Upload
dangque
View
217
Download
5
Embed Size (px)
Citation preview
THE OCCUPATIONAL PERFORMANCE MEASURE OF FOOD ACTIVITIES:
ITEM POOL DEVELOPMENT AND MEASUREMENT PROPERTIES
Abstract
OBJECTIVE: Occupational therapists have the knowledge and skills to improve nutritional
risk in community-living older adults, by improving performance of food-related
occupations. However, few tools are available to evaluate these interventions. We developed
the first item pool that measures community-living older adults’ occupational performance of
food activities.
METHOD: In phase 1 of the research we developed and item pool within a qualitative
exploratory study with five older adults. In phase 2 we designed the Occupational
performance measure of food activities. In phase 3 we formally assessed the measurement
properties of the 15-item measure, using survey responses from 77 community-living older
adults. Construct validity, reliability, and utility were evaluated.
RESULTS: A 13-item measure of the occupational performance of food activities showed
acceptable validity and reliability in three subscales (Cronbach’s alpha ranged from .70
to .75). The item pool demonstrated good utility. We had comparable results for
administration via self-completion survey (n = 38) and interview (n = 39).
CONCLUSION: This unique item pool showed promising validity and reliability for the
measurement of the occupational performance of food activities. Occupational therapists are
cautioned against modifying existing measurement tools without thorough testing of the
resulting new measure of occupational performance.
KEY WORDS: Geriatric Assessment, Cooking, Eating, Reliability and Validity
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 2
Introduction
Enjoyment of food and adequate nutrition are essential contributors to a good quality
of life as we age (Position paper of the American Dietetic Association: Nutrition across the
spectrum of aging, 2005). Older adults participate in various food activities like grocery
shopping, preparing simple and complex meals, and managing their diet, to both enjoy their
food, and eat well. Yet one in ten community-living older adults aged over 65 years in the
United Kingdom (UK is at risk of malnutrition (European Nutrition for Health Alliance
2006), while at least four in ten older adults admitted to hospital in the UK are malnourished
(Age Concern 2007, Age UK 2010, European Nutrition for Health Alliance 2005). When
nutritional risk is not improved, resulting malnutrition may lead to further dependence and
disability, increased length of hospital stay and costs of treatment, and an increased risk of
death (Bartali et al 2006, Feldblum et al 2009). The occupational performance of food
activities is, therefore, a key domain of concern for occupational therapists. Occupational
performance includes both the objective ability to perform an activity and the subjective level
of satisfaction with this performance (Creek 2003, Law et al 1997, Law and Baum 2001). For
this study, we defined the occupational performance of food activities as: an individual’s
ability to do the food activities that are meaningful to them, in a way that satisfies their life
needs, within the context of their environment, life stage, and life roles.
Literature review
Factors affecting the occupational performance of food activities
Food activities hold multiple meanings across the lifespan. Over the life course,
individuals develop certain food identities, such as ‘food lover’, ‘fussy eater’, or ‘vegetarian’.
These personal identities are enacted by consistently selecting or modifying food choices in
order to express them (Bisogni et al 2002). Food activities are also an important part of
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 3
family life, particularly in routines and rituals that contribute to the family’s identity and the
meaning of the food activities that are being shared (Fiese et al 2002). The meaning and
performance of food activities are also influenced by society, including nationality (De Groot
et al 2004) and ethnicity (Gilbert and Khokhar 2008). However, there are many changes in
later life that can affect many aspects of occupational performance of food activities. These
include changes in ability, meaning, and performance contexts.
Analysis of time-use diaries from the UK for the year 2000 shows time spent
preparing meals is likely to increase following retirement (Cheng et al 2007). Retired men, in
particular, may be shopping more and spending more time on other food activities, such as
meal preparation within a caring role (Atta-Konadu et al 2011, Locher et al 2010, Ribeiro et
al 2007). Declining health also changes the occupational performance of food activities, since
older adults with chronic conditions may become increasingly dependent with regard to food
activities as their health declines (Barichella et al 2008, Jönsson et al 2008, Medin et al
2010). Increasing social isolation is another change in later life that can affect the
occupational performance of food activities. Shahar et al (2001) found widowhood led to a
significant increase in the number of meals eaten alone and a decline in cooking frequency. In
addition, a qualitative study with Swedish women found meals eaten with others were
pleasurable, while women living alone viewed food as a necessity (Gustafsson and Sidenvall
2002). In another qualitative study, widowed men similarly spoke of eating to live, in contrast
to some of the married men who expressed pleasure in eating (Moss et al 2007).
Occupational therapists’ role in reducing nutritional risk
Occupational therapists have the specialized knowledge and skills to improve the
occupational performance of food activities, and so improve nutritional risk associated with
chronic disease and a loss of independence. We are able to address both the performance
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 4
components related to feeding, eating and swallowing (Clark et al., 2007); and the activity
limitations that restrict participation in other food activities, like shopping and cooking
(Roley et al., 2008). The International Classification of Functioning, Disability and Health
(ICF, WHO 2001) provides a complete range of activity and participation domains for
individual and societal functioning. We used the nine activity and participation domains of
the ICF that included food as a framework to review literature on the food activities of
community-living older adults. The assessment of occupational performance was reviewed
using the key word ‘assessment’ in combination with ‘occupational performance’ or ‘food’ or
keywords used in the food activities literature review. Assessments of databases used
included AMED, Medline, and PsychInfo, with ‘age over 65’ as a search limit.
Our literature review found that evaluation of occupational therapy practice in relation
to food and nutrition, particularly with adults over 60, is limited to a small number of studies.
