similar in lean and obese subjects, whereas vasodilatory responses were more pronounced in obese volunteers.

Regulation and counterregulation of lipolysis in vivo: different roles of sympathetic activation and insulin

To obtain further information on the regulation of lipolysis in vivo, the effect of increasing sympathetic nerve activity via lower body negative pressure (LBNP, -20 mm Hg) was studied in 11 healthy human subjects. Subcutaneous and muscle microdialysis as well as blood flow measurements were performed in the postabsorptive state and during an euglycemic hyperinsulinemic clamp. LBNP for 30 min in the postabsorptive phase resulted in an approximately 50% increase (P < 0.005) in the interstitial-arterial concentration difference for glycerol in adipose tissue, whereas no such effect was registered in muscle. Blood flow in adipose tissue and the forearm remained unaltered. During euglycemic hyperinsulinemic conditions (p-insulin 645 +/- 62 pmol/liter), both interstitial adipose tissue and arterial concentrations of glycerol were reduced. LBNP resulted in an increase in interstitial-arterial concentration difference in glycerol similar to that seen in the postabsorptive state (approximately 50%, P < 0.05). Muscle glycerol was not changed by either insulin or LBNP. Glucose infusion rate during the clamp was significantly decreased during LBNP (7.82 +/- 0.88 vs. 8.67 +/- 1.1 ml/kg.min, P < 0.05). We conclude that the sympathetic nervous activation by LBNP results in an increased lipolysis rate in adipose tissue both in the postabsorptive phase and during insulin infusion. On the other hand, muscle glycerol output was not affected by either LBNP or insulin. The data suggest that 1) lipolysis is regulated differently in muscle and adipose tissue, 2) postabsorptive lipolysis is mainly regulated by insulin, and 3) sympathetic nervous activation effectively inhibits the antilipolytic action of insulin by inducing insulin resistance.

Human adipose tissue glucose uptake determined using [(18)F]-fluoro-deoxy-glucose ([(18)F]FDG) and PET in combination with microdialysis

AIMS/HYPOTHESIS

To determine the lumped constant (LC), which accounts for the differences in the transport and phosphorylation between [(18)F]-2-fluoro-2-deoxy-d-glucose ([(18)F]FDG) and glucose, for [(18)F]FDG in human adipose tissue.

METHODS

[(18)F]FDG-PET was combined with microdialysis. Seven non-obese (29 +/- 2 years of age, BMI 24 +/- 1 kg/m2) and seven obese (age 32 +/- 2 years of age, BMI 31 +/- 1 kg/m2) men were studied during euglycaemic hyperinsulinaemia (1 mU/kg. min-1 for 130 min). Abdominal adipose tissue [(18)F]FDG uptake (rGUFDG) and femoral muscle glucose uptake were measured using [(18)F]FDG-PET. Adipose tissue perfusion was measured using [15O]-labelled water and PET, and interstitial glucose concentration using microdialysis. Glucose uptake (by microdialysis, rGUMD) was calculated by multiplying glucose extraction by regional blood flow. The LC was determined as the ratio of rGUFDG to rGUMD.

RESULTS

Rates of adipose tissue glucose uptake (rGUMD) were 36 % higher in the non-obese than in the obese patients (11.8 +/- 1.7 vs 7.6 +/- 0.8 micromol/kg. min-1, p < 0.05, respectively) and a correlation between rGUMD and rGUFDG was found (r = 0.82, p < 0.01). The LC averaged 1.14 +/- 0.11, being similar in the obese and the non-obese subjects (1.01 +/- 0.15 vs 1.26 +/- 0.15, respectively, NS). Muscle glucose uptake was fourfold to fivefold higher than adipose tissue glucose uptake in both groups.

CONCLUSION/INTERPRETATION

[(18)F]FDG-PET seems a feasible tool to investigate adipose tissue glucose metabolism in human beings. Direct measurements with [(18)F]FDG-PET and microdialysis suggest a LC value of 1.14 for [(18)F]FDG in human adipose tissue during insulin stimulation and the LC does not appear to be altered in insulin resistance. Furthermore, the obese patients show insulin resistance in both adipose tissue and skeletal muscle.

