Using joint models to disentangle intervention effect types and baseline confounding: an application within an intervention study in prodromal Alzheimer's disease with Fortasyn Connect.

BACKGROUND: Many prodromal Alzheimer's disease trials collect two types of data: the time until clinical diagnosis of dementia and longitudinal patient information. These data are often analysed separately, although they are strongly associated. By combining the longitudinal and survival data into a single statistical model, joint models can account for the dependencies between the two types of data. METHODS: We illustrate the major steps in a joint modelling approach, motivated by data from a prodromal Alzheimer's disease study: the LipiDiDiet trial. RESULTS: By using joint models we are able to disentangle baseline confounding from the intervention effect and moreover, to investigate the association between longitudinal patient information and the time until clinical dementia diagnosis. CONCLUSIONS: Joint models provide a valuable tool in the statistical analysis of clinical studies with longitudinal and survival data, such as in prodromal Alzheimer's disease trials, and have several added values compared to separate analyses.

Effects of Souvenaid on plasma micronutrient levels and fatty acid profiles in mild and mild-to-moderate Alzheimer's disease.

INTRODUCTION: Circulating levels of uridine, selenium, vitamins B12, E and C, folate, docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA) have been shown to be lower in patients with Alzheimer's disease (AD) than in healthy individuals. These low levels may affect disease pathways involved in synapse formation and neural functioning. Here, we investigated whether, and to what extent, circulating levels of micronutrients and fatty acids can be affected by oral supplementation with Souvenaid (containing a specific nutrient combination), using data derived from three randomized clinical trials (RCT) and an open-label extension (OLE) study with follow-up data from 12 to 48 weeks. METHODS: Subjects with mild (RCT1, RCT2) or mild-to-moderate AD (RCT3) received active or control product once daily for 12-24 weeks or active product during the 24-week OLE following RCT2 (n = 212-527). Measurements included plasma levels of B vitamins, choline, vitamin E, selenium, uridine and homocysteine and proportions of DHA, EPA and total n-3 long-chain polyunsaturated fatty acids in plasma and erythrocytes. Between-group comparisons were made using t tests or non-parametric alternatives. RESULTS: We found that 12-24-week active product intake increased plasma and/or erythrocyte micronutrients: uridine; choline; selenium; folate; vitamins B6, B12 and E; and fatty acid levels of DHA and EPA (all p < 0.001). In the OLE study, similar levels were reached in former control product/initial active product users, whereas 24-week continued active product intake showed no suggestion of a further increase in nutrient levels. CONCLUSIONS: These data show that circulating levels of nutrients known to be decreased in the AD population can be increased in patients with mild and mild-tomoderate AD by 24-48-week oral supplementation with Souvenaid. In addition, to our knowledge, this is the first report of the effects of sustained dietary intake of uridine monophosphate on plasma uridine levels in humans. Uptake of nutrients is observed within 6 weeks, and a plateau phase is reached for most nutrients during prolonged intake, thus increasing the availability of precursors and cofactors in the circulation that may be used for the formation and function of neuronal membranes and synapses in the brain.

Differences in nutritional status between very mild Alzheimer's disease patients and healthy controls.

BACKGROUND: Studies on the systemic availability of nutrients and nutritional status in Alzheimer's disease (AD) are widely available, but the majority included patients in a moderate stage of AD. OBJECTIVE: This study compares the nutritional status between mild AD outpatients and healthy controls. METHODS: A subgroup of Dutch drug-naïve patients with mild AD (Mini-Mental State Examination (MMSE) ≥20) from the Souvenir II randomized controlled study (NTR1975) and a group of Dutch healthy controls were included. Nutritional status was assessed by measuring levels of several nutrients, conducting the Mini Nutritional Assessment (MNA®) questionnaire and through anthropometric measures. RESULTS: In total, data of 93 healthy cognitively intact controls (MMSE 29.0 [23.0-30.0]) and 79 very mild AD patients (MMSE = 25.0 [20.0-30.0]) were included. Plasma selenium (p < 0.001) and uridine (p = 0.046) levels were significantly lower in AD patients, with a similar trend for plasma vitamin D (p = 0.094) levels. In addition, the fatty acid profile in erythrocyte membranes was different between groups for several fatty acids. Mean MNA screening score was significantly lower in AD patients (p = 0.008), but not indicative of malnutrition risk. No significant differences were observed for other micronutrient or anthropometric parameters. CONCLUSION: In non-malnourished patients with very mild AD, lower levels of some micronutrients, a different fatty acid profile in erythrocyte membranes and a slightly but significantly lower MNA screening score were observed. This suggests that subtle differences in nutrient status are present already in a very early stage of AD and in the absence of protein/energy malnutrition.