In a single-case experimental study with four older adults that assessed the effectiveness of
using a microwave to improve meal preparation in older adults, Kondo et al (1997) found that
the introduction of a microwave decreased time spent on meal preparation, but increased the
number of meals prepared. Frequency of using cooking appliances, number of food items
prepared, and time spent preparing meals, recorded by participants in a diary, were the
outcome measures for the study. In another study, comparing learning of kitchen skills both
in a clinical setting and at home for 44 adults with schizophrenia aged 27–62 years,
Duncombe (2004) found improvement in cooking skills for both groups following
intervention. The location of the training did not affect the level of improvement observed.
Assessment was based on modification of two kitchen assessments to create a new measure
of kitchen skills.
In a prospective randomized controlled trial with 46 adults aged over 60 in an in-
patient stroke unit, Liu et al (2004) compared a mental imagery intervention with standard
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 5
functional skills training, using observed performance of 20 household tasks that were rated
on a 7-point Likert scale. The study found patients in the mental imagery programme
improved more than the functional skills group and were more likely to transfer learned skills
to other tasks. In a fourth, uncontrolled prospective study investigating the effectiveness of
occupational therapy in improving eating in 36 in-patients with end-stage cancer, Lee et al
(2005) found a significant improvement in eating independence over 1 week, which was
maintained over 3 weeks. Assessment was completed using the Barthel Index. In contrast,
Logan et al (2003) found no change in independence in daily activities, including food
activities, in their study comparing Activities of Daily Living (ADLs) interventions with
leisure interventions in 309 adult stroke patients. Although the study reported using the
Barthel Index, the Extended Activities of Daily Living Scale (EADL), and the Nottingham
Leisure Questionnaire (NLQ), outcomes were measured as ‘independent’ or ‘not
independent’ in their analysis.
Issues in the assessment of the occupational performance of food activities
Although some of the studies supported the role of occupational therapy in addressing
performance of food activities, we also have concerns about how performance of food
activities was measured. The measures used included participant diaries, observed
performance rated on a Likert scale, and modifications of existing research instruments. Our
primary concern is related to the validity and the reliability of the tools used. Some of the
studies developed their own measures, without any rigorous testing for validity and reliability
(Kondo et al., 1997; Liu et al., 2004; Lee, Chan & Wong, 2005). The validity of selecting
items from existing scales, without considering if each item is a valid and reliable measure of
the task of interest, is also questionable. For example, Logan and colleagues (2003) used
individual items from the Barthel Index, Extended Activities of Daily Living Scale (EADL),
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 6
and Nottingham Leisure Questionnaire (NLQ), to measure specific performance components
of independence. Duncombe (2004) partially addressed reliability and validity of a modified
tool in her study, but neither of the validation studies have been published.
A second issue is the lack of assessment tools that adequately measure occupational
performance of food activities. Following a further literature review, we did not identify any
existing measures that address both subjective and objective dimensions of the occupational
performance of food activities. Although the nine activity and participation domains in the
ICF (WHO 2001) include food-related activities, when evaluating available tools to measure
these domains we found some were too specific, while others were not specific enough. For
example, the McGill Ingestive Skills Assessment (Lambert et al 2006) specifically assesses
performance components related to feeding, chewing, and swallowing. These components are
too specific for interventions addressing a range of food activities. Tools that were not
specific enough included food activities as a component of other ADL measures. For
example, ‘Feeding’ within the Barthel Index (Collin et al 1988) and preparing main meals,
washing up, and shopping as components of the ‘Domestic Domain’ in the Frenchay
Activities Index (Schuling et al 1993). Another problem is the focus on ‘independence’.
Measures of independence lack a subjective component. Some assessments, like the Barthel
Index and the ‘In the Kitchen’ subscale of the Nottingham Extended ADL Scale (Nouri and
Lincoln 1987), may also be insensitive to differences between participants or changes in
function. An alternative is to measure satisfaction with food-related life (Grunert et al 2007).
However, this does not assess which occupational performance problems may be decreasing
satisfaction with food-related activities.
Requirements for the development of a new food activities assessment tool
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 7
This paper describes how we addressed this lack of assessment tools, by developing and
pilot testing an item pool to measure occupational performance of food activities for
community-living older adults. Construct validity is the extent to which a measurement tool
measures the intended construct (O’Leary-Kelly and Vokurka 1998). The construct validity
of a new measurement tool needs to be developed and assessed through a multi-faceted
process (Clark and Watson 1995, O’Leary-Kelly and Vokurka 1998). The first step is to
ensure test items include all aspects of the construct of interest, otherwise referred to as
content validity (O’Leary-Kelly and Vokurka 1998). Secondly, the measurement properties
of the tool need to be assessed and items need to be refined. A wide variety of methods have
been used to assess and develop validity (for a review see Hattie, 1985) and the item pool
(see Clark & Watson, 1995) including criterion-based methods; indices based on internal
consistency like principal components analysis, factor analysis, and Cronbach’s α; and
evaluation based on item response theory, like Rasch analysis. The third step is to determine
whether the construct being tested relates to other constructs in the predicted way. This is
related to the external validity of the tool.
Aims of the study
Our primary goal was to develop a valid measure of occupational performance of
food activities that included both objective ability to perform food activities and subjective
satisfaction with that performance. In this paper we describe the iterative process of item pool
development, recommended by Clark and Watson (1995), which we followed. In Phase 1 we
selected an item pool of 15 items, based on an exploratory qualitative study with five older
adults and the existing literature. In Phase 2 we designed the Occupational Performance
Measure of Food Activities (OPMF). In Phase 3 we examined the extent to which our item
pool actually measured this intended construct, based on measures of internal consistency.
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 8
Ethical approval for the studies was granted by the Brunel University Research Ethics
Committee.
Method
Phase 1: Initial Item Pool Development with face and content validity
In Phase 1 we considered the range of activities older people associated with food,
and the measurable components of occupational performance. The aim of Phase 1 was to
develop an item pool for the OPMF that had face and content validity. The research question
guiding Phase 1 was: What daily activities do community-living older adults associate with
food?
