We hypothesized that chocolate preference would be related to health and psychological well-being in old men.
Design, setting and participants:
We have followed up a socio-economically homogenous group of men, born in 1919–1934, since the 1960s. In 2002–2003, a mailed questionnaire was used to assess the health and well-being (including questions related to positive life orientation, visual analogue scales and the Zung depression score) of survivors. In addition, candy preference was inquired. Those men who reported no candy consumption (n=108) were excluded from the analyses.
Psychological well-being in old age.
The response rate was 69% (1367 of 1991). Of the respondents, 860 and 399 preferred chocolate and other type of candy, respectively. The average age in both candy groups was 76 years. Of the respondents, 99% were home-dwelling, 96% were retired and 87% were presently married, without differences between the candy groups. Men preferring chocolate had lower body mass index and waist circumference, and they also reported more exercise and better subjective health (P=0.008) than other candy consumers. Variables related to psychological well-being were consistently better in those preferring chocolate. The differences were statistically significant in feeling of loneliness (P=0.01), feeling of happiness (P=0.01), having plans for the future (P=0.0002) and the Zung depression score (P=0.02).
In this socioeconomically homogenous male cohort, chocolate preference in old age was associated with better health, optimism and better psychological well-being.
The Academy of Finland, the Päivikki and Sakari Sohlberg Foundation, the Helsinki University Central Hospital and the Finnish Foundation for Cardiovascular Research.
Favourable effects of chocolate consumption on health and especially cardiovascular diseases have been suggested in numerous studies (Ariefdjohan and Savaiano, 2005; Buijsse et al., 2006; Ding et al., 2006; Engler and Engler, 2006). Dark chocolate is also together with wine, fish, fruits, vegetables, garlic and almonds, one component of the ‘Polymeal’, a combination of natural, health-promoting substances (Franco et al., 2004). Chocolate contains several ingredients with biological, for example antioxidant, vasoactive and antithrombotic activity (Ariefdjohan and Savaiano, 2005; Pearson et al., 2005; Ding et al., 2006; Engler and Engler, 2006; Schroeter et al., 2006). Interestingly, beneficial endothelial effects may be greater in older people (Ferri et al., 2006; Fisher and Hollenberg, 2006). Besides cardiovascular effects, ingredients of chocolate have also potential effects on mood (Parker et al., 2006). Because undernutrition is common in the community living elderly (Martin et al., 2006), the palatable and delicious chocolate might even offer possibilities to maintain well-being and health in old age.
However, final proof of clinical benefit requires studies with clinically meaningful end points, and they are still lacking for chocolate. Physiologic changes may have no clinical relevance and epidemiological observations can be biased by the possibility that – similarly to wine drinking (Mortensen et al., 2001; Rimm and Stampfer, 2002) – chocolate preferers may simply be protected by other beneficial effects of their lifestyle (‘the healthy user bias’). Thus, controlling for confounders, especially social class (Marmot et al., 2001), is important in observational studies. In the 2002–2003 survey of our long-term Helsinki Businessmen cohort (Strandberg et al., 2004), we also asked about candy preferences and consumption in the participants with an average age of 76 years. In the present cross-sectional analyses, we have related the use of chocolate vs other sugar confectionery to health and psychological well-being in old men. Our cohort is socioeconomically homogenous and thus the setting offered a clearer test than population studies of the effects of chocolate per se, because the possible influence of social class on candy preference was minimized.
Outline of the present study is shown in Figure 1.
Earlier examinations during the 1960s and 1970s
The cohort and earlier examinations have been described in detail (Miettinen et al., 1985; Strandberg et al., 2004, 2003). Initially healthy men, mostly business executives born in 1919–1934, had participated in structured health check-ups (with clinical examinations and laboratory tests including serum cholesterol and 1-h postload glucose) during the 1960s and early 1970s at the Institute of Occupational Health in Helsinki. Of the original population baseline data and follow-up information with national registers are available for 3277 men. In 1974, they were evaluated with questionnaires, clinical examinations and laboratory tests and they were asked how they rated their present health and physical fitness on a 5-step scale (‘very good’, ‘good’, ‘fair’, ‘poor’ and ‘very poor’). The men were also asked to recall their weight at age 25 (Strandberg et al., 2003), whereupon the weight gain up to midlife could be calculated.
