Depression, nutritional risk and eating behaviour in older caregivers

Abstract

Objective

This study investigated the relationship between depression, nutritional risk and dietary intake in a population of older caregivers.

Design

Mailed questionnaire with sub group participating in a home-based interview.

Participants and setting

Seventy-six community dwelling caregivers aged 50 y or over from Victoria, Australia.

Measurements

Questionnaires provided information on weight, height, hours of care, depressive symptoms, nutritional risk and appetite. The home-based interview assessed dietary intake and shopping, cooking and meal consumption habits.

Results

The sample had a mean±SD age of 70.3±12.8 y, BMI of 27.2±4.8 kg/m² and the time spent caring was 101.8±68.1 h/wk. Overall, 32% of caregivers had depressive symptoms, 21% were at risk of malnutrition and 21% reported their appetite was fair/bad/very bad. Caregivers with depressive symptoms (32%) compared to those with no depressive symptoms (53%) had a poorer appetite (p<0.05). Of the 20 caregivers who participated in the home interview, 25% reported they ate their meals alone.

Conclusion

A significant proportion of community dwelling older caregivers had depressive symptoms, were at risk of malnutrition and had poor appetites, although the majority were overweight or obese.

Introduction

The world population is aging, with the number of people aged 60 years or over expected to triple over the next 50 years (1). Depression is a serious, widespread medical condition that affects the health and quality of life of older people (2). A landmark study by the World Health Organization concluded that the significance of illness burden attributable to depressive illness increases with age, and is expected to grow further with predicted demographic shifts, most notable through a greater proportion of the “old-old” (3). Our own research has demonstrated that the prevalence of depression among older people in Australia is 28%: 18% of older people in our study were diagnosed with Major Depressive Disorder and 10% with Minor Depressive Disorder (4). Depression is considered a source of excess morbidity and mortality, associated with disability in activities of daily living, higher levels of medical co-morbidity, and the requirement for higher levels of supported care, such as a nursing home admission (5, 6, 7). Consistent with findings among depressed people in the broader population, depressed older people in the community have tended to be higher utilisers of general medical services compared to those without depression (8).

The effects of depression on appetite and weight are conflicting. Depression has been linked with an increased appetite (9, 10, 11). Cross sectional studies have indicated that depression is positively associated with obesity (12, 13), and prospective studies have demonstrated that depression is associated with increases in weight (14, 15). In contrast, higher levels of depression have been associated with a reduced appetite (9), higher degrees of nutritional risk (16, 17) and treatment with an antidepressant has been found to improve nutritional status (18). Furthermore, one of the consequences of long term depression can include unwanted weight loss (19, 20, 21).

As a consequence of the aging population, there will be an increase in the number of people with disabilities and chronic illness who will need support and assistance. Simultaneously, there has been a shift in aged care services from institutions to the community (22), with many of the elderly receiving care from family and/or friends in the home environment. Caregivers are of all ages, but 29% are aged 60 years or more, many are on limited incomes and the amount of time spent caring each week can be significant, with half of those aged 65 years or greater performing caring duties for more than 40 hours per week (23). Caregiving itself is an arduous task and is a risk factor for mortality (24) and morbidity, including psychological disorders such as increased levels of distress (25) and depression (26). Caregivers are likely to be susceptible to depression and poor eating habits, yet relatively few studies have assessed levels of depression, nutritional risk and dietary intake among community based caregivers.

The purpose of the present study was to obtain preliminary pilot data on the relationship between depressive symptoms, nutritional risk, appetite and dietary intake in a population of older caregivers, using a mailed questionnaire with a sub group participating in a home-based interview.

Methods

Subjects

Participants were caregivers of family members providing care in their own home, recruited from three caregiver agencies (Benetas, Uniting Aged Care and Baptcare) in Victoria, Australia, from both metropolitan and rural areas. These agencies are not for profit providers of aged care services in Victoria. Caregivers supported by these agencies were eligible for the study if they were 50 y of age or greater. The case managers obtained expressions of interest from caregivers willing to be involved in the study. The study was approved by the Deakin University Human Research Ethics Committee, and written informed consent was obtained from the sub group of participants who participated in the home interview.

