Previous studies have reported that people with visual impairment (VI) do not consume enough dairy products, meats and wholegrains(Reference Roebothan1) and do not consider the nutritional value of food before purchase(Reference Kostyra, Zakowska-Biemans and Sniegocka2,Reference Stevens, Bartlett and Cooke3) . It has been reported that people with macular degeneration in the UK do not consume the recommended daily amounts of nutrients for their age group(Reference Stevens, Bartlett and Cooke3). It has also been reported that those with ocular conditions such as macular degeneration and glaucoma do not have nutritious diets and are unsure about what foods they should consume to maintain optimal eye health(Reference Stevens, Bartlett and Cooke3–Reference Wong, Man and Fenwick13). The cost of malnutrition in the UK is £19·6 billion annually(14), with £16 billion being related to being overweight or obese(14). It is reported that malnourished adults account for 30 % of hospital admissions and 35 % of care home admissions in the UK(14).
Studies that have investigated the impact of VI on nutritional status have concluded that interventions are required to improve the diets and dietary habits of people with VI(Reference Bilyk, Sontrop and Chapman15–Reference Gopinath, Liew and Burlutsky17). These studies have suggested that the interventions could take the form of skills training(Reference Bilyk, Sontrop and Chapman15), development training packages for the young(Reference Smyth, Spicer and Morgese16) or rehabilitation packages for the elderly(Reference Gopinath, Liew and Burlutsky17). It has been reported that nutritional interventions save the National Health Service 172·2–229·2 million pounds due to reduced health care use(14).
Systematic review of the literature demonstrates that VI significantly impacts on nutritional status(Reference Jones and Bartlett18). Previous studies have used a variety of methods to assess nutritional status, such as nutritional screening tools to assess whether a person is at risk of undernutrition(Reference Muurinen, Soini and Suominen19), measuring BMI(Reference Montero20–Reference Wrzesinska, Urzędowicz and Motylewski22) and qualitative and quantitative analysis on the ability to acquire, cook and eat food(Reference Kostyra, Zakowska-Biemans and Sniegocka2,Reference Bilyk, Sontrop and Chapman15–Reference Gopinath, Liew and Burlutsky17,Reference De Faria, Da Silva and Ferreira23–Reference Pardhan, Latham and Tabrett26) . Some of these studies did not use representative sample size(Reference Roebothan1,Reference Bilyk, Sontrop and Chapman15,Reference Smyth, Spicer and Morgese16,Reference Acil and Ayaz21,Reference Nakamura, Otomo and Maeda24–Reference Baker27) , and some used non-standardised methods to categorise participants as visually impaired(Reference Muurinen, Soini and Suominen19,Reference Montero20) . Two studies conducted a dietary consumption assessment: one carried out a gross categorisation assessment of foods eaten into meat products, wheats and grains(Reference Roebothan1); the other carried out a detailed analysis of dietary consumption but the dietary consumption assessment was conducted for schoolchildren and was not done in the UK(Reference Montero20).
The present study is the first to investigate the impact of VI on nutritional status in older adults and whether dietary consumption is affected by shopping and cooking abilities.
Materials and methods
Survey design
Following a systematic review of the literature(Reference Jones and Bartlett18), a thirty-seven-question, cross-sectional questionnaire was designed to evaluate the impact of VI on dietary consumption, vision-related quality of life and activities of daily living(Reference Jones, Bartlett and Cooke28). The questionnaire was piloted and validated prior to the start of the study. Full details of the validation process and questionnaire design are reported elsewhere(Reference Jones, Bartlett and Cooke28).
Sample size
Using previously reported nutritional analysis data(Reference Stevens, Bartlett and Cooke3), sample sizes were calculated for individual nutrients. The effect sizes chosen for each nutrient were based on published mean and standard deviation data(Reference Stevens, Bartlett and Cooke3). The minimum sample size (n) required for a two-tailed t test at an α error level of 0·05 and a power (1 – β) of 80 % was calculated (see Table 1).
