Participants were recruited from the Hertfordshire Cohort Study (HCS), an established birth cohort study of men and women born between 1931 and 1939 in the county of Hertfordshire, UK [19, 20]. Between November 2019 and March 2020, 176 participants were visited at home by a trained researcher who administered a questionnaire that included information on medical history, medication use, lifestyle and social and psychological factors. The study received ethical approval from the East of England—Cambridgeshire and Hertfordshire Research Ethics Committee, reference number 11/EE/0196. All participants gave written informed consent before participating in this study.
At the visits, height was measured to the nearest 0.1 cm using a portable, free-standing stadiometer (Harpenden Pocket Stadiometer, London, England). Body weight was measured to the nearest 0.1 kg, with the participant wearing clothes but not shoes, using portable SECA digital scales (Model 835). BMI (kg/m2) was calculated (weight (kg)/ height (m)2).
Self-reported physical function was assessed using the SF-36 PF (Short Form-36 Physical Function) scale [21]. At the visits, various measurements were also performed, including the Short Physical Performance Battery (SPPB) test, which included the assessment of gait speed, standing balance and sit-to-stand performance (chair stand test) [22, 23]. Gait speed was measured using an eight-foot course, with participants being asked to walk at their usual pace while being timed using a stopwatch; participants could make use of assistive devices, such as canes, if necessary. Gait speed was calculated by dividing the distance walked by the time between the first and last step. For the chair stand test, participants were asked to move from a sitting position to a fully upright standing position five times as quickly as possible, with their arms crossed across their chest, while being timed from their initial sitting position until upright on the fifth repetition. The standing balance test involved a semi-tandem stand where participants were asked to place one foot in front of the other such that the big toe of one foot was touching the side of the heel of the other. If participants could not hold the semi-tandem stand for 10 s, they were asked to perform a side-by-side stand (standing with feet side by side). If they could hold the semi-tandem stand for 10 s, they were also asked to attempt a full tandem stand where they placed one foot in front of the other (touching heel to toe) and held this position for as long as they were able, up to 10 s. A physical performance score was derived from the above three tests, according to the SPPB scoring guidelines [23]. Participants who could not complete either the gait speed test or the chair rise test were given a score of 0. The remaining participants’ times were divided into quartiles and scored 1–4, the slowest to the fastest quartile. For the standing balance test, if participants could maintain balance in the tandem stand for at least 10 s, a score of 4 was given; if their time was ≥ 3 and < 10 s, they scored 3; if they maintained balance for < 3 s but were able to maintain a semi-tandem stand, they scored 2; if they were unable to perform the semi-tandem stand but could perform the side-by-side stand, they scored 1; and if they could do neither the semi-tandem nor the side-by-side stand, they scored 0. The scores for the three tests were then summed, with a maximum possible score of 12 and a minimum of 0. Scores of 9 or lower were considered to be indicative of poor physical performance.
Handgrip strength was measured using a handgrip Jamar dynamometer, three times for each hand and the maximum value was used for analysis [24]. Frailty was defined as the presence of three or more of the following Fried frailty criteria [25]: unintentional weight loss, weakness, self-reported exhaustion, slow gait speed and low physical activity. Weight loss was assessed by asking whether the participant had lost any weight unintentionally in the preceding 3–6 months; answering affirmatively was considered as unintentional weight loss. Weakness was defined as a handgrip strength of < 27 kg for men and < 16 kg for women [7]. Exhaustion was assessed by asking the participant how often in the preceding week they felt that “everything they did was an effort” or that “they could not get going”. Participants who reported feeling this way for either ‘a moderate amount of time’ or ‘most of the time’ were categorised as ‘exhausted’. Slow gait speed was defined as a gait speed of ≤ 0.8 m/s. Physical activity was assessed by the average amount of time (in minutes per day) spent walking outside, cycling, gardening, playing sports or doing housework in the preceding 2 weeks [26]. Low physical activity was defined as an activity time in the bottom fifth of the sex-specific distribution (≤ 58 min/day for men and ≤ 90 min/day for women).
Nutrition risk scores were calculated using the DETERMINE checklist [10]. This tool includes ten questions on age-related and contextual factors that are linked to poor nutrition in older age: illness leading to dietary changes; eating few meals/reduced appetite; eating few fruits, vegetables or milk products; high alcohol intake; eating difficulties due to tooth or mouth problems; not having enough money for food; eating alone; frequent medication usage; unintentional weight change; and physical difficulties with shopping, cooking or eating. Responses are weighted to calculate an overall nutrition risk score for each participant, by summing the ten scored items, with thresholds given to identify categories of risk: ‘low’ (0–2), ‘moderate’ (3–5) and ‘high’ (≥ 6) nutritional risk; total nutrition risk scores range from 0 to 21 [10]. Nutritional risk was also assessed using the MUST, which includes three scores: body mass index (BMI) (BMI ≤ 20 kg/m2 indicates risk), unintentional weight loss (unintentional weight loss during the preceding 3–6 months; ≥ 5% indicates at risk) and an acute disease effect score. From this information, total MUST scores are calculated and grouped into three risk categories: low risk (score 0), medium risk (score 1) or high risk (score ≥ 2) [18, 27].
Statistical analysis
Descriptive characteristics are given as mean with standard deviation (SD) for continuous normally distributed variables, median with interquartile range (IQR) for continuous variables with a skewed distribution, or counts and percentages for categorical variables, as appropriate.
The calculated nutrition risk score from the DETERMINE checklist was used as a continuous variable in regression analyses. The relationships between the nutrition risk score and gait speed, chair rises time, physical performance (SPPB) score, SF-36 physical functioning score, and grip strength were examined using multivariate linear regressions. Fisher–Yates rank-based inverse normal transformations were performed to create z-scores (FY z scores) to enable the comparison of effect sizes. The associations between the nutrition risk score and Fried frailty, tandem stand < 10 s, and low SPPB score (≤ 9) were examined using multivariate logistic regressions. Analyses were performed with adjustments for sex, age, age left education and number of comorbidities. Number of comorbidities was assessed by asking whether the participant had been told by a doctor that they had any of the following conditions: high blood pressure, diabetes, lung disease, rheumatoid arthritis, multiple sclerosis, cancer, vitiligo, depression, Parkinson’s disease, heart disease, peripheral arterial disease, osteoporosis, thyroid disease and stroke. Analyses were performed using Stata version 16.