The 2000 survey of health-related quality of life
In 2000, we sent a mailed questionnaire to all 2110 survivors of the initial cohort, and 1858 men (88% of eligible) responded. The questionnaire included items on demographic variables, present weight, lifestyle variables and diseases. In addition, the Finnish version of the RAND-36-Item Health Survey 1.0 (Aalto et al., 1999; Hays and Morales, 2001) was embedded in the questionnaire. We used the physical and mental component summary scores (PCS, MCS, 19) to compare respondents and non-respondents of the study performed in 2002–2003 (focus of the present analyses).
The 2002–2003 survey of health and well-being
In 2002–2003, we renewed our mailed questionnaire survey to all survivors (n=1991) of the initial cohort (re-mailed once for non-respondents). The questionnaire included the same items as the 2000 survey (symptoms and diseases, present medications, present weight), but this 2002–2003 questionnaire also included a question of the frequency of candy consumption (never, sometimes, weekly, daily), and the preferred sugar confectionary (chocolate, licorice, other candy). Except frequency, the amount of preferred candy consumed was not asked. Neither was the type of chocolate (dark chocolate, milk chocolate) further inquired, but both types of chocolate are consumed in Finland. Consequently, it was not possible to assess the intake of flavanols from chocolate (nor from other food items) in the present survey. Because the focus of the present analysis was on chocolate, we compared chocolate vs preferers of licorice or other candy. We also present baseline data of the men who reported that they never consumed candy (n=108), but these data are for descriptive purposes only. Statistical comparisons with consumers are not performed, because the reason for non-consuming was not asked, and may conceivably be due to some illness, which would confound the comparisons.
In addition, the 2002–2003 questionnaire contained several questions about attitudes towards life, which we have used in another study of the elderly (Pitkala et al., 2004). The following questions were asked: (1) Are you satisfied with your life? (yes/no), (2) Do you have zest for life? (yes/no), (3) Do you feel needed? (yes/no), (4) Do you have plans for the future? (yes/no), (5) Do you suffer from loneliness? (seldom or never/sometimes/often or always) and (6) Do you feel depressed? (seldom or never/sometimes/often or always). These domains have been suggested to be major components of psychological well-being among older people (WHO, 2003; Brown et al., 2004; GO, 2005). Positive life orientation was assessed to be present if the participant answered yes or seldom/never to these six questions, and it has been shown to predict mortality in the elderly (Pitkala et al., 2004).
The participants were also asked to rate their whole life course (life experiences, fullness of life, composition) using the Finnish school marks from 4 (worst) to 10 (best). Visual analogue scales (VAS; 10 cm) were used to assess self-rated health (0=worst, 10=best), present global happiness (0=very unhappy, 10=very happy) and personal work history (0=extremely stressful, 10=not at all stressful). Negative effect was further assessed with the Zung self-rated depression scale (Zung, 1965), widely used in epidemiological studies, embedded in the questionnaire. A person with a Zung score below 45 points was considered normal.
NCSS statistical software (www.ncss.com, Kaysville, UT, USA) was used for the analyses. T-test and non-parametric tests were used to compare continuous variables and χ2 tests to compare proportions. Multivariate regression was used to determine independent predictors with odds ratios (OR), and the 95% confidence intervals (CI). P-values <0.05 were considered statistically significant.