Table 1. Sample size calculations for each nutrient*
(Mean values and standard deviations)
DD, difference to detect; ES, effect size.
* Mean values for effect size calculations taken from Stevens et al.(Reference Stevens, Bartlett and Cooke3).
In total, 146 participants were recruited for the present study. Ninety-six participants were recruited for the VI group and fifty participants for the control group.
For fats, saturated fats, cholesterol and vitamins C, D and E, the sample size required to detect the desired effect sizes was large. The present study was therefore underpowered for these nutrients at powers (1 – β) 0·6, 0·3, 0·6, 0·5, 0·2 and 0·4, respectively. It would have been time-consuming and impractical to collect data for these nutrients in order to detect the desired effect sizes.
Inclusion and exclusion criteria
For both the VI and the control participants, the exclusion criteria were dietary restrictions relating to conditions such as coeliac disease, inability to communicate in English or inability to hear well over the telephone.
Following the criteria for the certification of VI, proposed by the Royal National Institute for the Blind (RNIB), participants were categorised:
Registered severely sight impaired (SSI) or sight impaired (SI);
Eligible for SSI or SI registration but not actually registered;
Not eligible for SSI or SI registration, but experiencing a level of VI that precludes driving. Or in other words, a reduction in vision that significantly impairs day-to-day activities
For the control group, participants were aged 50 years or over and had to demonstrate binocular visual acuity of at least better than 6/9·5, that is, a visual acuity that would meet the level of sight required to be able to drive legally.
Participant recruitment and setting
In all, 109 participants with VI were recruited from across the UK from October 2017 to July 2018. Advertisements were placed with the Macular Society, the RNIB and Visionary, a membership organisation for VI charities. Participants were also recruited by being directly approached by the researcher at Focus and Aston, low vision clinics in Birmingham. They were also approached by the researcher at Sight Concern, a support group for those with VI in Worcestershire, New Outlook, a sheltered accommodation in Birmingham, designed specifically for people with VI and at local macular society support groups.
Participants responded to the advertisements in the Macular Society Sideview magazine. In all written information, the Macular Society uses at least a size 16 font. They also produce ‘accessible’ versions of their publications in PDF form, which can be read aloud by screen readers. There are other types of text processing and screen readers available as apps as well, which people may use a mixture of. Additionally, the Macular Society offers the option for people to receive audio versions of publications – they provide this as a compact disk for their Sideview magazine and their leaflets are available on their website as mp3 files. The study was also advertised through RNIB Connect (radio) whereby participants provided their contact details to the researcher via email and telephone. The researcher then called the participants and read out the participant information sheet and arranged a convenient time and date to deliver a structured telephone interview.
Of the 109 VI participants recruited, only thirteen were aged under 50 years, and so although their data were included in the qualitative analysis(Reference Jones, Bartlett and Cooke28), a decision was made to restrict the dietary analysis to a subgroup of VI participants aged 50 years and over.
In all, fifty control group participants without VI were recruited from December 2018 to January 2019. The records of patients at the Aston University Eye Clinic who had given consent for their records to be accessed and to be contacted for research and teaching purposes were reviewed. Those who met the inclusion criteria were contacted through telephone and invited to take part.
Procedure for 24-h food recall
Participants were asked to recall over the telephone all the food and drink they had eaten over the previous 24 h for 3 d in the same week.
Studies using telephone interviews for 24-h recalls have reported that they are comparable to the standard in-person method(Reference Yanek, Moy and Raqueño29,Reference Galasso, Panico and Celentano30) . Concerns about this method in the literature pertain to non-covering bias, that is, excluding those unable to use a telephone or those without a telephone(Reference Kyrø, Skeie and Dragsted31); however, studies have also reported that the dietary intake reported over the telephone is comparable for participants of different ages, sex and BMI(Reference Bogle, Stuff and Davis32).
