Multiple sclerosis (MS), an immune-mediated demyelinating disease, is the leading cause of neurological disability among young and middle-aged adults (National Multiple Sclerosis Society, 2014), with a peak onset at about age 33 yr (Simmons, 2010). Given the onset of MS at such an early age, the effects of this disease may be more detrimental than those of many other neurological conditions, with its gravest impact on everyday functioning and overall quality of life. People diagnosed with MS are usually young, highly functioning people who struggle to adjust to a disease that is unpredictable and varying in its course. Moreover, at the age of diagnosis, many people are just beginning to make crucial life decisions regarding employment and family, which are hindered by their disease.

Perhaps one of the most devastating consequences of MS is loss of employment, which occurs with a frequency ranging from 24% to 80% (Aronson, 1997; Kornblith, LaRocca, & Baum, 1986). Although more than 90% of people with MS are gainfully employed before developing MS (LaRocca, Kalb, Kendall, & Scheinberg, 1982; Pompeii, Moon, & McCrory, 2005), 70%–80% become unemployed within 5 yr of diagnosis (Kornblith et al., 1986). Given that employment is critical to a sense of identity, self-esteem, social contact (Johnson et al., 2004), and overall mental and physical health (Linn, Sandifer, & Stein, 1985; Solari & Radice, 2001), identification and management of the factors most accountable for employment difficulties in people with MS are paramount.

Impairments in performance of activities of daily living (ADLs), social and community life, family life, household responsibilities, and instrumental activities of daily living (IADLs) have been noted in people with MS. In a cohort of 166 people with MS, Einarsson, Gottberg, Fredrikson, von Koch, and Holmqvist (2006)  found that only 52% were independent in personal ADLs and even fewer (30%) in IADLs. Additionally, only 35% reported having a normal frequency of social and lifestyle activities. Thus, ADLs and social and lifestyle activities were found to be greatly affected in about two-thirds of people with MS. Impairments in such vital life activities have been found to result in poor outcomes such as decreased quality of life (QOL) and well-being (Goverover et al., 2005; Goverover, O’Brien, Moore, & DeLuca, 2010).

By accurately identifying the factors linked to activity limitations (e.g., cooking) and participation restriction (e.g., employment), occupational therapy practitioners may be better able to address some of the potential, and possibly even modifiable, contributing factors. The result will be an improvement in practitioners’ ability to assist clients with MS in maintaining employment and other important life activities, thereby ultimately improving their overall QOL.

Although certain physical limitations and disease processes are known to contribute to a person’s ability to maintain employment and fulfill household responsibilities, cognitive impairment is also a significant contributor. Cognitive difficulties are well documented in people with MS (Chiaravalloti & DeLuca, 2008), of whom 43%–70% are affected. Impairments are most commonly seen in the domains of processing speed, learning, and memory and, to a lesser extent, in visuospatial processing and executive functions.

Additionally, it is well documented that cognitive impairments affect a person’s ability to engage in basic ADLs (i.e., basic self-care), IADLs (i.e., activities that support independent life, such as shopping and driving; Kalmar, Gaudino, Moore, Halper, & Deluca, 2008), and community and social life (Einarsson et al., 2006; Johnston, Goverover, & Dijkers, 2005; Miller & Dishon, 2006; Schultheis, Garay, & DeLuca, 2001). Several studies have demonstrated the important role of cognition in employment status (Beatty, Blanco, Wilbanks, Paul, & Hames, 1995; Benedict et al., 2006; Honarmand, Akbar, Kou, & Feinstein, 2011; Morrow et al., 2010; Strober et al., 2012; Strober, Chiaravalloti, Moore, & DeLuca, 2014). Impairments in processing speed, verbal learning, memory, and working memory consistently differentiate employed from unemployed people, with the largest effect observed on measures of processing speed (Strober et al., 2012, 2014).

The overall aim of the present investigation was to examine the relationship between measures of cognition and measures of activity and participation in a sample of people with MS and to identify common variables associated with levels of both activity and participation. Specifically, we sought to explore and identify variables associated with performance of the IADL of cooking breakfast and participation in employment. We chose breakfast preparation and employment to represent activity and participation, respectively, because both are universal and culturally acceptable and because both are part of people’s daily routine and of the foundation of human occupation (Pierce, 2001).

