**3. Results**

**Table 1** presents the descriptive statistics (mean ± SD) of diagnostic groups. First of all, the MANOVA models revealed that the interaction effects were not significant on sex, age, and education (p < 0.005). When considering the main effect of the 2D† in the MMSE score, the MANOVAs demonstrated significant effect on diabetes mellitus and depression, AD‐depres‐ sion, and AD‐diabetes mellitus, similarly MD‐diabetes mellitus, (p < 0.005). With the complete model, we can identify significant differences between the means of the diagnostic groups (F: 39.36, p: 0.000, R2: 0.614), indicating significant differences between the three dementias and cognitive performance.

For each diagnostic groups, we measure cognitive functioning (MMSE total scores) by calculating the difference of mean. We then convert the difference to a standardized effect size by dividing it by the pooled standard deviation for the diagnostic groups. These results are shown in **Table 2**. The margin of error (for a 95% confidence interval) for each estimate is shown. For example, the global cognitive functioning measured in effect size from group AD‐ depression is -3.02 standard deviation. Because the margin of error for this estimate is 0.28, the lower bound of its 95% confidence interval is -3.54, and the upper bound is -2.46. The cognitive performance in **Table 2** exhibits a strikingly consistent pattern for combination of all three types of dementia with diabetes. Global cognitive functioning is largest in the groups with AD and then decline steadily into the groups with VaD and MD.


MANOVA (complete model): sum of squares = 14913.79, gl: 18, F: 39.36, p: 0.000, R2: 0.614.

MMSE: mini‐mental state examination; 0D: group without "D", 1D group with one D, 2D group with 2D, 3D group with 3D; C‐E: mean of control group—mean of group (0D or 1D or 2D or 3D).

\* Dementia: Alzheimer (AD), vascular (VaD), and mixed (MD).

†D is dementia or diabetes mellitus or depression.

**Table 1.** Demographic characteristic of diagnostic groups.


MMSE: mini‐mental state examination; 0D: group without "D", 1D group with one D, 2D group with 2D, 3D group with 3D; C‐E: mean of control group—mean of group (0D or 1D or 2D or 3D).

\*Dementia: Alzheimer (AD), vascular (VaD), and mixed (MD).

†D is dementia or diabetes mellitus or depression.

**MMSE Ages Education Sex**

**Mean SD P Mean SD P Male**

134 30.2 3.3 **0.00** 71.8 7.6 0.009.1 5.5 **0.00** 46.5 53.5 **0.00** 0.37

28 29.2 5.1 **0.00** 71.6 7.0 0.037.0 6.2 **0.00** 52.5 47.5 **0.00** 0.67

84 30.2 3.2 **0.00** 69.4 6.7 0.008.9 5.9 **0.00** 56.5 43.5 **0.00** 0.33

10 28.7 6.7 **0.01** 67.8 8.1 0.016.9 6.2 0.35 45.6 54.4 **0.02** 0.72

25 18.7 5.3 **0.03** 78.2 8.9 0.035.8 5.1 0.68 75.0 25.0 0.07 0.35

13 21.4 4.5 0.52 74.5 10.0 0.424.1 3.7 0.76 23.8 76.2 **0.00** 0.21

8 21.8 4.2 0.95 74.3 7.4 0.907.0 6.6 0.71 75.0 25.0 0.86 0.08

24 21.5 4.2 0.48 76.0 8.5 0.246.7 6.6 0.89 55.2 44.8 0.41 0.12

19 19.7 6.2 0.53 80.0 8.3 0.185.6 4.2 0.86 57.1 42.9 0.10 0.53

10 20.8 6.5 0.37 78.3 6.6 0.235.9 5.0 0.33 45.8 54.2 0.10 0.39

14 20.8 6.0 0.46 76.7 7.9 0.416.6 6.5 0.30 57.1 42.9 0.88 0.22

7 15.3 7.7 0.47 77.3 9.8 0.229.7 6.0 0.48 21.1 78.9 0.30 0.39

27 18.0 4.9 0.67 77.8 9.2 0.456.6 5.4 0.30 26.3 73.7 **0.01** 0.07

12 17.9 6.4 0.38 78.9 7.8 0.877.5 4.9 0.26 16.7 83.3 0.13 0.31

) 12 17.4 5.8 0.51 76.7 8.9 0.225.8 6.8 0.64 41.4 58.6 0.76 0.24

MANOVA (complete model): sum of squares = 14913.79, gl: 18, F: 39.36, p: 0.000, R2: 0.614.

with 3D; C‐E: mean of control group—mean of group (0D or 1D or 2D or 3D).