The 24-h food recall is a methodological tool often used in dietary consumption studies, but presents advantages and limitations(Reference Castell, Serra-Majem and Ribas-Barba33). Advantages include short administration time, high precision when performed three or more times and low literacy requirements(Reference Galasso, Panico and Celentano30,Reference Castell, Serra-Majem and Ribas-Barba33–Reference Sharma, Rao and Jacob35) . Among the limitations falls the co-operation of the interviewee and their memory, in the case of the elderly, this can be compromised(Reference Caliendo36). In addition, difficulty of estimating the size of portions(Reference Souverein, de Boer and Geelen37) and recall bias can lead to over and under-reporting(Reference Castell, Serra-Majem and Ribas-Barba33,Reference Macdiarmid and Blundell38) .
Method
Materials
Materials included:
A password protected file of the participants’ names and contact details
A list of predefined questions for dietary analysis
A telephone equipped with a headset
Quiet surroundings
A digital voice recorder to collect verbal informed consent
A spreadsheet to record dietary information (separated into morning, afternoon, evening and snacks)
The interviewer received training on how to conduct the interview and input data into the dietary analysis software A La Calc by the project lead.
A telephone protocol was used in order to remain neutral and not react adversely to any responses given. The interviewer had a list of predefined questions. These questions were screened for clarity and wording by a focus group of six people with VI prior to the start of the study. The same interviewer conducted the interview for each participant.
Participants quantified the portions of foods consumed using the Zimbabwe Hand Method(Reference Kinshuck39–Reference Mash42), and this method has been shown to be more accurate than using household measures when measuring portion sizes(Reference Gibson, Hsu and Rangan43). The method was explained to participants at the start of the first telephone call, and they were reminded of how to quantify each food as they recalled each food item. This step was then repeated at each telephone call. This 24-h food recall exercise was carried out on two weekdays and one weekend day of the same week to ensure precision and validity of reporting(Reference Resnicow, Odom and Wang44).
The procedure was as follows:
To aid co-operation, verbal digitally recorded consent was taken at the start of each food diary; participants were reminded they could withdraw at any time if they wished.
The participants were first asked to recall foods eaten for breakfast, lunch and supper as well as any snacks consumed. They were asked about fluids they drank (alcohol, coffees, fruit juice, teas, milk).
To aid participants’ recall, they were probed to check if they had missed any information, that is, vitamin, supplements or other foods.
They were then asked to provide a detailed description of the food items. Examples of the questions asked include: what type of milk (full-fat, semi-skimmed and skimmed), whether milk, sugar and sweeteners were added to drinks, whether bread was white, seeded and wholemeal, whether cereal was fortified or unfortified and if vegetables were fresh or frozen.
Food quality was assessed where possible, participants were asked if spreads were cholesterol reducing and low in fat, as well as whether foods were baked or fried, shop bought or homemade.
To further support participants’ recall, they were asked one final time if they might have missed any other foods or drinks.
RDA analysis
The 3-d 24-h food recalls were analysed using nutritional software called A La Calc (Red Hot Rails LLP). This software provided a detailed nutritional analysis for each participant based on their self-reported food and drink consumption. This software has been used in previous research(Reference Stevens, Bartlett and Cooke3) and has been designed to be used by nutritionists, schools, consultants, manufacturers, and for research purposes. The software uses McCance and Widdowson’s composition of foods dataset to ensure an accurate breakdown of the nutrients contained within each food item entered(45). This UK nutrient database is maintained by the Food Standards Agency and contains the nutritional information of foods commonly consumed in the UK. All calculations are also compliant to the EC Directive 90/496/EEC(46). For each participant, the mean dietary consumption across the three reported days was calculated.
Data analysis
Statistical processing was performed using Microsoft Excel and exported to SPSS Software version 23.0 (IBM UK Ltd). The descriptive analysis is demonstrated in mean, standard deviation, median and interquartile range.