Method for phase 1
We recruited a convenience sample of five community-living older adults via email.
An invitation to participate was sent, by the interest group’s chairperson, to 60 older adults
who were on the email database of one community interest group for retirees. Two men and
three women (covering an age range of 55 to 85 years) responded to the invitation, and
agreed to participate. These participants varied in their age, gender, and marital status, but
were largely homogenous in their ethnicity and socio-economic status (see Table 1). These
participants volunteered because they were interested in food activities and because they
believed they had expertise in food and nutrition in later life, based on personal and
professional experience. Five participants is an appropriate sample size for exploratory focus
groups with older adults (Barrett and Kirk 2000, Toner 2009). The sample of 5 older adult
participants in this study also fell within the parameters of the 4 to 40 participants most often
included in qualitative studies (Holloway and Wheeler 2010). Pseudonyms were used to
maintain participant confidentiality throughout this component of the study.
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 9
Data collection included a focus group and follow-up semi-structured interviews with
each participant. Within the focus group, participants were asked ‘What kinds of daily
activities do you do that involve food?’ and ‘Why do you do these activities?’. In the follow-
up interviews we used visual methods. Participants either took photographs of their food
activities in the week preceding the interview, or selected up to six objects in their home that
were related to their food activities. Interviews started with the question ‘The purpose of this
interview is to understand your every day food activities. Please can you tell me about your
photographs / the objects you have chosen’.
Data were transcribed verbatim for analysis. We used NVIVO 8.0 software to conduct
a by-word frequency count of the 1000 most common words across the data set. All verbs
related to food were included in the analysis (e.g. eat, grill, nibble), because ‘doing’ is central
to the construct of occupational performance. We used framework analysis to group together
food-related verbs that were associated with the same food activities. Initial codes were based
on International Classification of Functioning, Disability and Health (ICF) activity and
participation codes related to food (World Health Organization, 2001). New codes were
created for activities that did not match these codes.
Findings for phase 1
The word count frequency yielded 51 food-related verbs. The participants talked
about seven food activities, matched with nine ICF codes (See Table 2).
Phase 2: Development of the OPMF
The aim of Phase 2 was to design a valid measure of the occupational performance of
food activities that could be administered to community-living older adults.
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 1 0
Item selection
The 15 items of the OPMF were based on food activities identified in Phase 1 and
occupational performance domains from the Canadian Occupational Performance Measure
(COPM, Law et al 2005). Our tool included three subscales: Food Activity Importance, Food
Activity Performance, and Food Activity Satisfaction. Each subscale included five food
activities.
The food activities included in the OPMF were selected by the first author, based on
the five food activities participants most clearly associated with food. These included
shopping, cooking, eating, eating out, and eating healthily. Whether or not drinking was a
‘food activity’ was debated by participants in the focus group. However, John noted: ‘No I’m
thinking of food as being solid’, while Edna described drinking as something that ‘can
encourage you to eat.’ This item was therefore excluded from the tool. Growing food was
only evident in Martin’s individual interview, when he talked about apples from his garden.
When asked ‘Do you grow any food on purpose or do you simply have an apple tree?’ Martin
replied, ‘Well, we put the fruit trees in the garden and we use the fruit, but I don’t
deliberately grow vegetables or anything like that.’ Growing food was also therefore
excluded from the item pool.
The measurable aspects of occupational performance included in the tool were
‘importance’, ‘performance’, and ‘satisfaction’. These domains were consistent with the
definition of occupational performance of food activities used in this study. In addition, these
domains are included in the COPM, which is the most widely used measure of occupational
performance in occupational therapy research and practice (Carswell et al 2004). Including
these domains improved the face validity of the measure. However, we would emphasize that
we do not view our measure as a modification of this tool.
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 1 1
Rating scale
Responses range from 1 to 10 for each item (see Table 3). For example, the
importance of eating was rated from 1 (not important at all) to 10 (extremely important). A
10-point scale is generally preferred by researchers for its higher reliability and validity
compared to scales with less than four response categories (Preston and Colman 2000). A 9-
point range was also expected to be more sensitive compared to other measures.
Administration and scoring
Questions were grouped by food activity, rather than by subscale, for ease of
administration (see Table 3). The three subscales of this tool were ‘Importance’,
‘Performance’, and ‘Satisfaction’ with food activities. There were five food activity items in
each subscale, with a maximum score of 10 for each item. Each subscale therefore had a total
of 50 points. The three subscale scores were added for a total score of 150 points.
Participants were asked to ‘Please answer the questions below thinking about all of
your activities that involve food. Food activities may include shopping, cooking, eating at
home, eating out, and looking after your health.’ For each of the five activities, questions
were phrased as follows (with the relevant subscale indicated): Importance: ‘How important
is (food activity — for example, grocery shopping) to you? A score of 1 means grocery
shopping is not important to you. A score of 10 means grocery shopping is extremely
important to you.’ Performance: ‘How well are you able to do your (food activity)? A score
of 1 means you are not able to do your (food activity) at all. A score of 10 means you are able
to do your (food activity) extremely well.’ Satisfaction: ‘How would you rate your
satisfaction with the way you do your (food activity)? A score of 1 means you are not
satisfied at all with the way you do your (food activity). A score of 10 means you are
extremely satisfied with the way you do your (food activity).’
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 1 2
Phase 3: Construct validity, reliability, internal consistency, and utility.
Once the OPMF was designed, we used survey methods to administer the 15-item
measure. The aim of Phase 2 was to test the construct validity, reliability, and utility of the
item pool. The research question guiding Phase 3 was ‘What are the measurement properties
of the Occupational Performance Measure of Food Activities total scale?’