Study design

Anonymous questionnaires with a reply paid envelope were mailed to all primary family caregivers who were aged 50 years or over on the databases of the agencies participating in the study. For caregivers who were affiliated with Baptcare agency, the questionnaire was mailed on two occasions, whereas for caregivers affiliated with Benetas and Uniting Aged Care agencies, there was only one mailout. Case managers from the agencies were informed that the survey was taking place and participants were encouraged to seek assistance from the case managers to complete the questionnaire. The questionnaire assessed the estimated amount of time spent caring each wk, estimated anthropometry measurements, depressive symptoms, nutritional risk, appetite and hunger and foods consumed when stressed. At the end of the questionnaire, participants could indicate if they were interested in participating in a home interview to assess dietary intake.

Questionnaires

Anthropometry Measurements

Weight and height were self reported by participants. Body mass index (BMI) was calculated as weight (kg) divided by height² (m²).

Depression assessment

Depression level was assessed using the short form of the Geriatric Depression Scale (GDS-15) (27), which is considered a valid measure of depression in the elderly (28, 29, 30). The GDS-15 has the following cut off scores: 0-4, no depression; 5-8, mild depression; 9-11, moderate depression; and 12-15, severe depression. The GDS-15 measures levels of depressive symptoms rather than making a diagnosis of clinical depression.

Nutritional risk

Nutritional risk was assessed using the Mini Nutritional Assessment Short-Form (MNA®-SF) (31), which consists of six questions on food intake, history of weight loss, mobility, presence of psychological stress, acute disease or neuropsychological problems and BMI (kg/m²). The MNA®-SF was developed as a short screening tool based on the longer 12-item MNA® assessment tool; both the MNA® and MNA®-SF are considered valid tools for assessment of nutritional risk in the geriatric population (31, 32). The presence of neuropsychological problems was not assessed as self assessment would not be appropriate; each subject was given a score of two for this question. Therefore, in calculating the MNA®-SF score the assumption was made that no subjects had neuropsychological problems. The maximum score for the MNA®-SF is 14, with 12 points or greater indicating not at nutritional risk and 11 points or below indicating possible malnutrition.

Appetite assessment

Appetite was assessed using ten items from the 29-item Appetite, Hunger and Sensory Perception (AHSP) questionnaire (33). Participants rated their response to 10 questions on taste, appetite and hunger on a five point Likert scale (Taste: It seems that all foods have the same taste; I still eat with relish; In general, I find foods taste, Appetite: Nowadays my appetite is generally; Nowadays I do not feel too much like eating; Everyday I feel like eating, Hunger: How often do you feel like eating your breakfast? How often do you feel like eating your lunch? How often do you feel like eating your dinner? How often do you feel like eating a snack?). Scores for each question ranged from 1 to 5, with a higher score on each question corresponding to a positive response, for example having a better appetite or a greater desire to eat meals more frequently.

Home interview

Dietary assessment

Dietary intake was assessed by interview using a food frequency questionnaire previously used by Booth et al (34). Participants were asked to indicate their daily consumption of fruit or 100% fruit juice, vegetables and dairy, and their weekly consumption of meat/meat equivalents. Frequency of food consumption was categorised as follows: no consumption, 1 serve or less, 2-3 serves, 4-5 serves and 6 serves or more. Dietary intake was also assessed using a 24 h recall and then daily energy, macronutrient and micronutrient intake was determined using Foodworks Professional Edition (version 5; Xyris Software, Brisbane, Qld, Australia). Data were also collected on shopping, cooking and meal consumption habits using a series of structured questions.