Normally distributed data that had two independent variables and a continuous variable were analysed using an independent t test (P < 0·05). The t test was used to analyse if dietary intake was influenced by sex in both the control and VI groups and living arrangements for the control group (living with family/living on own). A one-way between-groups ANOVA was used for normally distributed data that included one independent variable (grouping variable) that had three or more levels and one dependent continuous variable (P < 0·05). Post hoc analysis was performed using Tukey’s test. The one-way between-groups ANOVA test was used to analyse that dietary intake was influenced by shopping abilities (myself/myself with support/do not shop) and cooking abilities (do not cook/cook with support/cook myself), level of VI (do not drive/SI/SSI) and if level of VI was influenced by living arrangements (sheltered accommodation/family/living alone)
Where data were not normally distributed, the non-parametric equivalents, the Mann–Whitney U test (P < 0·05) and Kruskal–Wallis test with Bonferroni corrections for multiple comparisons were used (P < 0·02). The Mann–Whitney U test was used to determine if there was a significant difference between the ages of the two groups of this sample and the analysis of the dietary intake for males and females. The Kruskal–Wallis test was used to determine if living arrangements, shopping and cooking abilities and level of VI influenced dietary intake for nutrients that were not normally distributed.
Fisher’s exact test was used to determine if there was a relationship between level of VI and ability to shop and cook (P < 0·05).
Results
Demographics
Three-day 24-h recalls were analysed for sixty-four females and thirty-two males with VI. Ages of those with VI ranged from 51 to 96 years. The mean age was 76 (sd 11·7) years. The majority of the participants sampled were living with family members or on their own, were retired and were Caucasian.
VI in this sample was caused by multiple factors. For example, participants had congenital blindness due to measles or lost sight due to neurological conditions such as stroke. They also reported VI due to ocular trauma and retinal diseases such as diabetic retinopathy and macular degeneration. Genetic causes such as ocular albinism, macular dystrophies, and retinitis pigmentosa as well as corneal degenerations and optic nerve head disease, that is, glaucoma, were reported.
Those who were classified as SSI had been affected for longer compared with the other VI participants (H 17·2) (P < 0·01). In all, 81 % of the participants were registered SSI or SI with most being SSI (see Table 2).
Table 2. Demographic characteristics of participants with and without visual impairment
(Percentage (%) of people with and without visual impairment)
N/A, not applicable.
* These participants may have been eligible for sight impaired registration.
In all, twenty-six females and twenty-four males were recruited as part of the control group. The mean age was 75·4 (sd 7·2) years old. All the control participants were Caucasian and either lived with their family or on their own. In comparison with the VI group, a larger proportion of the control were in paid employment, either full time, part time or ad hoc (see Table 2).
The mean age of females with and without VI was 77·0 (sd 12) and 75·1 (sd 6·4) years, respectively, with no significant difference between groups (U 1033) (P = 0·07). The mean age for males with and without VI was 74·9 (sd 11·5) and 75·5 (sd 8·3) years, respectively, with no significant difference between groups (U 299) (P = 0·1).
Dietary consumption analysis
Dietary consumption compared with RDA
Table 3 displays the 3-d, mean and median results for macro- and micronutrients for the females and males in each group. These are compared with the RDA for each constituent for those aged over 74 years as reported by Public Health England(47).
Table 3. Nutrients consumed by females and males with and without visual impairment (VI) aged over 50 years compared with the recommended UK government guidelines (https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/618167/government_dietary_recommendations.pdf)
(Mean values and standard deviations; medians and interquartile ranges (IQR))
* Data not provided.
Similar amounts of macro- and micronutrients to RDA were found for the dietary consumption of participants with and without VI. Both groups were consuming fewer amounts of carbohydrates, dietary fibre, fats and vitamin D as recommended for their age group. Both groups were consuming sugars, Fe, protein, vitamin C and Ca in excess. The control group exceeded the recommended daily amounts of saturated fat intake.
Dietary consumption of participants with and without visual impairment
Females with VI consumed significantly fewer nutrients compared with their age-matched counterparts, including energy content, fats, saturated fats, protein, salt, Ca, cholesterol and vitamin C (see Table 3). Despite consuming fewer energy content, the amounts of vitamin D (U 704) (P = 0·29), fibre (t 1·4) (P = 0·10) and sugars (U 707) (P = 0·26) they consumed did not significantly differ from the control group.