Sample
We recruited 77 older adults by convenience sampling from the same community
interest group as Phase 1, through advertisement in a local shopping mall, and via
snowballing. Inclusion criteria were aged over 60 years, retired from full-time work, and
living in the community. Exclusion criteria were living in residential care and inability to
consent to the interview. Data were collected as part of a larger multi-stage mixed-methods
study. Initially, 38 older adults were recruited to complete a postal questionnaire.
Subsequently, a further 39 older adults were recruited for survey interviews (see Table 4).
However, using two methods of data collection to meet the aims of the larger study also
provided an additional opportunity to examine the utility of the measure for both postal
survey and interview administration. Using Pearson’s chi-squared analysis, the authors
compared group differences in the categorical demographic characteristics, using Fisher’s
exact test where the sample size was too small: that is, below 5 (Field 2009). As can be seen
in Table 4, there were no significant group differences in demographic characteristics
between the two groups, apart from living arrangements. The total sample size was adequate
because post-hoc analysis, using PASS 11 software, showed that this sample (n = 77)
achieved 100% power to detect the difference between the coefficient alpha under the null
hypothesis (α = 0) and alternative hypothesis (α = .70) using a two-sided F-test (p < .05).
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 1 3
Data Collection and analysis
Participants (N = 77) were asked to rate the 15 items of the OPMF (Table 3). Data
were analysed using SPSS 15.0 statistical analysis software for each of the following
measurement properties:
a) Construct validity: First we assessed whether each item was measuring
occupational performance as a concept (Kelley 1942) by calculating the unidimensionality of
the three subscales. We used the correlation coefficients between each item and the rest of the
scale (Hattie 1985). Items which did not reflect the pattern of responses on the rest of the
scale, indicated by a correlation coefficient between the item and the rest of the scale < 0.3,
were deleted (De Vaus 2002).
b) Reliability: Cronbach’s alpha is a measure of the lower bound of the reliability of a
test (Cronbach 1951). Cronbach’s alpha is also commonly viewed as a measure of the
internal consistency of a scale as it gives an indication of the ‘average’ degree of consistency
in responses. The criteria for satisfactory reliability was α > 0.7 (Bland and Altman 1997). In
response to concerns about the use of Cronbach’s alpha in occupational therapy research
(Spiliotopoulou 2009), we also used Cronbach’s (1951) correction formula to calculate the
mean inter-item correlation (ρ), which is independent of the number of items in the scale. A
mean inter-item correlation between .15 and .20 indicated satisfactory internal consistency
(Clark and Watson 1995).
c) Utility: Finally, we considered whether the item pool and rating scales could be
easily administered to a group of community-dwelling older adults. We examined results for
missing data and compared mean scores and the distribution of mean scores for the total
sample, and between the survey and interview groups.
Findings
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 1 4
Data characteristics: The Kolmogorov-Smirnov test of normality showed that none
of the 15 items were normally distributed. The Food Activity Importance (M = 39, SD =
5.89), Food Activity Performance (M = 43, SD = 5.52) and Food Activity Satisfaction (M =
40, SD = 5.47) subscales were significantly negatively skewed, ZSkewness > 3.29, p < .001. The
non-normal distribution of the data increases the risk of missing a significant effect, or Type
II error (Field 2009).
Unidimensionality: As illustrated in Table 5, Food Activity Importance sub-scale was
not unidimensional: There was a low correlation between both Eating Out Importance, r =
- .01, and Healthy Eating Importance, r = .27, and the rest of the scale. This means
participants’ scores on these items did not reflect the pattern of scores for items in the rest of
the scale and should be deleted (De Vaus 2002).
Reliability: Statistical analysis using Cronbach’s α showed acceptable internal
consistency of Food Activity Performance and Food Activity Satisfaction (Table 5). A 5-item
measure of Food Activity Importance had questionable internal consistency, α = .58.
Statistical analysis using Cronbach’s correction formula suggested all three sub-scales had
acceptable internal consistency (Table 4). Cronbach’s α For Food Activity Importance
increased to an acceptable level when Eating Out Importance, and Health Eating Importance
were deleted, α = .73.
Utility: In the survey group, one participant had missing values on the OPMF (n = 1,
2.63%). For each item, the minimum range was 5 points, the maximum was 9 points. Scores
were not normally distributed in each of the subscales (See Table 5). Data were transformed
using square root transformation for negatively skewed data (Field, 2009). The transformed
data met the assumptions for normality and homogeneity between groups, therefore
parametric methods were used.
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 1 5
As can be seen in Table 5, an independent samples t-test showed no significant mean
difference for the measure of OPMF, when comparing survey and interview groups, t(74) =
-.001, p = 1.00, 95% CI [-0.61, 0.61]. There was also no significant difference between
groups for each of the subscales. The Kolmogorov-Smirnov Z test showed no significant
difference in the distribution of scores between survey and interview groups. We found the
tool to be easy to administer in the survey interviews, with the time taken to administer it
varying, dependent on how much participants talked about their reasons for rating their food
activities in the way that they did. We found a useful question to be ‘Is there anything that
stops you doing your (food activity) in the way that you would like to?’. This question was
not included in the survey but was asked of 37 of the 39 participants in the interviews.
Discussion and implications
The OPMF showed promising results for validity and reliability. In this study, we
have only addressed the first two steps in developing content validity: We defined the
construct of occupational performance of food activities; and developed an item pool and
used 10 point rating scales to measure observable aspects of this construct. This was based on
consultation with our target group, from an early stage in the item pool development (Vogt,
King, & King, 2004). Our findings suggest a 3-item subscale for Food Activity Importance,
and 5-item subscale for Performance and Satisfaction, can be validly and reliably measured.
There is scope to test this item pool further, particularly in light of our positive findings for
the item pool’s utility.