Statistical analysis

Data were analysed using SPSS for WINDOWS (version 15.0.1; SPSS Inc, Chicago). Subjects’ baseline characteristics and scores on the GDS-15, MNA®-SF, Appetite, Hunger and Sensory Perception questionnaire and the food frequency questionnaire were analysed descriptively using means, standard deviations, counts and percentages. Normality of data was assessed using Kolmogorov-Smirnov statistics. All variables were normally distributed except for variables from the Appetite, Hunger and Sensory Perception questionnaire which were expressed as the median [IQR]; all other data were expressed as mean±SD or mean±SEM. Pearson’s correlations were used to analyse relationships between the GDS-15 and MNA®-SF and Spearman’s correlations to analyse the relationship between the GDS-15 and appetite. Chi-square and Mann-Whitney tests were used to analyse categorical data and Student’s t test for independent samples were used to assess continuous data. p<0.05 was considered to be statistically significant unless otherwise indicated.

Results

Subject Characteristics

A questionnaire was mailed to 314 caregivers and 76 (24%) completed the questionnaire. The characteristics of subjects are summarized in Table 1. Of the 76 subjects who completed the study, more women than men participated. Almost half of all subjects were over the age of 75 y (n=33; 43%). The mean BMI for the study participants was within the overweight weight range; two people were classified as underweight, 28% as normal weight, 36% as overweight and 22% as obese (35). The average amount of time spent caring was more than 100 h/wk, and there were 25 (33%) caregivers who reported caring for 168 h/wk (i.e. 24 h/d, 7 d/wk). Approximately one third of the subjects had depressive symptoms and one fifth was at risk of malnutrition. There was no difference in the mean score for depressive symptoms in males versus females, 4.1±0.6 versus 4.2±0.6 (p>0.05). For 15% and 26% of participants, there were no data for depressive symptoms and the MNA®-SF, respectively, as one or more questions were not completed on each questionnaire.

Table 1.

Characteristics of study participants (n=76)

	Mean±SD
Women/Men*	41/34
Age (y)	70.3±12.8
Weight (kg)	76.2±14.7
BMI (kg/m2)	27.2±4.8
Hours of care per wk (h)	101.8±68.1
	n (%)
Depression (GDS-15)
normal	40 (52.6%)
mild	16 (21.0%)
moderate	6 (7.9%)
severe	2 (2.6%)
N/A	12 (15.8%)
Nutritional assessment (MNA®-SF)
Not at risk	40 (52.6%)
Possible malnutrition	16 (21.1%)
N/A	20 (26.3%)

BMI: Body Mass Index; GDS-15: Geriatric Depression Scale; MNA®-SF: Mini Nutritional Assessment short-form; N/A, not available; *Number of women and men (data on gender missing for one participant)

Of the 76 participants who completed the questionnaire, 20 (26%) agreed to undertake the face-to-face interview. There were 12 women and 8 men with a mean±SD age of 72.3±12.1 y and a BMI of 28.2±5.1 kg/m². The mean±SD amount of time spent caring was 117.3±62.6 h/wk.

Questionnaire (n=76)

Table 2 shows the response to the 10 items from the AHSP questionnaire. Ten subjects (13.1%) totally agree/agree with the statement “Nowadays I do not feel like eating”. Twenty-one percent reported their appetite was fair/bad/very bad. Between 11 and 32% of subjects responded that they sometimes or seldom/never felt like eating a main meal or snack during the day.

Table 2.

Participant responses to the Appetite, Hunger and Sensory Perception questionnaire (n=76) (33)

Item	Participant response (n (%))
	Totally agree/agree		Disagree/totally disagree
It seems that all foods have the same taste	3 (3.9%)		70 (92.1%)
Nowadays I do not feel like eating	10 (13.1%)		61 (80.3%)
I still eat with relish	68 (89.4%)		5 (6.6%)
Everyday I feel like eating	64 (84.2%)		7 (9.2%)
	Very good/good	Fair	Bad/very bad
Nowadays my appetite is generally	59 (77.6%)	14 (18.4%)	2 (2.6%)
In general, I find foods taste	65 (85.5%)	10 (13.2%)	0
	Daily/often	Sometimes	Seldom/Never
How often do you feel like eating your breakfast?	63 (82.9%)	8 (10.5%)	4 (5.2%)
How often do you feel like eating your lunch?	65 (85.5%)	6 (7.9%)	3 (3.9%)
How often do you feel like eating your dinner?	66 (86.8%)	7 (9.2%)	2 (2.6%)
How often do you feel like eating a snack?	51 (67.1%)	17 (22.4%)	7 (9.2%)