Males with VI consumed significantly lower amounts of most nutrients compared with males from the control group (see Table 3). The amounts of vitamin C (U 307) (P = 0·20), vitamin D (U 304) (P = 0·18), vitamin E (t 1·2) (P = 0·20) and cholesterol (U 313) (P = 0·24) they consumed were not significantly different from that consumed by males without VI.
Dietary consumption and living arrangements
Living arrangements influenced the dietary consumption of participants with VI.
Those who lived with family members (6523 (sd 1699) kJ) or in sheltered accommodation (7360 (sd 1611) kJ) had a higher intake of energy content (F 2,93 = 5·7) (P < 0·01), compared with those living on their own (5552 (sd 1443) kJ). Those living independently were found to be eating an average of 1389 kJ less than those who lived in sheltered accommodation or with family. Post hoc Tukey’s test did not reveal any significant difference between those living with family and sheltered accommodation. Those living with family were found to be eating 16 g more fat (H 11·35) (P < 0·01) and 25 g more carbohydrates (H 11·52) (P < 0·01) compared with those living in their own home.
Among the control group, those living with family members showed no difference (t 1·8) (P = 0·08) than those living on their own.
Dietary consumption and level of visual impairment
Participants classified as SSI consumed an average of 25·7 mg less vitamin C than other VI participants (H 12) (P < 0·01).
Ability to cook was affected by level of VI with more SSI participants being unable to cook than other VI participants (Fisher’s exact test 25·9) (P < 0·01).
A one-way between-groups ANOVA revealed VI participants who cooked with support (7640 (sd 1657) kJ) consumed significantly (P < 0·05) more energy content (F (2,93) = 8·8) than those who did not cook (P < 0·01) (6293 (sd 1657) kJ) or cooked for themselves (5552 (sd 1397) kJ).
Post hoc comparisons using the Tukey’s honestly significantly different test revealed that those who cooked with support consumed an average of 1720 kJ more energy content than the other groups. Cooking with support also resulted in a higher dietary intake of carbohydrates (200 (sd 85) g) (F (2,93) = 4·8) (P = 0·01) when compared with not cooking (185 (sd 54) g) and when people with VI cooked by themselves (154 (sd 47) g). The dietary intake of fats (F (2,93) = 3·8) (P = 0·03) for those cooking with support was higher (64·8 (sd 14) g) than those who did not cook (54 (sd 23) g) or cooked independently (48 (sd 17) g).
Kruskal–Wallis with Bonferroni corrections revealed that those who received support consumed 6·7 mg more vitamin E (H 10·7) (P < 0·01) and 93·6 mg more vitamin C (H 23·89) (P < 0·01) than those who cooked by themselves or sourced ready meals.
Eating behaviours of participants with and without visual impairment
Meal preparation and shopping
All participants without VI stated that they had no difficulty cooking and could cook a hot meal if they were required to. The control group mainly reported no difficulty shopping, with 96 % stating they shopped independently. The 4 % who required support reported that physical limitations, such as arthritis, left them unable to lift heavy goods.
In contrast, 50 % of the participants with VI in this sample could not cook food by themselves. They required support, relied on a family member or purchased ready meals.
Only 29 % of participants with VI shopped independently, 42 % required support and 29 % did not shop but relied on family members or used meal delivery services. Level of VI affected ability to shop with more participants who were SSI or SI being unable to do so or requiring support (Fisher’s exact test 11·5) (P = 0·02). However, no relationship was found between reported shopping ability and dietary consumption.
When asked about food choices, participants with VI stated preference as the primary factor. Those without VI stated that perceived impact of foods on their health determined what they purchased (see Fig. 1).
Fig. 1. Main factors deciding the choice of foods purchased in a sample of participants with and without visual impairment (VI). 
, Participants with VI; 
, participants without VI.