The absence of missing data suggests older adults participating in our study had no
difficulty completing the rating scales we used. We also found our scale sensitive enough to
evaluate occupational performance in a well population. We found a broad range of scores
for each of our items, even though, as expected, scores were significantly skewed towards the
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 1 6
upper end of the distribution - especially in comparison COPM ratings in clinical groups of
older people (Cup, Scholte op Reimer, Thijssen, & van, 2003; Wressle, Lindstrand, Neher,
Marcusson, & Henriksson, 2003). Finally, the lack of difference in mean scores and the
distribution of scores between the survey and interview groups suggest this measure can be
equally useful with both methods of data collection. This is supported by Kjeken et al’s
(2005) research showing that individual interview (ICC 0.92 and 0.93) and survey responses
(ICC = 0.90 and 0.90) using similar rating scales provided excellent test-retest reliability for
their performance and satisfaction subscales respectively, while telephone interviews provide
good test-retest reliability (ICC = 0.73 and 0.73).
A final consideration is that we have developed a tool that measures multiple aspects
of the ‘occupational performance of food activities’, as defined for this study. We did not
identify any previous studies that have combined both objective and subjective dimensions of
occupational performance of food activities, even though it is clear in the international
occupational therapy literature that occupational
performance is multi-dimensional (for example, Creek 2003, Law et al 1997).
Limitations of the study and Implications for Research
An initial limitation of the study is that this 13-item measure is still at an early stage
in its development and testing. As such, we suggest caution using the item pool in a clinical
setting. Nevertheless, its use in research can provide useful information for reliability and
validity in different contexts if this data is included in the results of future studies. A second
limitation is that this tool is also based on the views of a specific ethnic and socio-economic
group of older adults living in one geographical area. Participants were predominantly of
white British origin, living in an urban environment, and had a relatively higher
socioeconomic status than may be typical in other research contexts. Nevertheless, cross-
referencing the tool with the domains of the ICF means the authors have taken some steps to
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 1 7
consider broader socio-cultural contexts. Further opportunities for research, therefore, include
testing with a bigger sample, both locally and internationally, before the content validity of
the item pool can be assured. The final limitation of this tool is the need to examine other
aspects of reliability; most notably test–retest reliability. This could lead to further
opportunities to evaluate the effectiveness of occupational therapy in improving nutritional
risk, and its consequences, in community-living older adults. The domains of the tool’s item
pool, and the rating scales used, are both similar to the widely used COPM. A final
implication for research is that this study adds to the limited evidence that 10-point rating
scales, such as those used in the COPM, are sensitive enough to use in occupational therapy
assessment with community-living older adults (Pearlman and Wallingford 2003), and in
studies comparing clinical and non-clinical groups (McNulty and Beplat 2008).
Implications for occupational therapy practice
Our study has also highlighted some implications for occupational therapy practice; in
particular for clinicians who have had difficulty finding appropriate assessments. As
highlighted by our literature review, finding assessments for a single domain of occupational
performance can be challenging. This was surprising to us in this case, given that food
activities are such an essential part of daily life. However, we would not recommend that
clinicians modify any existing measures of occupational performance, such as prescribing
activities to be rated using the COPM rating scales, unless they are able to ensure the validity
and reliability of what will essentially be a new measurement tool. One suggestion for
practice is focusing the interview for the COPM on a specific domain of occupational
performance, such as food activities (A successful example of this is Eckel et al’s [2012]
study of meal preparation among eight American community-living older women). As a
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 1 8
result, service users would still be able to select the food activities most important to them as
individuals.
Our study has also highlighted the need to include the importance of an activity in
measures of occupational performance, as already gathered in the COPM interview. The
inclusion of importance in our new measure can provide valuable information for
occupational therapists offering nutrition-related interventions. We believe that older adults
who rate their food activities as more important will be at lower nutritional risk, more
motivated to engage in occupational therapy interventions that address their food activities,
and may be more receptive to messages related to healthy eating. On the other hand, a low
level of food activity importance, alongside low scores for satisfaction, may be an indicator
of increasing nutritional risk.
Conclusion
We tentatively conclude that occupational performance of food activities can be validly and
reliably measured using a 13-item measure across the subscales of Food Activity Importance,
Performance, and Satisfaction in research studies. However, our tool requires further testing
of reliability and validity before it is ready for use in clinical practice. Nevertheless, our study
highlights the complex process required to develop and test a new measurement tool. We
therefore urge caution in the modification of any existing assessment tools of occupational
performance unless a comprehensive programme of research is planned.
Acknowledgements
This project was funded as a component of the British Geriatrics Society / Dunhill
Medical Trust Research Fellowship. We also extend our thanks to Professor Mary Law for
her constructive feedback in preparation of this paper.
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 1 9
REFERENCES
Age Concern (2007) Hungry to be heard: the scandal of malnourished older people in
hospital. Available at:
www.scie.org.uk/publications/guides/guide15/files/hungrytobeheard.pdf Accessed
03.04.12.
Age UK (2010) Still hungry to be heard: The scandal of people in later life becoming
malnourished in hospital. Available at: http://www.ageuk.org.uk/ Documents/EN-
GB/ID9489%20HTBH%20Report%2028ppA4.pdf?dtrk=true Accessed 03.04.12.
American Dietetic Association (2005) Position paper of the American Dietetic Association:
nutrition across the spectrum of aging. Journal of the American Dietetic Association,
105(4), 616–633.
Atta-Konadu E, Keller HH, Daly K (2011) The food-related role shift experiences of spousal
male care partners and their wives with dementia. Journal of Aging Studies, 25(3),
305–315.
Barichella, M., Villa, M. C., Massarotto, A., Cordara, S. E., Marczewska, A., Vairo, A., et al.
(2008). Mini nutritional assessment in patients with Parkinson’s disease: Correlation
between worsening of the malnutrition and increasing number of disease-years.
Nutritional Neuroscience, 11(3), 128-134.
Barrett J, Kirk S (2000) Running focus groups with elderly and disabled elderly participants.
