Sex-related differences in the association between frailty and dietary consumption in Japanese older people: a cross-sectional study

Japan is facing a super-aged society, and the number of older people with frailty is increasing. Frailty leads not only to disability, long-term care and higher mortality but also to increased health-related costs. Preventing frailty in older people is an urgent task in Japan and other developed countries. Currently, several frailty criteria exist, such as those proposed by Fried L et al. or Rockwood K et al. [1, 2]. These criteria have shown an association between frailty and high mortality rates. Fried L et al. showed greater frailty in those who were older age, female, and African American, with less education, lower income, poorer health, and higher rates of comorbid chronic disease. Rockwood K et al. also identified that people who were older age, female, cognitively impaired and incontinent were more frail. These two reports both showed an association between frailty and female sex. Furthermore, a systematic review and meta-analysis showed consistent results [3].

One of the most effective ways to prevent frailty is to consume appropriate nutrition. In addition to the intake of protein, the consumption of meat, fish, soybeans, milk, fruit and vegetables decreases the prevalence of frailty. The Mediterranean diet is also known to prevent frailty in older people [4‐6]. Houston DK et al. reported in their three-year cohort study that older men and women in the highest quintile in terms of protein intake lost approximately 40% less lean mass and appendicular lean mass than those in the lowest quintile in terms of protein intake [7]. The Framingham Third Generation Study also showed a significant association between high levels of protein intake and higher appendicular lean mass and quadriceps strength [8]. Moreover, the intake of protein and vitamin D demonstrated protective effects against frailty. There is a growing body of evidence that links frailty to nutrition and dietary intake in older adults. However, sex-related differences in the association between frailty and dietary consumption are under discussion. A multicenter cross-sectional study that recruited 2108 older women aged ≥65 years reported that high protein intake was associated with a low prevalence of frailty [9]. However, Shikany et al. showed no association between protein intake and frailty in 5925 older men aged ≥65 years [10].

Previous studies have shown sex-related differences in the association between dietary consumption and health-related outcomes such as cardiovascular disease and mortality [11‐15]. Fiber intake was reported to decrease mortality only in females and not in males [14]. The consumption of milk and dairy products was reported to be associated with high mortality rates only in females and not in males [12]. In addition, higher levels of consumption of red meat and processed meat were associated with higher mortality rates in females than in males [15].

However, few studies have focused on sex-related differences in dietary consumption in frail older adults. The aim of the present study was to investigate sex-related differences in the association between frailty and dietary consumption.

Of the 1226 participants who responded to the survey, 905 provided written informed consent and were included in the analysis. The participant characteristics are shown in Table 2. Table 1 shows the distribution of Kihon Checklist scores. The average score on the Kihon Checklist was 3.78 ± 3.9 (Fig. 1a, b, c). Significant sex-related differences was detected for 7 items (Question 3, 6, 7, 10, 16, 17 and 24). Women had higher scores than men on five out of those 7 items (Questions 6, 7, 10, 16 and 17). The numbers of normal (nonfrail), prefrail and frail individuals were 546 (60.3%), 214 (23.6%) and 145 (16.0%), respectively. Among men, 272, 81, and 67, were categorized into the normal (nonfrail), the prefrail, and the frail groups, respectively, while 274, 133, and 78 women were categorized to these groups, respectively (Fig. 1d). The prevalence of prefrailty and the prevalence of frailty were significantly higher in women than in men.

The chi-squared test revealed that men who lived alone ate significantly fewer vegetables (p = 0.001), potatoes (p < 0.001), and snacks (p = 0.005) than those who lived with a partner or with parent(s) or/and child (ren). Women who live alone ate significantly less, meat (p = 0.033) and fruit (p = 0.025). Compared with women, men consumed significantly more fish, salty food, and alcohol and less fruit (Fig. 2).