Attitudes towards diet and knowledge of healthy eating
In all, 59 % of participants with VI and 94 % without VI stated they were satisfied with their current health. In all, 61 % of participants with VI stated they were happy with their diet, giving this as the reason for why they would not change it. About 39 % who stated they would change their diets provided a variety of reasons. The main reasons given were ‘eat more fresh fruits, vegetables’, ‘have a diet that was varied and be aware of foods available’ and ‘improve knowledge of healthy eating’. Similarly, 62 % of the control group stated they would not change their current diet. Of these, 50 % believed they had already adopted healthy eating behaviours and 12 % stated they would not change their diet because they were happy with it. About 38 % of participants without VI who reported they would like to change their diets stated they would mainly like to ‘eat healthier foods’ or ‘be more disciplined with sugary foods’. Other reasons given were they would like to eat ‘more expensive foods like caviar’ and would consider changing their diets if ‘healthier foods tasted nicer’.
Participants were asked ‘can you name the five food groups for a balanced diet’. More of the control group were able to name the food groups compared with those with VI (see Fig. 2). The participants without VI strongly agreed that the foods we eat affect our health. Of the participants with VI, 18 % stated that they believed that our health is not affected by the foods we eat.
Fig. 2. Participants’ ability to name the five food groups for a balanced diet. , Participants with visual impairment (VI); , participants without VI.
Discussion
The present study is the first to report that older adults with and without VI are not meeting the recommended daily requirements as recommended by Public Health England(47). This finding suggests that additional factors other than VI could play a role in the undernourishment of participants in the present study. Factors reported in previous studies that cause a compromised nutritional status in older adults include physical changes associated with aging, as well as cognitive, psychological and social factors such as dementia, depression, isolation and limited income(Reference Mangels48). Researchers have also found that older adults have smaller appetites and feel that portion sizes of foods in shops are inappropriately large(Reference Bloom, Lawrence and Barker49).
For the first time using detailed dietary analysis, the present study reports that people with VI are consuming significantly fewer nutrients than age-matched controls. The present study supports the view that there are multi-factorial obstacles that make it difficult for people with VI to maintain healthy feeding, including difficulties shopping for, preparing and cooking food(Reference Kostyra, Zakowska-Biemans and Sniegocka2,Reference Stevens, Bartlett and Cooke3,Reference Bilyk, Sontrop and Chapman15,Reference Baker27) . People with VI have reported having an aversion to cooking(Reference Bilyk, Sontrop and Chapman15) and report that meals could take up to 2 h to cook(Reference Kostyra, Zakowska-Biemans and Sniegocka2). It has also been reported that people with VI eat more intuitively and the loss of visual cues may drive a reduced appetite in people with VI(Reference Linne, Barkeling and Rossner50–Reference Steenhuis and Poelman53).
The present study found that participants with VI who were living alone and cooking for themselves consumed significantly less food sources of energy content, fats, vitamin C and vitamin E nutrients than those with VI who lived with family or received support to cook. The reduction in energy content consumed by the participants with VI who were living alone (1389 kJ) almost equates to missing an entire meal, such as breakfast (1674 kJ) as recommended by UK government guidelines(54). The participants in the age-matched control group who were living alone also consumed less energy (799 kJ) than those living with family although this was not significant. It has been previously documented that older adults living alone have less favourable diets than those who live with family or receive support(Reference Davis, Randall and Lee55,Reference Robinson, Ritchie and Roth56) . Bereavement has been reported as a substantial change that has been linked to poor dietary intake and quality(Reference Whitelock and Ensaff57). A recent Canadian study suggested eating alone might act as reminder of bereavement and result in reduced pleasure from eating(Reference Atkins, Ramsay and Whincup58). Another study reported British men who were married and living with family had a better diet quality than those living alone(Reference Whitelock and Ensaff57). Lack of motivation to cook has also been reported as a contributory factor in older women who had lost their partner, who report preferring to cook less(Reference Bloom, Lawrence and Barker49). Other studies have reported that food wastage when buying for one could play a role in participant food choices and food quality with specific food groups being affected more so than others(Reference Whitelock and Ensaff57). Vegetables in particular were reported as the food group that participants had the greatest difficulty with when buying for one(Reference Whitelock and Ensaff57).
