We observed that dietary fat quality in the long-term care facilities was poor in 2007 but even worse in 2017/18, in spite of the official nutrition recommendations. This worsening was due to significant increase of SFA intake and consequently dietary fat quality indicators worsened in older long-term care residents. Higher SFA intake of the 2017 cohort was predicted by sugar intake, male gender, eating independently, eating higher amounts of foods, and not having dry mouth.
Very few studies have described fat composition profile of long-term care residents. In a study by Rodríguez-Rejón et al.  fat composition of Spanish long-term care facilities was reported, but quality indicators were not calculated. In their study, SFA intake was about half of that in 2017/18 cohort in our study. In a Canadian study, SFA intake evaluated from long-term care facilities’ menus was about the same as in the 2007 cohort, but MUFA and PUFA intakes were considerably higher than in either of the cohorts in our study .
Recommendations on fat quality are uniform across various national authorities and expert groups [14,15,16,17,18]. Generally, all authorities encourage reducing SFA intake to less than 10 E% or even reducing it even more and replacing it with MUFAs and PUFAs . In our study, fat quality in the long-term care residents was already poor in 2007, but even worse in 2017: SFA intake of the residents was on average 17 E% compared to the 13 E% in 2007. Moreover, in both of the cohorts, intake of PUFAs was lower than the recommended 5–10 E%. Malnutrition and its risk in the residents was high (> 80%) in both measuring points, which is common in long-term care facilities [5,6,7]. In order to avoid residents’ weight loss it is a common practice to increase energy content of the served meals by adding fat to the meals during preparation . Traditionally this has meant adding butter or cream to various foods and using whole milk products instead of low fat or fat free products. In the newly published recommendation in Finland targeted especially for older people, fat quality is seen as an important issue, and the use of vegetable oils and soft margarines are encouraged instead of butter and cream . This clearly has not happened in the long-term care settings. Although the data in both cohorts was gathered prior to the publication of the new recommendation, the general recommendation for diet quality has been published in 2014 .
Good fat quality may slow some aspects of age-related decline in health. High SFA intakes elevate liver fat and serum cholesterol, whereas increase of MUFA and PUFA seems to be beneficial for modulation of liver fat and lipid metabolism [27,28,29,30]. Moreover, high MUFA and PUFA diets may improve insulin sensitivity,  reduce type 2 diabetes risk  and improve cardiovascular outcomes . Healthy brain is also very much dependent on good cardiovascular health . Fatty acids take part in multiple functions in the body and interact with other dietary components as well as microbiome and thus dietary fat composition may be either pro- or anti-inflammatory [33, 34]. Cardiovascular disease, Alzheimer’s disease and frailty have all been associated with increased chronic inflammation [35,36,37]. It has been suggested that n-6:n-3 ratio is also important for cardiovascular health , although the optimal n-6:n-3 ratio for human health remains under debate . In our study, the n-6:n-3 ratio was reasonable good, but the problem with fat quality had more to do with low PUFA intakes in general. Thus, good fat quality and sufficient intake of n-3 fatty acids are important also for the oldest-old individuals.
Although the residents of the two cohorts were of similar age and did not differ in MNA score or BMI, the residents in the 2017/18 cohort had worse physical and mental health than the residents in the 2007 cohort. This reflects stricter national guidelines for admission to long-term care facilities . In Finland due to public policy institutionalized care for older people have been reduced to the minimum and people are expected to live in their homes as long as possible . Thus, only those who have very severe dementia, mobility disability, or other severe health complications due to multiple chronic diseases, are offered a place in a nursing home or assisted-living facility type of long-term care. This can also been seen in the participants of the 2017/2018 cohort, who had very poor cognition measured with CDR (Table 1) compared to the participants in the 2007 cohort.
However, factors related to nutritional care such as intensive nutritional care, use of oral nutrient supplements, dysphagia, chewing problems, or other nutrition related issues that might be associated with poor nutrition, were not associated with higher SFA intake in this study. Of specific nutritional issues, only dry mouth was inversely associated with SFA intake. This reflects higher food consumption, as those with dry mouth consumed considerably less energy than those who did not. Similarly, residents eating independently and residents reported consuming larger amounts of foods had higher SFA intake than those needing help with eating or those who only reported eating very little. Thus, all the nutritional care related associations were associated with amount of eaten foods.
It is quite interesting, that despite of official recommendations, the fat quality has worsened. This might be because during the past 10 years, specific education on how to identify and treat malnutrition in long-term care has been provided for nurses working in these facilities. This seems to have led to an increment of SFAs to the served foods in order to avoid weight loss of the residents. As sugar intake was one of the strongest predictors of higher SFA intake, it is likely that sugar is also added to the diet in order to make the offered foods more palatable for the residents at risk of weight loss.
The strengths of our study include its large sample of long-term care residents in both 2007 and 2017/18. To best of our knowledge this is the first study that specifically explores detailed fat composition and quality and how it has changed in recent years in the diets of these people. Trained nurses or nutritionists performed all the measurements in both 2007 and 2017/18 cohorts and all the questionnaires and measurements were validated. Moreover, demographic information, and nutritional supplements were retrieved from medical records, which increase the reliability of our results. However, our study also has many limitations. Food diaries may be subject to error. However, since trained nurses filled in the diaries for the residents, subjective under or over reporting is unlikely. More problematically, when only the 1–2 day food diaries are assessed, they may differ from the person’s average food intake over a longer period of time. However, although the individual food intake may vary on a daily basis, our results are relevant at the group level . Moreover, practices that favor using SFA sources such as whole milk products, spreads with high SFA content, butter and cream are not likely to vary considerably from day to day in long-term care facilities. One limitation has to do with the food diary data analyzing tool used in 2007 (Nutrica). Data obtained in the later cohort 2017/18 was more detailed than in 2007, and included also the amounts of trans fatty acids, n-3 and n-6 fatty acids. Therefore, very detailed comparison of fat composition between the two cohorts was not possible. Change in food diary analyzing tools between the two cohorts should not affect the reliability of the dietary intakes, since both programs are based on the same Finnish food database (Fineli) and are validated tools. The residents in the 2017/18 cohort had poorer health, cognition and mobility, which make them more vulnerable to malnutrition compared to the residents of 2007. However, the residents also had many similarities, and did not differ significantly in respect to nutrition between the cohorts. A further limitation in our study is that the mean time of stay in long-term care is only 2 years; therefore it was not possible to follow the same residents over time. However, although changes in dietary fat intake may have been due to numerous confounding factors e.g. dentition, BMI, health status or other dietary factors, in our study higher SFA intakes were only related to the amount of eaten foods.
The 2017/18 sample was randomly selected within voluntary facilities. The latter sample included six long-term care facilities originally included in the 2007 sample. Although, not all the same facilities were compared, all the long-term care facilities are operated by the city of Helsinki and all facilities should follow the same nutritional guidelines. Moreover, they have same resources, same kind of care and residents are alike spending their last years of their lives there. The participation of the facilities in this study was voluntary and the investigators could not influence the participation. Although fewer facilities participated in the cohort on 2017/2018, in fact higher number of residents took part of the study compared to the 2007 study. The results were obtained from long-term care facilities in the Helsinki metropolitan area with residents mostly of Caucasian origin. The data may thus not be applicable to other ethnic groups. Finally, due to observational nature of our study, no causal relationships can be drawn from these results.