Participants were 72 people diagnosed with clinically definite MS in accordance with criteria by McDonald et al. (2001) . Participant characteristics are presented in Table 1. Participants were ages 29–60 yr (mean [M] = 48.5, standard deviation [SD] = 8.3) with 12–20 yr of education (M = 15.5, SD = 2.2). Seven of the 72 participants (9.7%) were male. Fifty-four participants (75.0%) had relapsing–remitting MS, 7 had primary progressive MS, and 8 had secondary progressive MS; 3 participants had an unknown disease course. Mean duration of time since diagnosis with MS was 12.7 yr (SD = 8.1), ranging from 7 mo to 36 yr. Potential participants were excluded from the current study if they had a history of neurological illness or injury aside from MS, a history of major psychiatric illness, a history of alcohol or drug abuse, or severe visual impairment (could not read a short paragraph) or significant motor impairment (could not sign their name) that might interfere with the study procedures (i.e., testing and performance of study tasks). At least 1 mo had passed since participants’ most recent exacerbation, and they had been free of corticosteroid use for at least 1 mo.

Employment status was determined by participant statements describing their present work situation. Employment status was dichotomized as employed full time or part time or unemployed or disabled. People who reported being retired because of age were not included in the study.

Participants were asked to respond to five questions related to their cooking ability:

1. Do you cook? (Answers were 0, not cooking, or 1, yes, I cook.)

2. How often do you cook? (Answers could range from 0, never, to 4, always.)

3. How much previous cooking experience do you have? (Answers could range from 0, none, to 3, a lot.)

4. Have you ever cooked breakfast foods? (Answers were 0, no, or 1, yes, I cook breakfast foods.)

5. If you do cook breakfast foods, how often do you do so? (Answers could range from 0, never, to 4, always.)

Responses were summed for each participant, and overall scores could range from 0 (not cooking at all and no previous cooking experience) to 13 (cook often in the present and past). A cutoff score was set for cooking ability based on a median split: Participants with scores below the cutoff were labeled “nonactive cookers” (i.e., they did not cook often or did not cook at all), and those above the cutoff were labeled “active cookers” (i.e., they cooked frequently in the present and in the past).

Cognitive functions were evaluated using measures included in the Minimal Assessment of Cognitive Function in MS (MACFIMS) battery (Benedict et al., 2002, 2006). The MACFIMS was compiled to use as a cognitive test battery specific to people with MS by the Consortium of MS Centers. It contains the minimum number of measures necessary for the comprehensive cognitive assessment of people with MS; these measures are briefly described in the paragraphs that follow.

Processing speed and working memory were assessed using the Symbol Digit Modalities Test (SDMT)–Oral Version (Smith, 1982) and the Paced Auditory Serial Addition Test (PASAT; Gronwall, 1977). For detailed description of reliability and validity, see Sonder, Burggraaff, Knol, Polman, and Uitdehaag (2014) . The SDMT is a relatively quick assessment demonstrated to be valid for people with MS (Benedict, 2005; Benedict et al., 2006). The dependent variable in the current study was the number of the correctly completed numbers corresponding to geometric symbols during the 90-s trial. The PASAT is used to assess attentional processing and is valid and reliable to use with people who have MS (Tombaugh, 2006). The dependent measure for the PASAT was the number of the correct responses across both the 2-s and 3-s trials.

Visual–spatial memory was assessed using the Brief Visuospatial Memory Test–Revised (BVMT–R; Benedict, 1997), commonly used to evaluate visuospatial memory abilities in people with MS (Benedict et al., 2006). Interrater reliability of .97 and test–retest reliability ranging from .60 for Trial 1 to .84 for Trial 3 have been established (Benedict et al., 2006), as have construct and discriminant validity. The dependent measure representing the BVMT–R was the summed score of the three learning trials (ranging from 0, did not remember anything, to 36, remember all items) and the delayed recall score (range = 1–12).