When stressed, 18% reported they ate more, 58% ate the same and 17% ate less. Participants indicated the types of foods they consumed when stressed, and participants reported between one to six different foods. The majority of foods consumed when stressed (45 (65%)) were energy dense and typically high in refined carbohydrates and fat (e.g. chocolate, biscuits, cake, confectionery, potato crisps, take-away foods).

Depression and appetite

Table 3 displays nutritional data in relation to depression classification. There were no significant differences in age, amount of care provided each wk and BMI in those without depressive symptoms versus those with depressive symptoms. Individuals with depressive symptoms compared with those who had no symptoms of depression had a significantly lower score on the MNA®-SF by 13%, although both groups were not at nutritional risk. However, there was an inverse relationship between the GDS-15 and MNA®-SF, such that individuals who scored high on the GDS-15 (presence of depressive symptoms) were more likely to score low on the MNA®-SF (indicator of nutritional risk) (r=-0.380, p<0.05). Individuals with depressive symptoms compared to those without depressive symptoms also had significantly lower mean scores (mean difference range: -1.0 to -0.5) for the items related to appetite and taste, indicating a poorer appetite in those with depressive symptoms. Furthermore, there was a significant inverse relationship between the GDS-15 and all appetite items (except for “Everyday feel like eating” and “Feel like eating a snack”), indicating that individuals who scored high on the GDS-15 (presence of depressive symptoms) were more likely to score low on the appetite items (indicating a poorer appetite) (rho range: 0.3-0.8, p<0.01 for all).

Table 3.

Age, anthropometry, nutritional risk, and appetite, hunger and sensory perception in older caregivers classified as normal or having depressive symptoms (mild, moderate or severe)

	No depressive symptoms (n= 39)	Depressive symptoms (n=24)	p value*
Women/Men**	21/18	15/9	0.680
		Mean±SEM
Age (y)	70.4±2.0	68.0±2.8	0.479
Care per wk (h)	100.2±11.6	86.8±15.2	0.483
Weight (kg)	78.6±2.2	74.2±2.9	0.229
BMI (kg/m2)	27.8±0.7	27.0±1.1	0.536
MNA®-SF	13.1±0.2	11.6±0.5	0.004
		Median [IQR]
It seems that all foods have the same taste (range 1-5)	5.0 [1.0]	4.0 [1.0]	0.005
I still eat with relish (range 1-5)	5.0 [1.0]	4.0 [1.0]	<0.0001
In general, I find foods taste (range 1-5)	4.0 [1.0]	4.0 [1.0]	0.055
Nowadays my appetite is generally (range 1-5)	4.5 [1.0]	3.5 [1.0]	<0.0001
Nowadays I do not feel like eating (range 1-5)	4.0 [1.0]	3.5 [2.0]	<0.0001
Everyday I feel like eating (range 1-5)	5.0 [1.0]	4.0 [3.0]	0.004
How often do you feel like eating your breakfast? (range 1-5)	5.0 [0]	4.0 [2.0]	<0.0001
How often do you feel like eating your lunch? (range 1-5)	5.0 [0]	4.0 [2.0]	<0.0001
How often do you feel like eating your dinner? (range 1-5)	5.0 [0]	4.5 [2.0]	<0.0001
How often do you feel like eating a snack? (range 1-5)	4.0 [2.0]	4.0 [1.0]	0.465

BMI: Body Mass Index; MNA®-SF: Mini Nutritional Assessment short-form; *p<0.01 to adjust for multiple comparisons; ** Number of women and men

Home interviews (n=20)

There was no difference in mean age, BMI and hours of care provided between those who completed the home interview and those who did not (data not shown).

