Association of informal caregiving with body mass index and frequency of sporting activities: evidence of a population-based study in Germany

Cross-sectional data from the fifth wave of the German Ageing Survey (Deutscher Alterssurvey; DEAS) were used. DEAS is an ongoing longitudinal, population-based study of the German community-dwelling population in the second half of life (40 years and over) started in 1996. The sample was drawn by means of national probability sampling and was systematically stratified by region (West and East Germany), age and gender. Follow-up waves took place in 2002 (second wave), 2008 (third wave), 2011 (fourth wave), and 2014 (fifth wave). Each wave comprised a panel-sample (participants who had already been interviewed) and a new cross-sectional sample, except for 2011, which was a pure panel survey. Face-to-face interviews are conducted in each wave. After the oral interview, respondents are asked to fill out an additional written questionnaire (so called ‘drop-off’ self-report questionnaire). Please see for further details Klaus et al. [14].

In the fifth wave, about 6000 participants were interviewed for the first time (response rate: 25%) while over 4000 participants had already been interviewed in former waves (response rate: 61%). In total, the response rate corresponds to that in other large German survey studies [15]. Out of the 10,324 participants in the fifth wave, the subsample of n = 1380 who provided informal care and filled out the ‘drop-off’ self-report questionnaire including, e.g., psychological scales and physical (chronic) illnesses, was used for the present analysis.

Please note that an ethical statement for the DEAS study was not necessary because criteria for the need of an ethical statement were not met (risk for the respondents, lack of information about the aims of the study, examination of patients).

The BMI was computed from self-reported height (meter) and weight (kg) as weight divided by height-squared. In adults, the BMI is widely used as a body weight classification system (e.g., for excess weight). Furthermore, individuals were asked “How often do you do sports such as hiking, soccer, gymnastics, or swimming?” (daily; several times a week; once a week; 1-3 times a month; less often; never). The frequency format is based on the International Physical Activity Questionnaire (IPAQ) [16]. Furthermore, the item used in our study was validated [17].

In additional analysis, the frequency of sporting activities was replaced by three different outcomes; including frequency of … (1) strenuous physical activities, (2) moderate physical activities, and (3) light physical activities. Again, the frequency format was: daily; several times a week; once a week; 1-3 times a month; less often; never.

The exact wording was as follows:

Individuals were asked if they provide informal care with the question: “Are there people you look after or care for regularly due to their poor state of health, either on a private or volunteer basis?” (no; yes). Among those informal caregivers, individuals were asked (yes; no) “What kind of care or assistance do you provide to the person who you help? A: Do you help around the house? B: Do you look after him/her or keep him/her company? C: Do you perform care services for the person you assist? D: Do you help in another way?” In a narrow sense, care can be seen as “nursing care” (C: “Do you perform care services for the person you assist?”).

Furthermore, informal caregivers should rate how much time they spend per week providing informal care (average number of hours per week, ranging from 0 to 168 h).

Based on theoretical considerations and previous findings the following socioeconomic variables were included in the regression analysis: Age, sex, family status (distinguishing between: married and living together with spouse; others (married and living separately, divorced, widowed, and single)), occupational status (employed, retired, other), and individual monthly net equivalent income (OECD scale). Moreover, the sum of physical illnesses (e.g., insomnia, eye problems, ear problems, cardiac and circulatory disorders or cancer; ranging from 0 to 11) was used as covariate. For example, it has been shown that age is associated with physical activities [18]. Furthermore, it has been shown that BMI is positively associated with younger age [19]. Moreover, it has been found that employment status is associated with physical activity [20]. In addition, it has been demonstrated that (1) cardiac and circulatory disorders, (2) respiratory problems, asthma, shortness of breath, (3) stomach and intestinal problems, (4) cancer, and (5) gall bladder, liver or kidney problems are associated with weight loss [21, 22].