Episodic memory was assessed using the Open Trial Selective Reminding Test (OT–SRT; Buschke, 1973). Number of trials to reach criterion and number of items participants remembered at delayed recall served as the dependent variables. Chiaravalloti, Balzano, Moore, and DeLuca (2009)  established norms for OT–SRT administration in people with MS. The standard SRT, from which the OT–SRT was derived, has shown good internal consistency reliability (range = .41–.62, all significant) and test–retest reliability (.66–.73; Lezak, 1995).

Generative verbal fluency was assessed using the Controlled Oral Word Association Test (Benton & Hamsher, 1989). The total score for the three 1-min trials was the dependent variable for data analysis. Test–retest reliability was good (r = .74), and internal consistency was high (r = .83; Ruff, Light, Parker, & Levin, 1996).

Visual–spatial perception was assessed using the Judgment of Line Orientation (JOLO; Benton, Sivan, Hamsher, Varney, & Spreen, 1994). The JOLO is frequently used as a motor-free method of evaluating visuospatial processing. In patient populations, the test–retest reliability of the JOLO has been found to be as high as .90 (Franzen, 2000). Internal consistency was also found to be high (α = .90) in a rehabilitation sample (Qualls, Bliwise, & Stringer, 2000). The dependent variable from this measure was the number of correct choices (0 points for any other response), with a maximum possible score of 30.

Executive functions were measured with the Delis–Kaplan Executive Function System (D–KEFS) Sorting Test (Delis, Kaplan, & Kramer, 2001). The D–KEFS Sorting Test has been shown to effectively distinguish between healthy control participants and people with different clinical conditions, including MS (Benedict et al., 2006). Reliability has been reported to be consistently high (.80; Delis, Kramer, Kaplan, & Holdnack, 2004). The highest possible total confirmed correct sorts for the free-sorting condition for the first card deck is 8; total confirmed correct sorts are the number of correct sorts for which the verbal description is awarded 1 or more points. The examinee’s verbal description is scored in terms of accuracy and level of abstraction, with a maximum of 4 points per description of a particular sorting rule (free description score); see Delis et al. (2001)  for detailed scoring procedures. For the current study, the variables of interest were the total free description score and the total confirmed correct sorts.

To assess participants’ levels of depression and anxiety, the Chicago Multiscale Depression Inventory (CMDI; Nyenhuis et al., 1998) and State–Trait Anxiety Inventory (STAI; Spielberger, 1983), respectively, were administered. The CMDI summed total score was used as a dependent variable. The STAI is a standardized, well-established measure of anxiety, with higher STAI State and Trait scores indicative of increased anxiety. Both STAI State and Trait scores were used as dependent variables. Lastly, fatigue was measured with the Modified Fatigue Impact Scale (Fisk, Pontefract, Ritvo, Archibald, & Murray, 1994).

We recruited participants with MS through advertisements, support groups, and the participant database at the Kessler Foundation and invited them to participate in a larger study examining the use of self-generated learning to increase learning and memory in people with MS. Potential participants were screened by phone for the inclusion and exclusion criteria, and qualified participants were scheduled for testing. Before enrollment in the study, participants signed a consent form approved by the institutional review board that indicated their willingness to participate. Once enrolled in the study, participants completed the battery of neuropsychological tests and questionnaires to assess affect symptoms, fatigue, cooking ability, and employment status.

To explore which variables have a greater impact on employment and IADL status (i.e., cooking ability), we performed two data analyses. First, we examined the correlation between cognitive abilities, demographics, and affective symptoms among employed and unemployed participants and among active cookers and nonactive cookers using point biserial correlation. Second, we identified four functional groups: Group 1 consisted of 21 participants who were employed and cooked; Group 2 consisted of 30 participants who were unemployed and cooked; Group 3 consisted of 7 participants who were employed but did not cook; and Group 4 consisted of 10 participants who were unemployed and did not cook. To identify cognitive and affective variables that might be different among those four groups, we conducted two-way analyses of variance (ANOVAs) comparing participants who cooked and did not cook and participants who were employed and were unemployed on each of the cognitive, affective, and demographic measures.

Point biserial correlation coefficients are presented in Table 2. Employment status was significantly correlated with years of education; participants who were employed had more years of education. In addition, being unemployed was associated with greater reports of fatigue and worse performance on measures of processing speed (SDMT) and visual memory (BVMT–R). Cooking ability was significantly correlated with the cognitive domains of processing speed, verbal memory, and working memory (PASAT), with better performance related to cooking ability.

