Lower albumin levels are associated with frailty measures, trace elements, and an inflammation marker in a cross-sectional study in Tanushimaru

We enrolled subjects from the typical farming town, Tanushimaru, located in Kyushu, the southwestern island of Japan conducted in 2018 as a Japanese cohort of the Seven Countries Study [14]. The subjects of this study are 1368 participants (554 males and 814 females: mean age of 68.8 years) of a health check-up examination. As previously reported, the demographic background of the subjects in this area is similar to that of the general Japanese population [15].

The subjects’ medical history, history of cardio-cerebrovascular diseases, use of alcohol, and smoking were ascertained by a questionnaire. Alcohol intake and smoking were classified as current habitual use or not. Height and weight were measured, and body mass index (BMI) was calculated as weight (kilograms) divided by the square of height (square meters) as an index of obesity. Waist circumference was measured at the level of the umbilicus in the standing position. Blood pressure (BP) was measured in the supine position twice at 3-min intervals using an upright standard sphygmomanometer. Vigorous physical activity and smoking were avoided for at least 30 min before BP measurement. The second BP with the fifth-phase diastolic pressure was used for analysis. Hypertensive subjects were defined as those with systolic BP ≥ 140 mmHg and/or those with diastolic BP ≥ 90 mmHg and/or those receiving antihypertensive medication. Subjects with fasting plasma glucose (FPG) ≥ 6.99 mmol/l (126 mg/dl) and/or subjects taking oral hypoglycemic agents or receiving insulin injection were diabetic. Subjects with dyslipidemia were defined as those with low-density lipoprotein cholesterol (LDL-c) ≥ 3.62 mmol/l (140 mg/dl) and/or triglycerides ≥ 1.69 mmol/l (150 mg/dl) and/or high-density lipoprotein cholesterol (HDL-c) < 1.03 mmol/l (40 mg/dl) and/or those taking lipid-lowering drugs.

Fasting blood samples were centrifuged within 1 h after collection. Serum levels of IGF-1 and GH were measured by ELISA and ECLIA methods to 1363 subjects who could receive blood testing. The blood was submitted to the commercially available laboratory (SRL Inc. Fukuoka, Japan), and the intra- and inter-assay coefficient of variations of IGF-1 and GH, respectively, at the laboratory that performed the assays was 2.56% and 0.75%, and 3.06% and 0.73% [16]. Liver enzymes (alanine aminotransferase [ALT], aspartate aminotransferase [AST], and γ-glutamyl transpeptidase [γ-GTP]) were also measured. Estimated glomerular filtration rate (eGFR) was calculated by the following estimation formula that has been recommended by the Japan Society of Nephrology: eGFR (ml/min/1.73²) = (194 × Scr^−1.094 × age^−0.287) × (0.739 for females) [17]. In addition to these blood testing, trace elements (calcium, magnesium, iron, zinc, and phosphorus) and Hs-CRP as an inflammation marker were measured. The participants underwent measurements of handgrip strength using a calibrated strain-gauged dynamometer and expressed in kilograms (kg). They were asked to squeeze maximally with the dominant hand. The score of the dominant hand was used for analysis. Walk speed was measured over a 5-m course, marked out on level ground. Participants were asked to walk the course at their usual walking pace from a standing start. The time taken to complete the course was timed using a digital stopwatch and recorded to the nearest tenth of a second. The fastest time was used to derive walk speed in meters per second. Their cognitive functions were evaluated by the mini-mental state examination (MMSE). MMSE is designed to quickly measure global cognitive functioning, temporal and spatial orientation, attention, immediate and short-term memory, language, praxis, and calculation. Scores ranged from 0 to 30, with higher scores indicating better cognitive performance.

Carotid intima-media thickness (c-IMT) of the common carotid artery was determined by using duplex ultrasonography (Sonosite “TITAN,” ALOKA) with a 10-MHz transducer in the supine position. Longitudinal B-mode images at the diastolic phase of the cardiac cycle were recorded by a single trained technician who was blinded to the subjects’ background. We measured the only far wall of c-IMT. The images were magnified and measured on the screen and printed with a high-resolution line recorder (LSR-100A, Toshiba). We measured c-IMT according to the originally described method published in circulation [18]. Briefly, the c-IMT defined by Pignoli et al. [18] was measured as the distance from the leading edge of the first echogenic line to the leading edge of the second echogenic line. The first line represented the lumen-intimal interface; the collagen-containing upper layer of the tunica adventitia formed the second line. At each longitudinal projection, the site of the greatest thickness, including plaque, was sought along the arterial walls nearest the skin and farthest from the skin from the common carotid artery to the internal carotid artery. Three determinations of c-IMT of one artery were conducted at the site of the greatest thickness and at 2 other points, 1 cm upstream and 1 cm downstream from this site. The averaged value among the 6 IMTs (3 from the left and 3 from the right) was used as the representative value for each individual.

This study was approved by the Tanushimaru branch of the Japan Medical Association and by the local mayor, as well as by the ethics committee of Kurume University School of Medicine. All the participants gave informed consent. The Research Ethics Committee of the Kurume University School of Medicine (Process numbers 09019/2018) approved the study in conformity with the principles embodied in the declaration of Helsinki.