Dementia: Alzheimer (AD), vascular (VaD), and mixed (MD).

**Table 1.** Demographic characteristic of diagnostic groups.

†D is dementia or diabetes mellitus or depression.

MMSE: mini‐mental state examination; 0D: group without "D", 1D group with one D, 2D group with 2D, 3D group

) 37 18.7 7.4 **0.03** 76.3 8.7 0.036.4 5.1 0.11 66.3 33.7 0.10 0.24

**(%)**

 **Fe‐ male (%)**

 **P R2**

**n Mean SD P of**

**Control group (0D † )**

520 Update on Dementia

(1D † )

(1D † )

† )

Group diabetes mellitus

mellitus and depression (2D

(1D †

)

)

)

Group depression

Group diabetes

Group AD \*

Group AD‐diabetes mellitus (2D †

Group AD‐diabetes and depression (3D †

Group VaD‐Depression

)

(1D †

)

Group VaD‐diabetes mellitus (2D †

Group VaD‐diabetes and depression (3D †

Group MD‐diabetes mellitus (2D †

Group MD‐diabetes and depression (3D † )

Group MD \*

Group MD‐depression

(2D † )

\*

Group with VaD \* (1D † )

(2D † )

Group AD‐depression (2D † )

**corrected model ages**

**Table 2.** Mean and effect sizes of the MMSE by type of dementia.

We illustrate such cognitive performance gaps in **Figure 1**, which shows differences in terms of effect sizes, that is, the difference in mean scores divided by the standard deviation of scores for all groups. When comparing among the dementia groups, we observed that patients with AD presented better cognitive performance than patients with the diagnosis of VaD and MD, while patients with MD‐depression had worse cognitive performance than patients with any type of dementias and diabetes and depression. Which confirms the finding that the worst cognitive performance is evident in the groups od MD and depression (**Figure 1**).

**Figure 1.** Hedges' g of the MMSE by type of dementia. Alzheimer (AD), vascular (VaD), and mixed (MD) and "#" is the number of "D" a group has (0 or 1 or 2 or 3), and the letter "D" is dementia or diabetes mellitus or depression.

#### **4. Discussion**

To our knowledge, this is the first study to attempt to examine the specific patterns of per‐ formance on measures of neuropsychological functioning among those with diagnosis of dementia with co‐existing diabetes and depression. We found supporting evidence showing the variation in the neuropsychological functioning between individuals with 3Ds in contrast to dementia patients with either comorbid depression or diabetes. Overall, patients with dementia with coexisting diabetes and depression had greater cognitive impairment relative to dementia only, or healthy controls. There was a non‐significant trend for cognitive scores of dementia and depression group to fall between the dementia with diabetes groups. These results illustrate the importance of controlling for depression and diabetes when diagnosing cognition. Some studies suggest that depression is a risk factor for dementia and depression treatment may be a causal factor for dementia [22]. Furthermore, depression may increase vulnerability to and/or exacerbate existing cognitive deficits [23]. Additionally, Ritchie et al. [24] examined the association between depression and diabetes and report that 36.8% of the patients with dementia have depression, while only 10.6% have diabetes. Similar reports [25] show that 23.7% of patients with dementia have dementia and depression. While the precise neurobiological mechanisms underlying depression and cognitive abnormalities in type 2 diabetes are unknown, both cognitive impairment and diabetes have been observed among older adults with major depression. In our study, patients with dementia and diabetes had lower scores on cognitive performance relative to healthy controls, which is consistent with previously reports in other populations.