To identify variables that were common among the four functional groups of cooking and employment and those that were distinct, we used a 2 (employed vs. unemployed) × 2 (cook vs. do not cook) ANOVA to examine the effect of employment status and cooking ability on scores on each of the cognitive tests and affect symptoms. Significant results are presented in Table 3. We found a main effect of cooking ability; participants who cooked performed better on cognitive tests than participants who did not cook, regardless of their employment status. This finding was observed on the PASAT, F (1, 63) = 4.3, p = .04; the Controlled Oral Word Association Test, F (1, 63) = 4.9, p = .03; the SDMT, F (1, 63) = 8.4, p = .005; and the OT–SRT number of trials to reach criterion, F (1, 63) = 4.9, p = .03. The main effect of employment was not significant for any of the dependent variables. However, the interaction between cooking and employment on the BVMT–R scores approached significance for both immediate and delayed recall, F (1, 63) = 3.4, p = .06, η² = .05. Participants who cooked and were employed remembered more items (M = 53, SD = 8.7) than people who did not cook and were employed (M = 42, SD = 8.2). Affect symptoms, disease variables, and fatigue did not differ significantly among the groups.

This study examined the effect of cognition on both activity level (cooking ability) and participation (employment status) in daily life. The main finding of this study is that people with MS who reported that they cooked obtained higher scores on cognitive tests regardless of their employment status. Thus, cognitive abilities are significantly related to cooking ability when both cooking ability and employment status are considered.

Cooking ability was associated with better performance on measures of information processing speed, working memory and executive control, and verbal learning and memory. These results are consistent with those of a previous study in which better performance on a cooking activity, specifically preparation of a casserole, was associated with better learning and memory, executive control, and processing speed (Kalmar et al., 2008). Kalmar et al. (2008)  concluded that performance on measures of executive functions, processing speed, and new learning were predictive of the degree of independence in performance of activities of daily living. To our knowledge, our current study and Kalmar et al.’s study are the only ones to investigate the relationship between cooking ability and cognitive abilities in MS. Taken together, the findings indicate that deterioration in the frequency and quality of cooking ability may be an indicator of cognitive decline in MS.

Only 40% of our participants were employed. Most recently, Schiavolin et al. (2013)  reported a mean unemployment rate in people with MS of 59%. Previous studies have demonstrated that loss of employment following the diagnosis of MS occurs with a frequency ranging from 24%–80% (Julian, Vella, Vollmer, Hadjimichael, & Mohr, 2008). Furthermore, unemployment increases consistently with severity and duration of the disease (Berg, Lindgren, Fredrikson, & Kobelt, 2006; Julian et al., 2008; MacLurg et al., 2005; Putzki et al., 2009; Strober et al., 2012). Interestingly, we did not observe such a pattern in the current study. Instead, we found that duration of illness was not associated with employment status and that MS disease course did not differ significantly among the four functional groups.

Unemployment in our sample was associated with poorer performance on measures of information processing speed and visual learning and memory, greater fatigue, and less education, consistent with previous findings. In particular, Morrow et al. (2010)  found that among people with MS employed at baseline, declines in processing speed and verbal memory were the most consistent predictors of clinically meaningful functional decline. Additionally, over the course of 3 yr, participants whose working status was stable deteriorated less than those who had a negative change in working status. The findings of the current study replicated previous findings in which processing speed was a pivotal cognitive skill related to employment status (Strober et al., 2012, 2014).

In terms of differences among participants who were employed and cooked, who were unemployed and cooked, who were employed but did not cook, and who were unemployed and did not cook, results demonstrate that participants who cooked performed at higher levels on cognitive tests than those who did not cook, regardless of employment status. Thus, participants in Group 2 (were unemployed and cooked) obtained higher scores on cognitive tests than those in Group 4 (were unemployed and did not cook). In addition, Group 1 (were employed and cooked) obtained higher scores on cognitive tests than Group 3 (were employed and did not cook). Thus, although the literature provides considerable evidence to support differences in cognitive abilities between employed and unemployed people with MS, in the current study, when we divided participants into groups based not only on employment but also on cooking ability, we found significant group differences in terms of cooking ability but not employment status.