In 2002–2003, 1374 men (69% of eligible) with an average age (with s.d.) of 76 years responded to our survey. According to the earlier 2000 survey (where response rate was 88%), non-respondents in 2002–2003 reported to have significantly more hypertension, diabetes, congestive heart failure, cancer and memory disturbances than respondents. The non-respondents also had clearly lower PCS (43.6, s.d. 10.2 vs 46.3, s.d. 9.0, P<0.0001), as well as lower MCS (50.7, s.d. 11.3, vs 53.7, s.d. 9.6, P<0.0001) in 2000 than respondents.
Of the respondents, 108 reported that they did not use candy. Of the 1259 men who consumed candy and reported their preference, 860 men preferred chocolate and 399 men other sugar confectionary. Among them, 99% were home living, 96% were retired and 86% were currently married, and there were no statistical differences between the candy preference groups. The frequency of candy consumption was moderate – over two-thirds consumed candy ‘sometimes’ – and the consumption pattern was similar in the various preference groups (Table 1).
In this follow-up study, it was also possible to compare the earlier characteristics of the candy preference groups defined in 2002–2003. These comparisons showed that the traditional risk factor levels during the 1960s, and the self-rated health and physical fitness in 1974 were similar in the candy preference groups (data not shown). However, the chocolate group had gained less weight from the age 25 years up to midlife than the other candy group (8.6 kg, s.d. 7.7 vs 10.2 kg, s.d. 8.0, respectively, P=0.003). Weight gain among those men, who did not consume candy, was 10.4 kg (s.d. 7.2).
Characteristics related to health and well-being in 2002–2003
Characteristics of the study groups in 2002–2003 are shown in Table 2. Statistical analyses are only performed between the candy consumer groups, and the group reporting no candy consumption is shown for comparison. Chocolate preferers had slightly longer education, and interestingly, leaner waists, lower body mass index (BMI) and less diabetes than other candy preferers. Other reported cardiovascular risk factors were similar between the groups. No candy consumers were in many respects similar to consumers, but they had more hypertension and especially more diabetes, and their plasma glucose was higher than that of consumers.
Factors related to psychological well-being and attitudes towards life are shown in Table 3. Although not all differences are statistically significant, the data are consistently better in the chocolate group than in the other candy group. Chocolate preferers rated their health significantly better (P=0.008), felt themselves happier (P=0.01) and less lonely (P=0.01), and had a better Zung depression score (P=0.02). Chocolate users also more often reported that they have plans for the future (P=0.0002).
Because the clearest difference between the candy preference groups was in having plans for the future, we further tested the difference in this item (yes/no) with multivariate analysis. Chocolate preference had an independent predictive value (OR 1.45, 95% CI 1.05–2.01, P=0.02) after adjustment for age, years of education, weight gain from the age 25 to midlife, self-rated health and physical fitness in 1974, present BMI, regular exercise (yes/no) and diabetes (yes/no). Chocolate preference also remained an independent predictor of the feeling of happiness assessed with VAS (continuous variable) after adjustment for the aforementioned covariates (P=0.008).
Finally, we tested whether the frequency of consumption would be related to well-being among chocolate preferers. In age-adjusted analyses there were no consistent relationships between psychological characteristics and frequency of use (data not shown).
The present study suggests that among old men with an average age of 76 years, chocolate preference over other type of candy is associated with better health and psychological well-being – but also with a healthier lifestyle. Our study is cross-sectional and cannot prove cause and effect, but the results may offer one explanation for the apparent health benefits associated with chocolate consumption.
A strength of our study is that the male cohort is socioeconomically quite homogenous, all men were former executives and businessmen, virtually all were home-dwelling, most of them were retired and currently married. Reverse causation – that is, sicker men would have shunned chocolate – is not an explanation for the findings, because the control group also consumed other sugar confectionary and the consumption pattern was similar. Limitations of our study include the fact that we neither had detailed information of the amount of chocolate consumed, nor information about these men's consumption habits earlier in life. Especially the latter may be very important, if we think about lifelong effects on health. Furthermore, we only asked about the consumption of chocolate as a confectionary, not as cocoa and related products. However, in Finland especially coffee and tea – not cocoa – are the principal hot drinks among middle-aged and older men. An earlier unpublished survey of our cohort showed that coffee, tea and cocoa drink were consumed by 93, 50 and 8% of the men, respectively. According to this survey, there was no difference in the use of coffee and tea between chocolate and other candy preferers (unpublished observations). The response rate in 2002–2003 was moderate, 69%, but according to the 2000 survey, the non-respondents probably included several men in bad health and with cognitive disorders possibly institutionalized and therefore the candy preferences would nevertheless have changed.