Lumped constant for [(18)F]fluorodeoxyglucose in skeletal muscles of obese and nonobese humans

Quantitative 2-[(18)F]fluoro-2-deoxy-D-glucose ([(18)F]FDG) positron emission tomography (PET) has been widely used to calculate glucose utilization in skeletal muscle. FDG-PET results depend partly on the lumped constant (LC), which accounts for the differences in the transport and phosphorylation between [(18)F]FDG and glucose. In this study, we estimated the LC for [(18)F]FDG directly in normal and in insulin-resistant obese subjects by combining FDG PET with the microdialysis technique. Eight obese [age 29.4 +/- 1.0 yr, body mass index (BMI) 33.6 +/- 1.0 kg/m(2)] and eight nonobese (age 25.0 +/- 1.0 yr, BMI 23.1 +/- 1.0 kg/m(2)) males were studied during euglycemic hyperinsulinemia (1 mU. kg(-1).min(-1) for 150 min). Muscle blood flow was measured using (15)O-labeled water and PET. Muscle [(18)F]FDG uptake (rGU(FDG)) was calculated with Patlak graphic analysis. Interstitial glucose concentration of the quadriceps femoris muscle was measured simultaneously with [(18)F]FDG scanning using microdialysis. Muscle glucose uptake (by microdialysis, rGU(MD)) was calculated by multiplying glucose extraction by regional muscle blood flow. A significant correlation was found between rGU(MD) and rGU(FDG) (r = 0.78, P < 0.01). The LC was determined as the ratio of the rGU(FDG) to the rGU(MD). The LC averaged 1.16 +/- 0.16 and was similar in the obese and nonobese subjects (1.15 +/- 0.11 vs. 1.16 +/- 0.07, respectively, not significant). In conclusion, the microdialysis technique can be reliably combined with FDG PET to measure glucose uptake in skeletal muscle. Direct measurements with these two independent techniques suggest an LC value of 1.2 for [(18)F]FDG in human skeletal muscle during insulin stimulation, and the LC appears not to be sensitive to insulin resistance.

Estimations of muscle interstitial insulin, glucose, and lactate in type 2 diabetic subjects

Previous measurement of insulin in human muscle has shown that interstitial muscle insulin and glucose concentrations are approximately 30-50% lower than in plasma during hyperinsulinemia in normal subjects. The aims of this study were to measure interstitial muscle insulin and glucose in patients with type 2 diabetes to evaluate whether transcapillary transport is part of the peripheral insulin resistance. Ten patients with type 2 diabetes and ten healthy controls matched for sex, age, and body mass index were investigated. Plasma and interstitial insulin, glucose, and lactate (measured by intramuscular in situ-calibrated microdialysis) in the medial quadriceps femoris muscle were analyzed during a hyperinsulinemic euglycemic clamp. Blood flow in the contralateral calf was measured by vein plethysmography. At steady-state clamping, at 60-120 min, the interstitial insulin concentration was significantly lower than arterial insulin in both groups (409 +/- 86 vs. 1,071 +/- 99 pmol/l, P < 0.05, in controls and 584 +/- 165 vs. 1, 253 +/- 82 pmol/l, P < 0.05, in diabetic subjects, respectively). Interstitial insulin concentrations did not differ significantly between diabetic subjects and controls. Leg blood flow was significantly higher in controls (8.1 +/- 1.2 vs. 4.4 +/- 0.7 ml. 100 g(-1).min(-1) in diabetics, P < 0.05). Calculated glucose uptake was less in diabetic patients compared with controls (7.0 +/- 1.2 vs. 10.8 +/- 1.2 micromol. 100 g(-1).min(-1), P < 0.05, respectively). Arterial and interstitial lactate concentrations were both higher in the control group (1.7 +/- 0.1 vs. 1.2 +/- 0.1, P < 0. 01, and 1.8 +/- 0.1 vs. 1.2 +/- 0.2 mmol/l, P < 0.05, in controls and diabetics, respectively). We conclude that, during hyperinsulinemia, muscle interstitial insulin and glucose concentrations did not differ between patients with type 2 diabetes and healthy controls despite a significantly lower leg blood flow in diabetic subjects. It is suggested that decreased glucose uptake in type 2 diabetes is caused by insulin resistance at the cellular level rather than by a deficient access of insulin and glucose surrounding the muscle cell.