Because of their skewed distribution, natural logarithmic transformations were performed for the values of triglycerides, Hs-CRP, IGF-1, and GH. These values after the analysis using log (natural)-transformed values were presented in the original scale (Tables 1 and 2). In Tables 1 and 2, these values are presented as the geometric mean and range. Sex (men = 0, women = 1), smoking habits (non-smoker and former smoker = 0, current smoker = 1), alcohol intake (non-drinker and former drinker = 0, current drinker = 1), and medications for hypertension, dyslipidemia, and diabetes (no = 0, yes = 1) were used as dummy variables. The mean serum albumin levels were classified into the following 4 groups: G1 (3.2–3.9 mg/dL), G2 (4.0–4.3 mg/dL), G3 (4.4–4.6 mg/dL), and G4 (4.7–5.3 mg/dL). The mean parameters stratified by the 4 albumin levels groups were compared using an analysis of variance (ANOVA). Age and sex-adjusted means of parameters stratified by the 4 serum albumin groups were compared using an analysis of covariance (ANCOVA). For the categorical parameters, the χ² test was used to test differences among groups. Univariate regression analysis was performed using a linear regression analysis. Using the significant factors detected by the univariate analysis, multiple stepwise regression analysis was performed. P values of < 0.05 were considered to indicate statistical significance. All of the statistical analyses were performed using the SAS software program (version 9.4, SAS Institute, Cary, NC, USA).

We found that handgrip strength, times of walk 5 m, GH, and IGF-1 were significantly associated with the 4 albumin groups. Reports from Smit et al. [6] suggested that serum albumin levels were lower in frail US adults. They defined frailty based on the four domains by Fried et al. [19] as follows: (1) slow walking, (2) muscle weakness, (3) exhaustion, and (4) low physical activity. Lower serum albumin may have been shown to be associated with longer times of walk followed by greater loss of muscle mass and may be an early indicator of impending muscle strength decline as suggested by Schalk et al. [20].

There are many causes of frailty. Of these, alterations in hormones represent major factors involved in the pathophysiology of frailty [21]. The clear-cut effects of GH and IGF-1 on serum albumin levels in healthy individuals have not been demonstrated. Especially, although the effects of GH on bone and muscle in GH-replete individuals have been studied [22], further investigation remained to be required. Our data in the multiple stepwise regression analyses showed that GH is inversely associated with albumin levels, which can be interpreted by the fact that the declining activity of GH with age in part responsible for better nutritional status [23]. In the final analysis, serum IGF-1 is positively associated with albumin levels. As recent epidemiological survey supported our study [24], it was indicated that lower IGF-1 levels were associated with lower handgrip strength and worse physical performance. This may be also because human aging is accompanied by a decrease in serum IGF-1 secretion.

Cognitive dysfunction is one of the important factors of frailty measures. Our study showed that MMSE score was significantly and independently associated with serum albumin. Ng et al. reported low albumin was an independent risk marker for cognitive decline in community-living 1664 Chinese older adults [25]. Although the data is limited to elderly hip fracture patients, lower serum albumin levels are independently associated with lower MMSE scores [26].

Of trace elements we have measured, calcium, zinc, magnesium, and iron were positively and significantly associated with albumin levels in the final stepwise regression analysis. It is well known that calcium and iron are closely associated with human nutritional status [27]. On the other hand, Peng et al. [8] suggested serum zinc was a significant independent risk factor for hypoalbuminemia and hospital admissions. Ongan et al. [28] also reported that iron, zinc, and magnesium intake of malnourished elderly persons was lower. Recently, Japanese investigators have suggested that zinc intake was associated with mental disorder, whereas calcium, magnesium, and iron intake were not associated with mental disorders [29]. The impact of dietary mineral intake on mental health is still controversial.

The present study revealed that lower albumin levels are strongly associated with an inflammatory marker such as Hs-CRP. Hypoalbuminemia, which can be associated with various diseases, is frequently observed in dialysis patients, as previously reported [30, 31]. One suggested that levels of inflammatory biomarkers such as CRP and interleukin-6 are increased in hypoalbuminemic compared with normoalbuminemic dialysis patients [30]. The other suggested that hypoalbuminemia and inflammation may play an important role in atherogenesis [31].

Our previous study suggested that lower albumin levels have a strong impact on cerebro-cardiovascular mortality [3]. The enrolled subjects were free from apparent cerebro-cardiovascular disease at baseline, and their serum albumin levels were within the normal range; however, lower albumin levels and Hs-CRP can affect the high mortality on future cerebro-cardiovascular events in this population, because the inflammation is one of the important risk factors for atherosclerosis [32]. The precise mechanism underlying the relationship between lower albumin and an inflammation marker should be clarified in future studies.

Clinical roles of albumin are frequently discussed in hospitalized patients [33‐35]. However, there are few reports on the association between lower albumin levels and related factors in a general population. In the clinical practice, lower albumin levels [36] are commonly discovered in association with nutritional deterioration and disease-related inflammatory response. Along with the evolution of the disease itself, this condition might be a result of the aging process, with levels of albumin decreasing with advancing age. However, the association between the lower albumin levels and age has not been fully elucidated; therefore, the association should also take into account diseases and other age-related conditions rather than age alone.

The limitations of this study were as follows. First, because our study design is cross-sectional, we cannot clarify the changes of time trend in albumin. Prospective studies are planned to investigate the role of the factors in the subjects with lower albumin levels in our future studies after long-term follow-up. Second, we performed only a single blood testing to evaluate an association between lower albumin and related parameters. The quadrant shift may occur in the same subjects. Third, the pathophysiological mechanism underlying the association between low albumin levels and related factors was not revealed from our observational study. Nevertheless, clear positive or inverse relationships between lower albumin levels and related factors were striking and further investigation should be required. Fourth, we have no data regarding gastrointestinal tract function in this study.