As shown in Table 3, after adjusting for age, body mass index and family arrangement in the multinomial regression analyses, older men who consumed meat less than twice per week had a higher prevalence of frailty than those who consumed meat daily (odds ratio: 2.76 (95%CI: 1.12–6.77), p = 0.027). We detected a greater number of significant associations between the consumption of dietary elements and frailty in women than in men. Women who consumed fish, meat, vegetables, seaweeds and potatoes less than twice per week had a higher prevalence of prefrailty than those who consumed those dietary elements daily. The odds ratios were as follows: fish 2.12 (1.11–4.05), p = 0.023; meat 2.22 (1.18–4.18), p = 0.013; vegetables 2.14 (1.05–4.37), p = 0.036; seaweeds 2.17 (1.19–3.95), p = 0.011; and potatoes 1.96 (1.03–3.74), p = 0.040. Women who consumed fish, meat, vegetables, potatoes and snacks less than twice per week had a higher prevalence of frailty (odds ratios: fish 2.45 (1.02–5.89), p = 0.045; meat 4.05 (1.67–9.86), p = 0.002; vegetables 5.03 (2.13–11.92), p < 0.001; potatoes 3.84 (1.63–9.05), p = 0.002; and snacks 2.16 (1.02–4.56), p = 0.043). Women who consumed vegetables 3–6 times per week also had a higher prevalence of frailty compared with those who consumed vegetables daily (odds ratio: 3.88 (1.69–8.87), p = 0.001).

Two-way ANOVA revealed significant interaction between sex and vegetable consumption in participants who lived with a partner. Post-hoc test indicated that frailty was significantly less prevalent in women who consumed vegetables daily compared with those who consumed vegetables less than twice /week (p < 0.001). Meanwhile, although there tended to be an interaction between sex in fish and dairy products in participants who lived alone and between sex and soy bean products in participants who lived with a partner, such interaction was not statistically significant (Additional file 2: Table S1).

The dietary variety score was lower in frail men and women than in normal participants, but the differences were not significant (Men: normal; 4.3 ± 2.7, prefrailty; 3.7 ± 2.6, frailty; 3.6 ± 2.5, p = 0.057, Women: normal; 4.4 ± 2.7, prefrailty; 4.0 ± 2.6, frailty; 3.9 ± 2.4, p = 0.134). The multinomial regression analyses showed that frail and prefrail women had significantly lower dietary variety scores than nonfrail women, but the same trend was not observed in men (odds ratio, women: prefrailty; 0.879 (95%CI; 0.793–0.975), p = 0.015, frailty; 0.835 (95%CI; 0.727–0.959), p = 0.011. Men: prefrailty; 0.911 (95%CI; 0.806–1.029), p = 0.134, frailty; 0.872 (95%CI; 0.759–1.002), p = 0.053) (Table 4).

In summary, the present study demonstrated that the prevalence of prefrailty and frailty was higher in women than in men. Sex-related differences in the association between frailty and dietary consumption were clarified. A higher prevalence of frailty was detected in men who consumed less meat than in men who consumed more meat, and a higher prevalence of frailty was detected in women who consumed less fish, meat, vegetables, potatoes and snacks than in women who consumed those dietary elements with greater frequency. A low dietary variety score was significantly associated with frailty only in women.

Previous reports revealed an association between frailty and dietary consumption without considering sex-related differences. In their systematic review of longitudinal studies, Feng Z et al. reported that lower levels of consumption of fruits and vegetables were significantly associated with frailty [4]. Moreover, low protein intake has been reported to be connected not only to frailty but also to low lean mass, appendicular lean mass and muscle strength [7, 8, 20]. A Mediterranean diet is also known to prevent frailty. However, only a few studies have considered sex-related differences, despite the fact that sex-related differences in the association between the prevalence of frailty and dietary consumption have been consistently demonstrated. A study that analyzed only men showed no association between protein intake and frailty [10]. However, another study demonstrated a significant association between protein intake and frailty [9].