Applied Ergonomics, 31(6), 621–629.
Bartali, B., Frongillo, E. A., Bandinelli, S., Lauretani, F., Semba, R. D., Fried, L. P., et al.
(2006). Low nutrient intake is an essential component of frailty in older persons.
Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 61(6), 589-
593.
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 2 0
Bisogni CA, Connors M, Devine CM, Sobal J (2002) Who we are and how we eat: a
qualitative study of identities in food choice. Journal of Nutrition Education and
Behavior, 34(3), 128–139.
Bland, J. M., & Altman, D. G. (1997). Cronbach's alpha. British Medical Journal, 314(7080),
572.
Carswell, A., McColl, M. a., Baptiste, S., Law, M., Polatajko, H., & Pollock, N. (2004). The
Canadian occupational performance measure: A research and clinical literature review.
Canadian Journal of Occupational Therapy, 71(4), 210-222.
Cheng S, Olsen W, Southerton D, Warde A (2007) The changing practice of eating: Evidence
from UK time diaries, 1975 and 2000. British Journal of Sociology, 58(1), 39–61.
Clark, G. F., Avery-Smith, W., Wold, L. S., Anthony, P., Holm, S. E., Eating and Feeding
Task Force, et al. (2007). Specialized knowledge and skills in feeding, eating, and
swallowing for occupational therapy practice. American Journal of Occupational
Therapy, 61(6), 686-700.
Clark, L. A., & Watson, D. (1995). Constructing validity: Basic issues in objective scale
development. Psychological Assessment, 7, 309-319.
Collin, C., Wade, D. T., Davies, S., & Horne, V. (1988). The Barthel ADL index: A
reliability study. International Disability Studies, 10, 61-63.
Creek, J. (2003). Occupational therapy defined as a complex intervention. London: College
of Occupational Therapists.
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 2 1
Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika,
16(3), 297-334.
Cup, E. H., Scholte op Reimer, W. J., Thijssen, M. C., & van, K. (2003). Reliability and
validity of the Canadian occupational performance measure in stroke patients. Clinical
Rehabilitation, 17(4), 402-409.
De Groot LC, Verheijden MW, De Henauw S, Schroll M, Van Staveren WA, SENECA
Investigators (2004) Lifestyle, nutritional status, health, and mortality in elderly people
across Europe: A review of the longitudinal results of the SENECA study. Journals of
Gerontology Series A-Biological Sciences & Medical Sciences, 59(12), 1277–1284.
De Vaus, D. (2002). Surveys in social research (5th ed.). London: Routledge.
Duncombe, L. W. (2004). Comparing learning of cooking in home and clinic for people with
schizophrenia. American Journal of Occupational Therapy, 58(3), 272-278.
Eckel E, Schreiber J, Provident I (2012) Community dwelling elderly women and meal
preparation. Physical and Occupational Therapy in Geriatrics, 30(4), 344–360.
European Nutrition for Health Alliance (2006) Malnutrition among older people in the
community: Policy recommendations for change. Available at:
http://www.bapen.org.uk/professionals/publications-and-resources/bapen-reports/
malnutrition-among-older-people-in-the-community Accessed 13.01.14.
European Nutrition for Health Alliance (2005) Malnutrition with an ageing population: a
call for action. Available at:
http://www.rcn.org.uk/__data/assets/pdf_file/0018/12537/malnutrition.pdf Accessed
13.01.14.
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 2 2
Feldblum, I., German, L., Bilenko, N., Shahar, A., Enten, R., Greenberg, D., et al. (2009).
Nutritional risk and health care use before and after an acute hospitalization among the
elderly. Nutrition, 25(4), 415-420.
Field, A. (2009). Discovering statistics using SPSS (3rd ed.). London: SAGE.
Fiese BH, Tomcho TJ, Douglas M, Josephs K, Poltrock S, Baker T (2002) A review of 50
years of research on naturally occurring family routines and rituals: cause for
celebration? Journal of Family Psychology, 16(4), 381–390.
Gilbert PA, Khokhar S (2008) Changing dietary habits of ethnic groups in Europe and
implications for health. Nutrition Reviews, 66(4), 203–215.
Grunert, K. G., Dean, M., Raats, M. M., Nielsen, N. A., Lumbers, M., & Food in Later Life,
T. (2007). A measure of satisfaction with food-related life. Appetite, 49(2), 486-493.
Gustafsson K, Sidenvall B (2002) Food-related health perceptions and food habits among
older women. Journal of Advanced Nursing, 39(2), 164–173.
Hattie, J. (1985). Methodology review: Assessing unidimensionality of tests and items.
Applied Psychological Measurement, 9, 139-164.
Holloway I, Wheeler S (2010) Qualitative research in nursing and healthcare. Chichester:
Wiley-Blackwell.
Jonsson, A. C., Lindgren, I., Norrving, B., & Lindgren, A. (2008). Weight loss after stroke: A
population-based study from the Lund stroke register. Stroke, 39(3), 918-923.
Keller, H. H. (2005). Reliance on others for food-related activities of daily living. Journal of
Nutrition for the Elderly, 25(1), 43-59.
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 2 3
Kelley, T. L. (1942). The reliability coefficient. Psychometrika, 7(2), 75-83.
Kjeken, I., Slatkowsky-Christensen, B., Kvien, T. K., & Uhlig, T. (2004). Norwegian version
of the Canadian Occupational Performance Measure in patients with hand osteoarthritis:
Validity, responsiveness, and feasibility. Arthritis & Rheumatism: Arthritis Care &
Research, 51(5), 709-715.
Kondo, T., Mann, W. C., Tomita, M., & Ottenbacher, K. J. (1997). The use of microwave
ovens by elderly persons with disabilities. American Journal of Occupational Therapy,
51(9), 739-747.