Impact of dietary fat quantity and quality on skeletal muscle fatty acid metabolism in subjects with the metabolic syndrome.

Insulin resistance is characterized by disturbances in lipid metabolism in skeletal muscle. Our aim was to investigate whether gene expression and fatty acid (FA) profile of skeletal muscle lipids are affected by diets differing in fat quantity and quality in subjects with the metabolic syndrome (MetS) and varying degrees of insulin sensitivity. 84 subjects (age 57.3±0.9 y, BMI 30.9±0.4 kg/m(2), 42 M/42 F) were randomly assigned to one of four iso-energetic diets: high-SFA (HSFA); high-MUFA (HMUFA) or two low-fat, high-complex carbohydrate diets, supplemented with 1.24 g/day of long-chain n-3 PUFA (LFHCCn-3) or control oil (LFHCC) for 12 weeks. In a subgroup of men (n=26), muscle TAG, DAG, FFA and phospholipid contents were determined including their fractional synthetic rate (FSR) and FA composition at fasting and 4h after consumption of a high-fat mixed-meal, both pre- and post-intervention. Genes involved in lipogenesis were downregulated after HMUFA (mean fold change -1.3) and after LFHCCn-3 (fold change -1.7) in insulin resistant subjects (< median of (S(I))), whereas in insulin sensitive subjects (>median of insulin sensitivity) the opposite effect was shown (fold change +1.6 for both diets). HMUFA diet tended to decrease FSR in TAG (P=.055) and DAG (P=.066), whereas the LFHCCn-3 diet reduced TAG content (P=.032). In conclusion, HMUFA and LFHCCn-3 diets reduced the expression of the lipogenic genes in skeletal muscle of insulin resistant subjects, whilst HMUFA reduced the fractional synthesis rate of DAG and TAG and LFHCC n-3 the TAG content. Our data indicate that these diets may reduce muscle fat accumulation by affecting the balance between FA synthesis, storage and oxidation.

Adipose triglyceride lipase and hormone-sensitive lipase protein expression in subcutaneous adipose tissue is decreased after an isoenergetic low-fat high-complex carbohydrate diet in the metabolic syndrome.

The objective was to determine the contribution of dietary fat quantity and composition to lipolysis and lipolytic gene expression in humans in relation to obesity, insulin resistance, and the metabolic syndrome (MetS). Men and women with the MetS were randomly assigned to one of four isoenergetic diets: a high-fat saturated fat diet (n=10), a high-fat monounsaturated fat diet (n=7), and two low-fat high-complex carbohydrate (LFHCC) diets, one supplemented with 1.24 g/day long-chain n-3 PUFA (LFHCC: n=7, LFHCCn-3: n=8). Subcutaneous adipose tissue biopsies were taken before and after the 12-week dietary intervention period. ATGL and HSL mRNA and protein expression was determined. Whole body rate of appearance of free fatty acids (Ra(FFA)) was determined by intravenous infusion of [(2)H(2)]-palmitate in a subgroup of men (n=20). Adipose tissue ATGL and HSL mRNA and protein expression was not affected by alterations in dietary fat composition. Pooled analysis comparing the low- and high-fat diets showed that ATGL and HSL protein expression was significantly reduced after the LFHCC diets (P=.04), irrespective of long-chain n-3 PUFA. Moreover, LFHCC diets lowered fasting insulin, HOMA(IR), and (LDL)-cholesterol concentrations (P≤.05). Changes in ATGL and HSL protein expression was positively associated with changes in whole body Ra(FFA) (P<.03). The low-fat high-complex carbohydrate diets reduced ATGL and HSL protein expression and significantly improved circulating lipids and insulin sensitivity. Under isoenergetic conditions, dietary fat quantity, rather than composition, may be most important for modulating subcutaneous adipose tissue ATGL and HSL protein expression.

Transcriptional metabolic inflexibility in skeletal muscle among individuals with increasing insulin resistance.