We illustrate such cognitive performance gaps in **Figure 1**, which shows differences in terms of effect sizes, that is, the difference in mean scores divided by the standard deviation of scores for all groups. When comparing among the dementia groups, we observed that patients with AD presented better cognitive performance than patients with the diagnosis of VaD and MD, while patients with MD‐depression had worse cognitive performance than patients with any type of dementias and diabetes and depression. Which confirms the finding that the worst

**Figure 1.** Hedges' g of the MMSE by type of dementia. Alzheimer (AD), vascular (VaD), and mixed (MD) and "#" is the number of "D" a group has (0 or 1 or 2 or 3), and the letter "D" is dementia or diabetes mellitus or depression.

To our knowledge, this is the first study to attempt to examine the specific patterns of per‐ formance on measures of neuropsychological functioning among those with diagnosis of dementia with co‐existing diabetes and depression. We found supporting evidence showing the variation in the neuropsychological functioning between individuals with 3Ds in contrast to dementia patients with either comorbid depression or diabetes. Overall, patients with dementia with coexisting diabetes and depression had greater cognitive impairment relative to dementia only, or healthy controls. There was a non‐significant trend for cognitive scores of dementia and depression group to fall between the dementia with diabetes groups. These results illustrate the importance of controlling for depression and diabetes when diagnosing cognition. Some studies suggest that depression is a risk factor for dementia and depression treatment may be a causal factor for dementia [22]. Furthermore, depression may increase vulnerability to and/or exacerbate existing cognitive deficits [23]. Additionally, Ritchie et al. [24] examined the association between depression and diabetes and report that 36.8% of the patients with dementia have depression, while only 10.6% have diabetes. Similar reports [25] show that 23.7% of patients with dementia have dementia and depression. While the precise neurobiological mechanisms underlying depression and cognitive abnormalities in type 2

**4. Discussion**

522 Update on Dementia

cognitive performance is evident in the groups od MD and depression (**Figure 1**).

Overall, type 2 diabetes has been associated with mild cognitive deficits, most frequently in the domains of verbal memory, processing speed, and to a lesser degree, executive functioning (see review) [5]. Another study suggests that there is a protector effect of insulin on surface plasma insulin receptors, although it is possible that the improvement of cognitive function is due to better glucose control rather than a direct effect on the neurons [27]. However, meth‐ odological and study design differences, such as variations in sampling, assessment instru‐ ments, degree of diabetes severity, and the presence of comorbid illnesses, have resulted in inconclusive results. Consistent with the literature on diabetes research, research examining the relationship between depression and cognitive functioning is filled with mixed results, as mentioned above. There is evidence suggesting that the pattern of cognitive impairment varies by depression subgroup or severity (e.g., major versus minor depression) [30, 31]. In general, depression has been linked with a range of declines in cognitive domains, including memory, executive functioning, attention, and psychomotor speed [32].

The present study differs from others in that previous studies typically relied only on self‐ report for depression with varying measurement instruments. In our study, we also corrobo‐ rated our depression status with the medical prescription of any antidepressant.

The present study had several limitations. First, the specific treatments of the patients with diabetes were not verified, so that it is impossible to draw conclusions about the influence of diabetes treatment on cognitive impairment in this sample. Time with any of the conditions was not available, so the effect of short versus long term cannot be confirmed. In addition, other comorbidities, especially those affecting vascular, neuronal, or metabolic status, were not taken into account. The nature of the design does not allow for explanation of the mecha‐ nisms of the relationships observed. Another important limitation of this study is that we included only the MMSE, which is not a diagnostic instrument, to assess the cognitive function. We understand this test is widely used for its ability to follow cognitive changes over time [33]. Despite these limitations, with our large sample size, rigorous study design, and broad socioeconomic, and educational characteristics of our participants, we believe that it is possible to make valid inferences to the elderly population residing in Mexico City.

The patients with triple diagnoses of dementia, depression, and diabetes demonstrated greater cognitive dysfunction relative to those with double or single diagnosis of dementia. Additional research is needed to unravel this relationship, as to whether the cognitive impairment accrued in patients with DDD. These findings highlight the importance of identifying depression among diabetics and patients with dementia. Since depression is readily treatable, remission should lead to improved cognitive function and quality of life. The role of neuropsychology is expanding due to the increasing demand for differential diagnosis and to draw conclusions about patients' abilities to function independently. Further research is necessary to define and recognize patients with dementia with comorbid conditions.