Lambert, H. C., Gisel, E. G., Groher, M. E., Abrahamowicz, M., & Wood-Dauphinee, S.
(2006). Psychometric testing of the McGill ingestive skills assessment. American
Journal of Occupational Therapy, 60, 409-419.
Law, M., Baptiste, S., Carswell, A., McColl, M., Polatajko, H., & Pollock, N. (2005).
Canadian occupational performance measure (4th ed.). Ottawa: CAOT Publications
ACE.
Law, M., & Baum, C. (2001). Measurement in occupational therapy. In M. Law, C. Baum &
W. Dunn (Eds.), Measuring occupational performance: Supporting best practice in
occupational therapy (pp. 3-20). Thorofare: SLACK Incorporated.
Law, M., Cooper, B. A., Strong, S., Stewart, D., Rigby, P., & Letts, L. (1997). Theoretical
contexts for the practice of occupational therapy. In C. H. Christiansen, & C. M. Baum
(Eds.), Occupational therapy: Enabling function and wellbeing (). Thorofare,N.J.:
SLACK Incorporated.
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 2 4
Lee, W. T., Chan, H. F., & Wong, E. (2005). Improvement of feeding independence in end-
stage cancer patients under palliative care--a prospective, uncontrolled study. Supportive
Care in Cancer, 13(12), 1051-1056.
Liu, K. P., Chan, C. C., Lee, T. M., & Hui-Chan, C. W. (2004). Mental imagery for
promoting relearning for people after stroke: A randomized controlled trial. Archives of
Physical Medicine & Rehabilitation, 85(9), 1403-1408.
Locher JL, Robinson CO, Bailey FA, Carroll WR, Heimburger DC, Saif MW, Tajeu G,
Ritchie CS (2010) Disruptions in the organization of meal preparation and consumption
among older cancer patients and their family caregivers. Psycho-oncology, 19(9), 967–
974.
Logan, P. A., Gladman, J. R., Drummond, A. E., & Radford, K. A. (2003). A study of
interventions and related outcomes in a randomized controlled trial of occupational
therapy and leisure therapy for community stroke patients. Clinical Rehabilitation, 17(3),
249-255.
McNulty, M. C., & Beplat, A. L. (2008). The validity of using the Canadian occupational
performance measure with older adults with and without depressive symptoms. Physical
& Occupational Therapy in Geriatrics, 27(1), 1-15.
Medin J, Larson J, von Arbin M, Wredling R, Tham K (2010) Elderly persons’ experience
and management of eating situations 6 months after stroke. Disability and Rehabilitation,
32(16), 1346–1353.
Moss SZ, Moss MS, Kilbride JE, Rubinstein RL (2007) Frail men’s perspectives on food and
eating. Journal of Aging Studies, 21(4), 314–324.
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 2 5
Nouri, F. M., & Lincoln, N. B. (1987). An extended activities of daily living scale for stroke
patients. Clinical Rehabilitation, 1, 301-305.
O'Leary-Kelly, S. W., & Vokurka, R. J. (1998). The empirical assessment of construct
validity. Journal of Operations Management, 16, 387-405.
Pearlman, V., & Wallingford, M. S. (2003). Intergenerational wellness programming in
occupational therapy. Journal of Intergenerational Relationships, 1(2), 67-78.
Preston, C. C., & Colman, A. M. (2000). Optimal number of response categories in rating
scales: Reliability, validity, discriminating power, and respondent preferences. Acta
Psychologica, 104, 1-15.
Ribeiro O, Paúl C, Nogueira C (2007) Real men, real husbands: caregiving and masculinities
in later life. Journal of Aging Studies, 21(4), 302–313.
Roley, S. S., DeLany, J. V., Barrows, C. J., Brownrigg, S., Honaker, D., Sava, D. I., et al.
(2008). Occupational therapy practice framework: Domain & practice, 2nd edition.
American Journal of Occupational Therapy, 62(6), 625-683.
Schuling, J., de Haan, R., Limburg, M., & Groenier, K. H. (1993). The Frenchay activities
index: Assessment of functional status in stroke patients. Stroke, 24, 1173-1177.
Shahar DR, Schultz R, Shahar A, Wing RR (2001) The effect of widowhood on weight
change, dietary intake, and eating behavior in the elderly population. Journal of Aging
and Health, 13(2), 186–199.
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 2 6
Spiliotopoulou, G. (2009). Reliability reconsidered: Cronbach's alpha and paediatric
assessment in occupational therapy. Australian Occupational Therapy Journal, 56, 150-
155.
Toner J (2009) Small is not too small. Qualitative Social Work, 8(2), 179–192.
Vogt, D., King, D., & King, L. (2004). Focus groups in psychological assessment: Enhancing
content validity by consulting members of the target population. Psychological
Assessment, 6(3), 231-243.
World Health Organisation. (2001). International classification of functioning, disability and
health: ICF.
Wressle, E., Lindstrand, J., Neher, M., Marcusson, J., & Henriksson, C. (2003). The
Canadian Occupational Performance Measure as an outcome measure and team tool in a
day treatment programme. Disability and Rehabilitation, 25(10), 479-506.
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 2 7
Table 2: Framework Analysis of Focus Group and Interviews Showing Food Activity Verbs Used by Community-living Older adultsTheme ICF Code and Descriptor Associated words used by
participants (N = 5)Shopping* Shopping (d6200): Obtaining, in exchange for money, goods and
services required for daily living (including instructing and supervising an intermediary to do the shopping), such as selecting food, drink, cleaning materials, household items or clothing in a shop or market; comparing quality and price of the items required, negotiating and paying for selected goods or services, and transporting goods.
Going (shopping)/ shop / shopping
Basic economic transactions (d860): Engaging in any form of simple economic transaction, such as using money to purchase food or bartering, exchanging goods or services; or saving money.