Two hypothetical explanations for the findings can be visualized. First, chocolate would have specific, favourable effects on mood and health of these elderly men. Second, and more likely, chocolate preference is simply associated with other characteristics that make the true association. A recent systematic review, however, suggested that chocolate ingredients and consequently its consumption may protect against cardiovascular disease (Ding et al., 2006). In keeping with this, in a cohort of elderly men without a history of cardiovascular disease, consumption of cocoa-containing foods was actually inversely associated with blood pressure and 15-year mortality (Buijsse et al., 2006). These cardiovascular effects of chocolate are usually related to flavonoids (flavanols), which have been associated with less cardiovascular risk in epidemiological studies (Hertog et al., 1993; Keli et al., 1996; Arts and Hollman, 2005). However, large randomized trials on long-term cardiovascular outcomes and chocolate are lacking. In our cohort, chocolate preference was neither associated with less reported cardiovascular disease nor with hypertension. Diabetes was, however, less common among chocolate preferers, but this is probably explained by smaller weight gain during midlife and lower BMI.
Chocolate has also interesting hypothetical associations with mood states and it has been claimed to act, for example, as an antidepressant (Parker et al., 2006). Although our results might seem to support this view, the mediating mechanisms are complex (Parker et al., 2006) and the effects are probably short-lived. It must be noted that in our study, two-thirds of the chocolate preferers reported to consume chocolate only ‘sometimes’, and there was no consistent relationship between frequency of use and psychological characteristics. Therefore, although we did not ask closer the present or lifelong amount of chocolate consumed, we consider that specific chocolate effects on psychological well-being are unlikely in our study population with moderate consumption. Of note, putative health effects of chocolate have also been suspected to be lost during the present commercial processing methods of cocoa, but a recent study suggested that the amount of non-fat cocoa solids (more in dark chocolates) was nevertheless the primary contributor of antioxidants in cocoa-containing products (Miller et al., 2006).
We are thus left with the more probable explanation that chocolate preference is simply an indicator of underlying psychology and healthier lifestyle, which mediates the beneficial effects. Despite the similar adult socioeconomic status, there were social differences between the candy preference groups. These differences included childhood home conditions and education, and the chocolate preferers seemed to be better off in life.
However, nutritional problems being common even among home-dwelling elderly and it might be worthwhile to study closer the possibilities of chocolate and related products to improve the nutritional state of old people. Chocolate is palatable and delicious, it has a high-energy content, and in our study it was preferred by almost two-thirds of the elderly men. On the other hand, the harms of the high saturated fat content of some chocolate products should be carefully weighed against the benefits. We plan to gather more detailed information on chocolate consumption and also relate it to frailty in a forthcoming survey of the present cohort.
In conclusion, our study suggests that chocolate preference over other type of candy is associated with better health and psychological well-being among old men. The association probably reflects underlying psychological differences dictating candy preference rather than ‘therapeutic’ effects of chocolate. This possible caveat may put the epidemiological studies of putative health effects of chocolate in a new and different light. Further studies of chocolate, elderly health and nutrition are clearly warranted.
Ethical approval: The research project has been approved by the Ethical Committee of the Department of Medicine, University of Helsinki. This work was funded by Finnish Foundation for Cardiovascular Research, the Sohlberg Foundation.
About this article
Aging Clinical and Experimental Research (2015)