Dietary intake

Data derived from the home interview summarises the frequency of intake (serves/d) of fruits, vegetables, dairy and meat/meat equivalents (Figure 1). The median [IQR] intake for dairy and meat/meat equivalent either exceeded or met the recommended daily intakes (36) [dairy: 3.0 [3.0] serves/d (recommended- 2 serves/d); meat/meat equivalent: 1.0 [0.8] serves per day (recommended- 1 serve/d)]. However, for fruit intake, the median was less than the recommended daily intake [(1.8 [1.5] (recommended- 2 serves/d)], and 50% of subjects consumed one serve or less per day. For vegetable intake, the median intake was 2.5 [3.0] serves/d, which is less than the recommended 5 serves/d. Moreover, 15 (75%) of participants had less than 3 serves of vegetables per day.

Figure 1.

Frequency of consumption of fruits, vegetables, dairy and meat/meat equivalents (serves/d) in those participating in the home interview (n=20)

Dietary intake compared to recommended intake

Both men and women were having excessive intakes of energy (mean ± SD: 7.9±2.4 MJ/d; recommended intake: 6.2–6.7 MJ/d (37)) and percentage saturated fat (mean±SD: 14.0±4.8%; recommended intake: 8-10% (37)), but percentage fat met current recommendations (mean±SD: 32.8±8.7 %; recommended intake: 20-35% (37)). Both men and women were having inadequate intakes of calcium (mean±SD: 825.2±382.2 mg/d; estimated average requirement: 1100 mg/d (37)).

Shopping, cooking and meal consumption habits

With regards to shopping habits obtained from the face to face home interview (n=20), all caregivers responded that they participated in the shopping, and three of these caregivers stated they required assistance with this task. Thirteen caregivers when prompted expressed problems during shopping, which included difficulties with accessing and locating grocery items and reading food labels. Seventeen caregivers stated they did the cooking in their household and of these, four stated they received assistance with meal preparation. Seven reported eating take away meals on a regular basis (range: 1/month to 2/wk) which typically consisted of fast foods or restaurant meals. Five caregivers responded they ate their meals alone.

Discussion

In a population of older caregivers, we found that more than one third had depressive symptoms, one fifth were at risk of malnutrition, more than half were either overweight or obese and depressive symptoms were associated with a poorer appetite. Depression symptoms were common among the caregivers, with more than one third having either mild, moderate or severe depressive symptoms. This is comparable to the level of depressive symptoms measured using the GDS-15 in other populations of community dwelling elderly (38, 39, 40), although Osborn et al reported a greater prevalence where 77% of older people had depressive symptoms (41). In a systematic review investigating depressive disorders using diagnostic criteria (DSM, RDC, ICD), 22% of caregivers of dementia patients had a depressive disorder, which is less than what we detected in the present study (42). The response rate to our questionnaire was low (24%), possibly reflecting the high demands and strain typically experienced by caregivers. Consequently, the proportion of caregivers with depressive symptoms may have been underestimated in this study, as those with more severe forms of depression may not have participated in the study.

Obesity across the globe is increasing at an alarming rate (35) and is becoming a significant problem in the elderly. In Australia, 57% of adults aged 75 years or more are either overweight or obese (43). In the present study, the mean BMI was 27 kg/m² and we also found a high prevalence of overweight and obesity, with 58% of the caregivers falling into this weight range. In agreement with our findings, Aggarwal et al (44) and Castro et al (45) also reported a mean BMI of 27 Kg/m² in populations of older caregivers. Obesity is associated with co-morbidities such as hypertension, cardiovascular disease, type 2 diabetes and dyslipidemia (46), as well as a reduced quality of life (47). Obesity results from chronic energy imbalance, where energy intake exceeds energy expenditure over a long period (35). The high incidence of overweight and obesity in caregivers in the present study may be due to reduced levels of exercise, although physical activity was not measured. The burden of caring may mean that caregivers have less time to participate in healthy behaviours such as regular physical activity. Caregivers in the present study reported caring for approximately 100 hours each week or 14 hours per d, which equates to a significant proportion of the day. Burton et al found that high level caring was associated with insufficient time to exercise (48). In contrast to this, one study found that caregivers were more likely to walk more as a form of exercise when compared to non caregivers (49).