Bought/ buy/ buying/ spending/ spend / spent
Cooking* Preparing meals (d630): Planning, organizing, cooking and serving simple and complex meals for oneself and others, such as by making a menu, selecting edible food and drink, getting together ingredients for preparing meals, cooking with heat and preparing cold foods and drinks, and serving the food. Inclusions: preparing simple and complex meals
Boil / cook/ cooking/ cooked/ cooks/ fried/ making/ makes/ mix/ offer (to others)/ prepare (food)/ roast/ roasts/ set (the table)/ use (cooking utensils)
Doing housework (d640): Managing a household by…storing food…, washing counters, walls and other surfaces; collecting and disposing of household garbage... Storing daily necessities (d6404): Storing food drinks… and other household goods required for daily living; preparing food for conservation by canning, salting or refrigerating, keeping food fresh and out of the reach of animals.
freeze/ keep (storage)
Eating* Eating (d550): Carrying out the coordinated tasks and actions of eating food that has been served, bringing it to the mouth and consuming it in culturally acceptable ways, cutting or breaking food into pieces, opening bottles and cans, using eating implements, having meals, feasting or dining. Exclusion: drinking (d560)
Ate/ chew / eat/ eaten/ eats/ enjoy (food and drink)/ nibble/ nibbling/ snack/ taste/ use (eating utensils)
Drinking Drinking (d560): Taking hold of a drink, bringing it to the mouth, and consuming the drink in culturally acceptable ways, mixing, stirring and pouring liquids for drinking, opening bottles and cans, drinking through a straw or drinking running water such as from a tap or a spring; feeding from the breast.
Drinking/ drink
Eating Out*
Not included Going (restaurant / lunch club/ church)
Eating healthily*
Managing diet and fitness (d5701): Caring for oneself by being aware of the need and by selecting and consuming nutritious foods and maintaining physical fitness.
Avoid / lose (weight)
Growing Food
Taking care of plants, indoors and outdoors (d6505): Taking care of plants inside and outside the house, such as by planting, watering and fertilizing plants; gardening and growing foods for personal use.
Gardening/ give (produce to others)/ grow/ growing/ grown/ keep (the garden)
*Included in the Occupational Performance Measure of Food Activities
Table 3: ‘Occupational Performance Measure of Food Activities’ Item Pool Questions organized by Food Activity
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 2 8
Domain, and Rating Scales Used
Subscale Questions
Grocery Shopping
I1 How important is grocery shopping to you?
P2 How well are you able to do your grocery shopping?
S3 How would you rate your satisfaction with the way you do your grocery shopping?
Cooking
I How important is cooking to you?
P How well are you able to cook?
S How satisfied are you with the way you cook?
Eating
I How important is eating to you?
P How well are you able to eat?
S How satisfied are you with your eating?
Eating Out
I How important is eating out to you? For example going to restaurants, a coffee shop, or a lunch club
P How well are you able to eat out?
S How satisfied are you with the eating out you do now?
Eating Healthily
I How important is eating healthily to you?4
P How well are you able to eat healthily?
S How satisfied are you with your healthy eating?
1. Food Activity Importance: 1 (Not important at all) to 10 (Extremely important)2. Food Activity Performance: 1 (Not able) to 10 (Extremely well)3. Food Activity Satisfaction: 1 (Not at all satisfied) to 10 (Extremely satisfied)4. Omitted from the final version of the tool following analysis of unidimensionality
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 2 9
Table 4: Demographic Characteristics of Phase 2 Participant Groups, with Chi-squared and Fisher’s Exact Test Results showing significance of differences between Survey and Interview Groups
Survey Interview Total Sample Between Group Differences
n = 38 % n = 39 % N = 77 %)Gender χ2 (1) = 0.29, p = .78
Male 6 16 8 21 14 18Female 32 84 31 79 63 82
Age χ2 (2) = 3.50, p = .19
60 – 69 8 21 15 38 23 3070 – 79 14 37 14 36 28 3680 – 90 16 42 10 26 26 34
Marital Status χ2 (3) = 1.69, p = .68
Never Married 7 18 5 13 12 16Married 8 21 11 28 19 25Divorced or Legally Separated
6 16 9 23 15 19
Widowed 17 45 14 36 31 40
Living Arrangements
Fisher’s Exact Test = 6.20, p = .04
Living as a Couple 8 21 13 33 21 27Living Alone 29 76 20 51 49 64Living with Family 1 3 6 15 7 9
Ethnic Group Fisher’s Exact Test = 1.59,
p = 1.0White British 35 36 71White Irish 2 1 3White, other 1 1 2Any other mixed background
0 1 1
Socio-economic Classification
Fisher’s Exact Test = 1.37, p = .83
Managerial and Professional Occupations
26 68 26 67 52 68
Intermediate Occupations
9 24 9 23 18 23
Lower Supervisory and Technical Occupations
2 5 1 3 3 4
Semi-routine and Routine Occupations
1 3 3 8 4 5
M e a s u r i n g F o o d A c ti v i ti e s P a g e | 3 0
Table 5: Results of Statistical Analysis of Phase 2 Data for Validity and Reliability, and Comparison of Scores Between Postal Survey and Interview Groups, using SPSS 15.0 software
Tests of Validity and Reliability
Kolmogorov – Smirnov Test of Normality
Independent Samples t – test Kolmogorov-Smirnov Z Test
Lower r
α ρ D df p t df p 95% CI Z p
Importance .01 .58 .22 .13 76 .003 .72 74 .48 [- .26, .55] .82 .52Performance .46 .75 .37 .12 76 .005 - .60 74 .55 [- .59, .32] .67 .76Satisfaction .34 .70 .31 .13 76 .003 - .11 74 .92 [- .41, .37] .42 1.00Occupational Performance of Food Activities
- .001 74 1.00 [- .61, .61] .66 .78