More research is necessary to examine this finding for several reasons. First, we did not have access to information regarding the type of employment of our participants. Thus, a person could be employed in the capacity of, for example, receptionist or chef, professions that require different sets of cognitive skills and training. In addition, cooking ability was limited to breakfast foods, which vary considerably in range of foods and cognitive requirements for preparation.

Few studies have explored how people with MS experience participation in domestic life activities such as meal preparation and associated nutritional behaviors (Plow & Finlayson, 2012; Plow, Finlayson, & Cho, 2012). Plow et al. (2012)  found that people with MS indicated that nutritional self-efficacy and self-management skills, rather than functional limitations, may interfere with healthy nutritional behavior. Their main difficulties in food preparation and food shopping may stem from fatigue and mobility impairments (Plow & Finlayson, 2012).

The impact of cognition on functional outcomes is unquestionable; both cooking and employment require various cognitive skills such as planning, organization, mobility, execution, and follow-through. It may be that people with MS choose to leave their work because they feel they cannot cope with the work requirements. However, they may still prepare breakfast foods. Thus, when a person with MS reports that he or she does not cook in addition to not working, this may be an indicator of significant cognitive impairment.

In the current study, affective symptoms such as depression and anxiety were not associated with either employment status or cooking ability, and such symptomatology did not differ among the functional groups. Previous research has been inconsistent regarding affective symptoms and their role in employment status in people with MS. For instance, Honarmand et al. (2011)  found that employment was predicted by both people’s physical and cognitive disability status and their depression and personality traits. Strober et al. (2012), however, did not find a significant difference between employed and unemployed people on self-reported scales of depression. It is possible that in many studies (including this one), participants with major psychopathology such as depression were excluded; therefore, the sample of participants in the studies could represent a restricted range of affective symptoms. More research is needed to clarify the role of emotional symptomatology in activity and participation in people with MS.

This study has several limitations. First, although the sample was characterized by various measures describing disease-related variables, a measure characterizing disease severity and neurological disability (e.g., Expanded Disability Status Scale scores; Kurtzke, 1983) was not available. Disease severity and neurological disability should be evaluated in future studies in regard to their relationship to activity and participation. In addition, participants in this study were restricted to those with intact visual and general motor functions. Given that motor skills and vision are important for functioning in both cooking and employment, results of this study are limited in their generalizability. Additionally, our sample consisted of 7 men and 65 women, which does not reflect the male–female ratio in the MS population. Serial cross-sectional assessments, however, have provided evidence of a general increase in women with relapsing–remitting MS (Koch-Henriksen & Sørensen, 2010).

Second, cooking and employment were assessed by means of self-report. Although employment status is a relatively objective variable, cooking is not. The question “Are you employed?” requires a yes-or-no answer. The question “Do you cook?” may yield a more variable response (once a week vs. every night, complex meal vs. cereal) and, thus, make the response more subjective. Therefore, future studies should assess the performance of cooking more objectively.

Additionally, with regard to employment, participants reported only whether or not they were employed, resulting in two binary categories of employment. Therefore, more detailed information about employment, such as the nature of the work, level of training, and use of accommodations or reduced hours, could provide richer information. Cooking ability was limited in this study to breakfast food preparation, as noted, and we did not gather data on complexity and levels of difficulty. Similarly, we did not assess extraneous factors that may have played a role in participants’ choice to leave the workforce or continue cooking breakfast. Such factors might include socioeconomic status (e.g., Does the participant have to work?), interest level (e.g., Does the participant enjoy working or cooking?), and personality (e.g., Does the person value such tasks?). These variables should be considered for inclusion in future work.

To our knowledge, this is the first study to examine the relationship between measures of cognition and measures of activity and participation in a sample of people with MS. The findings of the current study represent a first step in understanding the relationship and differences between cooking and employment and cognitive abilities in people with MS. Our findings provide preliminary evidence with the following implications for occupational therapy practice:

This research was supported by the National Multiple Sclerosis Society (Grant No. RG 3935A2/2). This study was presented in part at the annual meeting of the Consortium of Multiple Sclerosis Centers, Orlando FL, June 2013.