Table 1 shows the sample characteristics stratified by type of informal caregiving. Among the caregivers, 65.1% provided help around the house, 83.2% looked after their relative, 27.9% provided nursing care services, and 67.9% provided any other type of help. The mean age of all informal caregivers was 63.4 years (SD: 10.4 years) with a range from 40 to 91 years. 60.0% were female.

The BMI was 26.8 kg/m² (SD: 4.5 kg/m²) on average. As Table 1 displays, there was a statistical significant difference between caregivers who provide nursing care services as compared to those who only provide other types of informal care; the BMI of the group providing nursing care services was about 0.7 kg/m² higher than the BMI of the group not providing nursing care services. Equally, there was a significant deviation in the frequency of doing sports between both aforementioned groups. Additionally, the frequency of doing sports deviated between the groups who provided help around the house and those who did not. With respect to other types of informal caregiving, there were no significant differences regarding frequency of doing sporting activities and BMI.

Table 2 splits the sample by the median of informal caregiving time per week. The median was 5 h per week and people who provided informal care of exactly 5 h per week were assigned to the group below the median so that this group is larger with n = 732 participants compared to n = 604 participants above the median. The latter group above the median was more likely to be female and somewhat older. In addition, there were significant differences in all considered potentially confounding variables. The BMI was significantly higher in the group that provided care more than 5 h per week. Besides, the frequency of doing sports deviated significantly between the two groups.

Table 3 displays the results of the multiple regression analyses with the BMI as dependent variable. Model 1 to 4 (marked as (1), (2), (3), and (4) in Table 3)) include the types of informal caregiving, model 5 (marked as (5) in Table 3) the time of informal caregiving per week. The same principle holds for Table 4. All models are fully adjusted for the various potential confounders (sex, age, marital status, number of illnesses, income, and occupational status). The regression models showed no association of BMI with three out of the four types of informal caregiving. Only the provision of nursing care services was associated with an increased BMI of about 0.6 kg/m². For example, for an individual of 1.70 m height performing nursing care services this increase in BMI equals an increase in weight of about 1.68 kg (ceteris paribus). Spending more time on providing informal care was associated with an increased BMI; each additional hour per week was associated with a higher BMI of about 0.01 kg/m².

With respect to the control variables, female gender, the number of illnesses, and net equivalent income per capita were positively associated with the BMI, whereas a higher age was negatively associated with the BMI in all five models.

The results of multiple ordered probit regression analyses with frequency of doing sporting activities are depicted in Table 4. Help around the house was associated with a lower frequency of doing sports activities (model (1)) after adjustment of the covariates. Equally, providing nursing care services was associated with doing fewer sporting activities (model (3)). The informal caregiving types of ‘looking after someone’ and ‘any other help’ were not associated with the frequency of doing sporting activities (models (2) and (4)). With respect to the time spent on informal caregiving, each additional hour spent on informal caregiving was associated with a lower frequency of sporting activities (model (5)).

Regarding the control variables, a higher number of illness was associated with a lower frequency of doing sports in all five models, and higher net equivalent income per capita was associated with higher frequency.

In additional analysis (please see Additional file 1: Tables S1, Additional file 2: Tables S2 and Additional file 3: Tables S3), the frequency of sporting activities was replaced by the frequency of (1) light (2) moderate, and (3) strenuous physical activities. While providing nursing care services and caregiving time were associated with the frequency of light physical activities, only caregiving time was significantly associated with the frequency of moderate physical activities. In contrast, none of the caregiving factors was associated with the frequency of strenuous physical activities.

Using a representative sample of individuals in the second half of life, this study aimed at examining whether caregiving time and caregiving activities are associated with BMI and the frequency of sporting activities. The results show that the BMI increased with caregiving time and performing nursing care services, whereas it was not associated with the other caregiving activities (help around house; looking after someone; any other help). Equally, the frequency of sporting activities decreased with caregiving time and performing nursing care services, whereas the outcome measure was not associated with the other caregiving activities (except for help around house). Additional analysis showed that nursing care services were only associated with the frequency of light physical activities, whereas it was neither significantly associated with moderate nor with strenuous physical activities.