Obesity can also result from long term excessive intake of energy. In the present study, the recommended energy intake was exceeded by 18-27%. It should be highlighted that under reporting of energy intake by 24 hour dietary recalls, the dietary assessment tool used in the present study, are more prevalent in overweight and obese individuals (50, 51), therefore energy intake reported by caregivers in this study is likely to be an underestimate. Despite this apparent over consumption of energy, a significant proportion of caregivers in this study appeared to be at nutritional risk and potentially consuming inadequate amounts of dietary nutrients. One fifth of the caregivers in the present study were at risk of malnutrition as assessed by the MNA-SF, a screening tool developed especially for the elderly (31). Malnutrition is a serious problem among the elderly and is associated with adverse clinical outcomes such as poor wound healing (52), increased mortality (53) and lower functional ability (54); the caregivers in the present study may be at risk of developing these conditions. Many studies have documented the incidence of malnutrition in hospitalised and institutionalised elderly (55), but to the best of our knowledge, no studies have reported malnutrition in older community dwelling caregivers. However, our findings are similar to other populations of community dwelling elderly, which have found 10 to 22% were at risk of malnutrition as determined by the MNA (56, 57, 58). We also found that those caregivers who had depressive symptoms were more likely to be at nutritional risk; these findings are unique as it provides further insight into the interplay between depression and malnutrition in this vulnerable group of elderly caregivers.

Dietary intake was determined in a sub group of the caregivers during the home interview by a food frequency questionnaire. We found that 50% of caregivers consumed one serve or less of fruit per day, which is well below the Australian recommended intake of two serves per day (36). Moreover, 75% of caregivers had less than three serves of vegetables per day, where the recommended intake is five serves per day (36). This finding is in agreement with Castro et al who reported that caregivers consumed less fruit when compared to non caregivers (45) and in a population of community dwelling old-old non-caregivers, only 50% met the recommended requirements for fruit and vegetable intake (59). Inadequate consumption of fruit and vegetables is associated with a greater risk of developing age related chronic diseases (60). We found that depressive symptoms were associated with a poorer appetite; potentially contributing to the sub optimal dietary intake of fruits and vegetables, although we couldn’t examine this due to the small sample size.

Several potential limitations of this study need to be highlighted. Depressive symptoms were evaluated with the GDS-15, which is a screening tool used to detect depressive symptoms rather than a diagnostic tool. The cross-sectional design of this study prevents the detection of causal relationships between depression, nutritional risk and eating behaviour. Questionnaires obtained self reported data, for example weight and height, which can be under and overestimated, respectively. We were unable to investigate the relationship between levels of depression in the caregiver and the medical condition of the person receiving the care. It has been recently reported that caregivers of dementia patients have higher rates of depression when compared to caregivers of other medical conditions (42). This study is limited by a small sample size due to the low response, but this is not unexpected as caregivers are unlikely to have the time to participate in such a study.

Conclusion

This preliminary pilot study revealed that caregivers represent a vulnerable group across a number of levels, where a significant proportion of caregivers had depressive symptoms, were at risk of malnutrition, although the majority were overweight or obese. Although the levels of depressive symptoms were comparable to other populations of community dwelling elderly, the findings of this study are unique in that it demonstrates that many were at risk of malnutrition even though the majority were overweight or obese. Future prospective studies will need to ascertain the relationship between depression, nutritional risk and eating behaviour in this sample of caregivers. These findings emphasise the need for routine screening of depressive symptoms in elderly caregivers.

Acknowledgements: The authors acknowledge Lisa J Willenberg and Paige van der Pligt for collecting the data for this study. They also thank Prof. Sing Kai Lo for his advice regarding the statistical analyses of these results.

Financial disclosure: None of the authors had any financial interest or support for this paper.