There are several reasons why sex-related differences in the association between dietary consumption and frailty are detected. First, Sharkey JR and colleagues discussed that compared with men, women require more calcium and vitamin D, which particularly affect bone development, bone maintenance and muscular function [21]. Additionally, falls, osteoporosis and fractures are more prevalent in older women than in older men. For these reasons, the association between frailty and the frequency of consumption of fish (which is rich in calcium and vitamin D) was detected in women but not in men. Our data showed that more women than men had a fear of falling while walking. Second, chronic inflammation may affect sex-related differences in the association between frailty and dietary consumption. Canon ME and colleagues reported that decreases in sex hormones in postmenopausal women lead to chronic inflammation and elevated levels of C-reactive protein and interleukin-6 [22]. Chronic inflammation promotes proteolysis and muscle catabolism; therefore, chronic inflammation may help explain the association, especially in women, between the inadequate intake of protein such as meat and fish and frailty. Third, we should take into consideration sex-related differences in the relationship between comorbidities and dietary consumption because comorbidities constitute one of the most strongly predictive factors for frailty [1]. Previous reports demonstrated the relationships between dietary consumption and comorbidities such as cardiovascular disease, cerebrovascular disease and cancer as well as the relationship between dietary consumption and mortality. Du et al. followed 451,665 participants, and they showed that the daily consumption of fresh fruits reduced systolic blood pressure, blood glucose, ischemic stroke, hemorrhagic stroke, major cardiovascular disease and death from cardiovascular causes [19]. In other studies, a higher incidence of cardiovascular disease was detected in participants with lower intake levels of vegetables [23, 24], fiber [14], protein [25], and white meat [15] and those with higher intake levels of milk [12, 26], red meat, and processed meat [15, 27]. Although some studies could not demonstrate or did not mention sex-related differences in the association between dietary consumption and comorbidities [22‐24, 27, 28], many studies indicated significantly a higher incidence of comorbidities in women than in men [12, 14, 15, 25, 26]. We hypothesize that greater frequency of consumption of meat, fish and vegetables may prevent older women from developing comorbidities, resulting in protection against frailty.

The total score of Kihon Checklist was significantly higher in women than in men particularly in physical questionnaires such as climbing stairs, stand up from a chair and fear of falling. Also, the odds ratios in the association between dietary consumption and frailty were higher in women than men. These results are probably because low meat and fish consumption have an effect on higher prevalence of physical frailty in women compared with men. Our results are consistent with previous studies that indicate higher protein intake is associated to lower risk of frailty only in women [29].

In this study, there were three food categories that indicated inconsistent results, namely, dairy products, fat or oil in men, and seaweeds in women. These three food categories were significantly associated with prefrailty, but not frailty, and this might be attributed to two reasons. First, these results arose from the distribution of the participants that consumed each food category. For example, only 2 prefrailty participants consumed dairy products 3–6 times/week (Additional file 1: Figure S1 g), and thus statistical error might have occurred. Second, there were more participants with prefrailty than those with frailty, and statistical significance tended to be more detected in the prefrailty than in the frailty group.

Some limitations of this study should be mentioned. First, the present study is a cross-sectional study and cannot provide direct evidence of causality. It is possible to interpret that frail older men and women cannot consume more fish, meat and vegetables. Second, the mortality and the incidence of comorbidities are different between those who consume red meat and those who consume white meat and between those who consume animal protein and those who consume vegetable protein. The type of meat or protein being consumed by older men and women should be considered. Third, because the Kihon Checklist is based on a self-report questionnaire, the results might have been affected by the participant’s memory, and the checklist should thus be adjusted according to the participant’s cognitive function. Finally, we assessed food frequency qualitatively rather than quantitatively. Energy intake should be evaluated to support our results.

As a conclusion, the present study demonstrated sex-related differences in the association between dietary consumption and frailty. Women were identified as having a greater association between food categories and frailty than men. The consumption of fish, meat, vegetables and potatoes is recommended to prevent frailty in women.