Muscle glucose uptake is effectively activated by ischemia in type 2 diabetic subjects

It has previously been shown that Wortmannin, a phosphatidylinositol 3-kinase inhibitor, inhibits glucose transport activated by insulin but not by ischemia, suggesting the importance of an activating mechanism that bypasses the insulin signal. To evaluate the relevance of this insulin-independent pathway in insulin-resistant subjects, the ability of ischemia to stimulate glucose uptake was investigated in 9 patients with type 2 diabetes and in 9 healthy control subjects (fasting glucose level 9.4 +/- 0.8 vs. 5.1 +/- 0.1 mmol/l, P < 0.001, in type 2 diabetic patients and control subjects, respectively; fasting insulin level insulin 8.1 +/- 2.6 vs. 4.5 +/-0.7 mU/l, P < 0.05, respectively) matched for sex, age, and BMI. Arterial plasma and interstitial concentrations of glucose and lactate (measured by subcutaneous and muscle microdialysis) were recorded in the forearm before, during, and after ischemia induced locally for 20 min. During ischemia, the muscle interstitial glucose concentration decreased significantly from 7.7 +/- 0.6 to 5.4 +/- 0.4 mmol/l (P < 0.01) and from 4.4 +/- 0.3 to 3.6 +/- 0.3 mmol/l (P < 0.05) in type 2 diabetic patients and control subjects, respectively. The arterial-interstitial (A-I) glucose concentration difference was 1.7 +/- 0.6 and 0.7 +/- 0.3 mmol/ at basal, and it increased significantly to 3.5 +/- 0.7 (P < 0.01) and 1.4 +/-0.3 mmol/l (P < 0.05) during ischemia in each group, respectively. Interstitial lactate increased significantly during ischemia from 0.8 +/- 0.1 to 1.1 +/- 0.1 mmol/l (P < 0.05) and from 0.5 +/- 0.1 to 0.9 +/- 0.2 mmol/l (P < 0.05), respectively. The A-I glucose concentration difference was abolished immediately postischemia and regained after approximately 15 min, whereas high interstitial lactate levels remained elevated throughout the study. Subcutaneous interstitial glucose concentrations remained unchanged during ischemia and postischemia in both groups, whereas the interstitial lactate concentration in adipose tissue increased during ischemia from 1.4 +/- 0.2 to 2.0 +/- 0.2 mmol/l (P < 0.05) and from 1.1 +/- 0.1 to 1.8 +/- 0.3 mmol/l (P < 0.05) in type 2 diabetic patients and control subjects, respectively. Plasma glucose and lactate levels were unchanged in both groups during the study period. The results show that in muscle, but not in adipose tissue, glucose uptake is efficiently activated by ischemia in insulin-resistant type 2 diabetic subjects, suggesting the activation of a putative alternative pathway to the insulin signal in muscle cells.

Validation of an endogenous reference technique for the calibration of microdialysis catheters

In order to validate the use of urea as an internal reference for the calibration of microdialysis catheters, both subcutaneous and muscle measurements were taken in 11 healthy volunteers. The equilibration calibration of urea, glucose and lactate, and the internal reference (3H-glycerol retrodialysis) calibration of glycerol were performed in both the muscle and fat. These data were then compared with those taken from the arterial plasma. Based on the average arterial-interstitial fluid concentration difference, the urea microdialysis relative recovery was estimated from dialysate and plasma content of urea. Interstitial urea was then used as an endogenous reference for the calculation of the relative recovery of each substance. This was carried out after defining the relative recovery ratios (RR) that were obtained from in vitro measurements on crystalline solutions with the assumption that RR in vivo = RR in vitro. The data demonstrate that the interstitial fluid concentrations of glucose, lactate and glycerol can be calculated from the knowledge of plasma and microdialysate content of urea together with the RR in vitro. Also, the data that were obtained using urea as an endogenous reference showed a significant correlation with those that were achieved using either equilibration or internal reference (retrodialysis) techniques for calibrating the microdialysis devices.