Participants with VI in the present study were less able to recall the five food groups for a balanced diet. Those with VI were mainly making food choices irrespective of its nutritional value, whereas those without VI made food choices based on how healthy foods were. To improve dietary consumption knowledge of where to obtain healthy ready meals, support with cooking and supporting the knowledge of the recommended portion sizes of food may therefore be helpful for people with VI. The results of the present study suggest that interventions are required to improve the nutritional awareness of people with VI. These could take the form of skills training or rehabilitation(Reference Bilyk, Sontrop and Chapman15) to support activities of daily living.
Strengths
Participants from across the UK took part in the present study, and so the study was not restricted by geographical location. The method of using 24-h recalls has been reported to be affected by age, and a trend of under-reporting of foods consumed has been reported. In an attempt to reduce this bias, the 24-h food recalls were collected for three non-consecutive days as they have been reported to have precision and when multiple days are assessed validity(Reference Resnicow, Odom and Wang44). The 24-h food recall was also the first question asked at the initial telephone call to attempt to reduce this bias.
Limitations
The results of the present study are subject to limitations. The present study was performed over a 3-d period of the same week. This method would significantly influence the dietary intake analysis, as these data were not representative of what participants ate throughout the year. Future studies should perform the dietary analysis on multiple days throughout the year to capture the macro- and micronutrients consumed more completely.
The same interviewer collected the data for each participant, and the dietary analysis may therefore be subject to interviewer bias. Participants also required notice for the 24-h food recalls, and therefore, the recalls were not truly spontaneous; this time to prepare may have also influenced the results of the present study.
The thirty-seven-question item survey was disseminated prior to the second and third telephone calls. The questions asked may have influenced the participants eating habits for the subsequent phone calls although the researchers did not find a significant variation in the dietary consumption reported at the follow-up telephone calls.
Participants could not always report with accuracy about the quality of the food consumed, for example, if they went to a pub or restaurant, they could not report if the food was prepared with heart healthy oil or not, which may have affected the accuracy of reporting and therefore the dietary consumption analysis.
VI may have also affected the ability of participants to relay portion sizes accurately and therefore have affected the dietary analysis for this group.
The aim of the present study was to recruit participants from all ages and ethnicities; however, very few participants who were under the age of 50 years, identified as Black Asian ethnic minority, and were in employment participated.
Measurements such as BMI, waist circumference and activity levels would be useful in future studies to evaluate the nutritional status of people with VI more completely.
Conclusion
The present study is the first to highlight that older adults with VI in the UK are eating fewer nutrients when compared with their age-matched counterparts. Both adults with and without VI are not meeting the recommended amounts nutrients according to government guidelines. These results suggest local and government-led initiatives should be implemented to support the diets of older adults in the UK, and these initiatives could include healthy eating workshops, café clubs or skills training and rehabilitation.
Ethical statement
The present study was conducted according to the guidelines laid down in the Declaration of Helsinki, and all procedures involving human subjects/patients were approved by the Aston University School of Life and Health Sciences Ethics Committee, no. 1398. Verbal informed consent was obtained from all subjects/patients. Verbal consent was witnessed and formally digitally recorded by the first author.
Consent for publication
All participants gave verbal digitally voice recorded, informed consent for their data to be published.
Acknowledgements
The authors would like to thank the participants from Aston University Low Vision Clinic and Eye Clinic, Focus Birmingham, New Outlook Northfield Birmingham, Macular Society, the RNIB and Sight Concern, Worcestershire, for their contribution to the present study.
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
N. J. contributed to the acquisition of data, analysis and interpretation of data. H. E. B. made substantial contributions to conception and design. Both authors participated in drafting the article and revising it critically. Both authors gave final approval of the version to be submitted for review.
The authors declare that there are no conflicts of interest.