While several studies examined whether informal caregivers and non-caregivers differ in lifestyle factors [11, 23], little is known about the associations between caregiving time as well as caregiving activities with lifestyle factors among informal caregivers. For example, it was found that 26.1% of full-time informal caregivers were obese compared with 22.6% of part-time informal caregivers in Thailand (n = 60,569) [24]. In addition, it has been shown that women providing care to an ill/disabled spouse had a higher BMI (mean BMI was 27.0) compared to women providing care to an ill parent/ill others (both, with mean BMI of 26.1) in 11 states of the USA (n = 54,411) [10]. However, physical activity (metabolic equivalent hours (MET-hours) per week) was similar between women providing care to an ill/disabled spouse and women providing care to an ill parent.

Our study adds further insights into the relationship between these lifestyle factors and informal caregiving time as well as several caregiving activities. The current study showed that the BMI is positively associated with caregiving time and performing nursing care services, whereas it was not associated with the other caregiving activities. A possible explanation might be that long caregiving time as well as performing nursing care services are associated with adverse factors such as stress or depressive symptoms [3] because (1) informal caregivers are generally not used to perform nursing caregiving activities and because (2) an increase in informal caregiving time is often the result of an increase in cognitive impairment of care-recipients [25]. This is supported by the fact that according to our calculations (German Ageing Survey, fifth wave), caregiving time was associated with increased depressive symptoms [26] (r = .19, p < .001) and increased stress [27] (r = .16, p < .001). These factors including depressive symptoms are in turn associated with a higher BMI [28] or a reduced frequency of physical activities [29].

Likewise, the frequency of sporting activities decreased with caregiving time and performing nursing care services, whereas it was not associated with the other caregiving activities. In the German Ageing Survey (fifth wave), the evaluation of leisure time (ranging from 1 = “very bad” to 5 = “very good”) was associated with caregiving time (r = −.30, p < .001). This might support the idea that an increase in caregiving time restricts leisure time. Furthermore, performing nursing care activities might be a source of stress, which might exhaust individuals’ resources. This is supported by the fact that performing nursing care services was associated with increased perceived stress (r = .13, p < .001) in the German Ageing Survey (fifth wave).

Thus, individuals might have a lack of energy or they might be too tired to exercise. However, this is speculative and consequently should be investigated in future studies. It should be acknowledged that caregiving time and performing nursing care services were only weakly associated with both outcome measures. For example, the partial η² for these two explanatory variables was .003 each (0.3% of the variability in BMI explained) in the models presented in column 3 and column 5 of Table 3. Consequently, the clinical significance might be limited or even small.

In contrast to performing nursing care activities, the other caregiving activities (except for help around house) might have a stronger focus on supervision or are strongly related to everyday household activities. In many cases, informal caregivers might be familiar with these activities, which might help to cope with the demands of informal care. Thus, it is assumed that individuals did not perceive the other caregiving activities as demanding, stressful or challenging. Consequently, it appears plausible that the other caregiving activities are not associated with both outcome measures.

This is one of few studies providing insights into the relationship between informal caregiving time, caregiving activities and the lifestyle factors of BMI and sporting activities. Data from the present study were gathered from a large, population-based study among community-dwelling individuals aged 40 and over. Data on caregiving time as well as on four caregiving activities were provided. A limitation of the study is that we cannot exclude a potential recall bias in participants’ estimations of average caregiving time. Thus, it is also likely that self-reported BMI is biased downwards because subjects generally tend to underestimate weight and overestimate height [30]. In addition, this study is cross-sectional, which does not allow analyzing the causal inference of the variables of interest. Thus, we cannot dismiss the possibility that changes in body mass index (BMI) and changes in the frequency of sporting activities affect the probability of starting informal caregiving. However, we assume that this is rarely the case.

Moreover, because a sample selection bias cannot be ruled out, our findings might be difficult to generalize to individuals with, e.g., low education. However, it has been shown that selectivity effects were rather small in the DEAS study [31].