Reduced bleeding through temporary balloon occlusion in hip and knee revision surgery

We used temporary balloon occlusion of the iliac or femoral artery to reduce blood loss in major hip or knee operations in 15 cases in 13 patients. The balloon was introduced by an interventional radiologist in the afternoon of the day before surgery or in the morning before. A latex occlusion balloon was inserted via the transfemoral ipsi- or contralateral route. The patients received two 40 mg doses of low molecular weight heparin. At the beginning of the operation, saline was injected into the predetermined volume to inflate the balloon. The balloons were inflated during 1-6 hours. In each case, the balloon occlusion clearly reduced bleeding in the surgical field and facilitated surgery. The perioperative bleeding was reduced by half, as compared to a retrospective control group. We measured the intraarterial blood pressures distally to the balloon in 2 patients. They decreased from 120 to 40 mm Hg and 155 to 50 mm Hg, respectively, after inflation. Two complications occurred, one bleeding due to catheter dislocation the night before surgery and one postoperative necrosis of the tip of a toe in a patient in whom the deflated balloon was not extruded until the day after surgery and the dose of heparin was too low.

Validation of the 'internal reference technique' for calibrating microdialysis catheters in situ

In vivo calibration of microdialysis catheters with [3H]glucose as internal reference was done in rat (n = 17) and human (n = 12) subcutaneous tissue. The estimated interstitial glucose level was compared with the glucose concentration in venous plasma which, in turn, has been shown to be identical to the interstitial glucose concentration. In subcutaneous tissue of anaesthetized male Sprague-Dawley rats, interstitial glucose was significantly overestimated (43%, P < 0.005, n = 8, and 19%, P < 0.005, n = 9, in normoglycaemic and hyperglycaemic animals, respectively). Furthermore, fractional outflux of [3H]glucose decreased continuously during prolonged perfusion of the microdialysis catheter. In contrast, in human subcutaneous tissue microdialysed with two catheters, correct measurements of interstitial glucose could be achieved and the precision was comparable to that obtained with equilibration calibration in vivo. The average relative error of the mean result of two catheters was 8.9% at a perfusate flow rate of 1 microL min-1. It may be suggested that calibration in vivo of microdialysis catheters with internal references may be used in human subcutaneous tissue. However, it is necessary to validate the calibration technique in each different tissue under reproducible experimental conditions since accumulation of the reference substance in the tissue may create artefactual results.

Comparison of iopromide versus iohexol in aortobifemoral arteriography. A Swedish multi-center study of 446 patients

A double-blind randomized, clinical trial was conducted in 9 hospitals comparing the use of non-ionic contrast media (CM) iopromide 300 (Ultravist) and iohexol 300 (Omnipaque) during peripheral arteriography in a total of 446 patients. After premedication with morphine-scopolamine each patient was given two consecutive injections of 50 ml CM at a rate of 12 ml/s above the aortic bifurcation. Both CM were well tolerated. There were no differences between the two substances as far as general tolerance, pulse rate, blood pressure, sensation of heat or pain after CM injection were concerned.

Autotransplanted teeth with early-stage endodontic treatment: a radiographic evaluation

Autotransplantation of teeth has been performed for many years and with varying degrees of success. The present report concerns the long-term results of autotransplantation of 36 teeth with completed root formation. Endodontic treatment was started before or within 12 weeks after transplantation. At the final check-up 32 teeth were present, 24 of them without signs of inflammation, replacement, or internal resorption. Careful radiographic examination during the follow-up period and adequate therapeutic measures with inter alia renewed endodontic treatment are of importance for successful autotransplantation.

Female work capacity during the menstrual cycle: physiological and psychological reactions

Changes in physical and mental work capacity during the menstrual cycle were studied in 12 healthy woemen with severe menstrual distress. Physiological and psychological tests were performed before, during, and after menstruation. Heart rate, pulmonary ventilation, oxygen uptake, blood lactate concentration, and perceived exertion were measured during work on a bicycle ergometer ar two submaximal work loads corresponding to 40 and 70 %, respectively, of individual maximal oxygen uptake. Mental work capacity was studied with the aid of performance tests of psychological functions such as attention, short-term memory, perceptual speed, perception of time, and reaction time. No change in heart rate or oxygen uptake could be observed over the three phases of the menstrual cycle. However, pulmonary ventilation during work varied significantly. It was highest in the menstrual phase. At the same heart rate exercise on the bicycle ergometer was perceived as more exerting in the menstrual phase than in either the premenstrual or postmenstrual phase. Among the performance tests significant results were obtained only in a test of reaction time, which was slightly impaired during the menstrual phase. The results of the performance tests do not however support the assumption that menstruation affects a woman's mental work capacity.