Disturbances in skeletal muscle lipid metabolism may play an important role in development of insulin resistance (IR). The aim was to investigate transcriptional control of skeletal muscle fatty acid (FA) metabolism in individuals with the metabolic syndrome (MetS) with varying degrees of insulin sensitivity (S(I)). 122 individuals with MetS (NCEP-ATP III criteria) at age 35-70 years, BMI 27-38 kg/m(2) were studied (subgroup EU-LIPGENE study). Individuals were divided into quartiles of S(I) measured during a frequently sampled insulin modified intravenous glucose tolerance test. Skeletal muscle normalized mRNA expression levels of genes important in skeletal muscle FA handling were analyzed with quantitative real-time PCR. The expression of sterol regulatory element binding protein 1c (SREBP1c), acetyl-CoA carboxylase 2 (ACC2), diacylglycerol acyltransferase (DGAT1), and nuclear respiration factor (NRF) was higher in the lowest two quartiles of S(I) (<50th) compared with the highest two quartiles of S(I) (>50th). Interestingly, peroxisome proliferator-activated receptor coactivator 1α (PGC1α), peroxisome proliferator-activated receptor α (PPARα), and muscle carnitine palmitoyl transferase 1b (mCPT1), important for oxidative metabolism, showed a complex mRNA expression profile; levels were lower in both the most "insulin sensitive" (IS) as well as the most "IR" individuals. Lipoprotein lipase (LPL) mRNA was reduced in the lowest quartile of S(I). Enhanced gene expression of SREBP1c and ACC2 in the IR state suggests a tendency towards FA storage rather than oxidation. From the lower expression of PGC1α, PPARα, and mCPT1 in both the most "IS" as well as the most "IR" individuals, it may be speculated that "IS" subjects do not need to upregulate these genes to have a normal FA oxidation, whereas the most "IR" individuals are inflexible in upregulating these genes.

Skeletal muscle fatty acid handling in insulin resistant men.

Disturbances in skeletal muscle lipid metabolism may precede or contribute to the development of whole body insulin resistance. In this study, we examined fasting and postprandial skeletal muscle fatty acid (FA) handling in insulin resistant (IR) men. Thirty men with the metabolic syndrome (MetS) (National Cholesterol Education Program-Adult Treatment Panel III) were included in this sub-study to the LIPGENE study, and divided in two groups (IR and control) based on the median of insulin sensitivity (S(I) = 2.06 (mU/l(-1))·min(-1)·10(-4)). Fasting and postprandial skeletal muscle FA handling were examined by combining the forearm balance technique with stable isotopes of palmitate. [(2)H(2)]-palmitate was infused intravenously to label endogenous triacylglycerol (TAG) and free FAs (FFAs) in the circulation and [U-(13)C]-palmitate was incorporated in a high-fat mixed meal (2.6 MJ, 61 E% fat) to label chylomicron-TAG. Muscle biopsies were taken to determine muscle TAG, diacylglycerol (DAG), FFA, and phospholipid (PL) content, their fractional synthetic rates (FSRs) and degree of saturation, as well as messenger RNA (mRNA) expression of genes involved in lipid metabolism. In the first 2 h after meal consumption, forearm muscle [(2)H(2)]-labeled TAG extraction was higher in IR vs. control (P = 0.05). Fasting percentage saturation of muscle DAG was higher in IR vs. control (P = 0.016). No differences were observed for intramuscular TAG, DAG, FFA, and PL content, FSR, and muscle mRNA expression. In conclusion, increased muscle (hepatically derived) TAG extraction during postprandial conditions and increased saturation of intramuscular DAG are associated with insulin resistance, suggesting that disturbances in skeletal muscle FA handling could play a role in the development of whole body insulin resistance and type 2 diabetes.

Fasting and postprandial remnant-like particle cholesterol concentrations in obese participants are associated with plasma triglycerides, insulin resistance, and body fat distribution.

Elevated plasma concentrations of remnant-like particle cholesterol (RLP-C) are atherogenic. However, factors that determine RLP-C are not fully understood. This study evaluates which factors affect RLP-C in the fasting and postprandial state, using multiple regression analyses in a large cohort of lean and obese participants. All participants (n = 740) underwent a test meal challenge containing 95 energy % (en%) fat (energy content 50% of predicted daily resting metabolic rate). Fasting and postprandial concentrations of circulating metabolites were measured over a 3-h period. Obese participants (n = 613) also participated in a 10-wk weight loss program (-2510 kJ/d), being randomized to either a low-fat or a high-fat diet (20-25 vs. 40-45en% fat). Postprandial RLP-C was associated with fasting RLP-C, waist:hip ratio (WHR), HOMA(IR) (homeostasis model assessment index for insulin resistance) (P < 0.001), and age, independently of BMI and gender [adjusted R(2) (adj. R(2)) = 0.70). These factors were also related to fasting RLP-C (P < 0.010), along with gender and physical activity (adj. R(2) = 0.23). The dietary intervention resulted in significantly lower fasting RLP-C concentrations, independently mediated by weight loss, improvements in HOMA(IR), and the fat content of the prescribed diet. However, after inclusion of plasma triglyceride (TG), HDL-cholesterol, and FFA concentrations in the models, HOMA(IR) and WHR no longer significantly predicted fasting RLP-C, although WHR remained a predictor of postprandial RLP-C (P = 0.002). Plasma TG was strongly associated with both fasting and postprandial RLP-C (P < 0.001). In conclusion, plasma RLP-C concentrations are mainly associated with plasma TG concentrations. Interestingly, the high-fat diet was more effective at decreasing fasting RLP-C concentrations in obese participants than the low-fat diet.